Updated driving impressions in snow on page 4

Well, this has taken forever. I think it was October when I proposed the possibility of using a kit based control unit for my DCCD trans. PHATsuby was kind enough to make some measurements of a STi with a 'scope, and I've used that info pretty much as my target. I ordered and assembled the K8004 kit from Velleman in October/November, but one thing or another kept me from hooking it up and "smoke" testing it. (Not the least of which was a complete specC driveline swap).

Well, I installed it last week and have been using it in some light snow / ice conditions since. So far, it 'seems' to work well.

Some disclaimers:

- I have never driven a STi with DCCD in slippery conditions, so I can't offer comparisons
- I don't own an oscilliscope
- If you try this and you burn out your DCCD coil, its NOT MY FAULT.

My background. I was trained as an Electronic Engineering Technologist. (A three year program that is quite different than the American version). I spent five years servicing Analytical Instrumentation before getting into sales roles then software (Gas Chromatographs, Atomic Absorption Spectrometers, Mass Spectrometers etc.). I have done very little electronic design.

The circuit design is from the Velleman kit, but they pretty much copied the design from the chip makers technical sheet for the SG3525 PWM chip. Rather than trying to cobble together a more targeted circuit, I went with the kit because I believe that having everything on a PCB will be more reliable.

The cicuit will vary a 12volt Pulse Width Modulated waveform between 0 and 100 %. It can drive up to 6.7A before circuit protection kicks in. I chose to use a frequency of 1kHz, but the device can range from 100 Hz to 5 kHz.

First, the circuit:

http://www.we-todd-did-racing.com/wetoddimage.wtdr/wNzk3OTc1NnM0MTNkZmQzMXk1NDE%3D.jpg

I used the Subaru thumbwheel. It was about $120 Canadian, by far the most expensive part of this project. I took it apart to inspect it, and found that its made by Nissan. It looks exactly like the dash dimmer switch in the Nissan Altima; except for the DCCD print on the face of the switch. I'd lay odds that the Nissan part would be half the price and would behave the same.

I assembled the kit (took me about 90 minutes going slow). Then I went about building connectors and wiring for installation in the car. This was a bigger problem than I wanted. I went through four Molex connectors and was unable to get reliable pin connections, even after buying a Molex crimp tool. I gave up and just ran wire from the circuit board to the connections in the car.

http://www.we-todd-did-racing.com/wetoddimage.wtdr/wNzk0MTQ4NnM0MTNkZmQzMXk1NDE%3D.jpg

There are seven wires. Three for the thumbwheel, two for power, and two for the DCCD leads. I used an inline minifuse rated at 2 amps.

Results.

In a parking lot, with DCCD wheel set to 'open' I had massive rear wheel spin. I moved the wheel to 1/3 lock and tried again. It just dug in and went! I haven't played with full lock yet because I want to work up to it. I'm nervous about sending too much current to the coils and there seems to be some difference of opinion on what duty cycle equates to full lock. Right now I'm happy.

Rather than waiting till everything was perfect, I thought I'd post now. Otherwise it may never happen!

My next project will be to incorporate some other features like brake / handbrake defeat so the diff will go open, and possibly a more "active" controller with inputs like throttle position and speed. Who knows, maybe even an accelerometer ;)

That truly is awesome!

I know that in Chekoslovaquia someone is making this type of Control for Rally cars.
It as 8 positions of Front/Rear traction, than it as wires to Brake pedal,Hand brake and Accelerator. It costs a lot more than you spend. Congratulations.

Thanks guys.

WOW! Great job man, that looks awesome. Sounds like you know your stuff.

I'm currently a Computer Engineer. One of the fields I'm interested in going into is embedded systems for Automotice Applications. Interesting stuff. I assume the coil is used to provide a magnetic force in the DCCD to move parts to transfer power, or something like that?

awesome fragment! congrats! thats sweet work man, props on being the first to take the leap. Be careful man, i still want to double check those measurements. I am trying to see if PDX will take some measurements since they do such thorough work, then we could be sure about the frequencies and voltages.

how long have you driven around with it on 1/3 lock?

also did you try it at only 5v? the readings i took the max voltage was 5, i dont know how much it matters.

congrats again.


Ben

Nice job! It's awesome to see somebody come up with a home-grown alternative that works. You smart guys make me sick. You want to trade brains? I have some awesome memories of a summer spent in Bavaria with an Austrian girl that I'll throw in for free! :)

Your words of caution are heard. To be honest, I also assembled clues from others like patr (Patrick Richard) that have more real world experience than I. I would love to have the use of a scope for a weekend to thoroughly map the STi's behavior, and to map my beast in its image, but sometimes just doing it is better than waiting too long to do it 'right'. I'm hoping for snow and ice so I can test more.

Nice job! It's awesome to see somebody come up with a home-grown alternative that works. You smart guys make me sick. You want to trade brains? I have some awesome memories of a summer spent in Bavaria with an Austrian girl that I'll throw in for free! :)

Hmm. Tempting, but then my wife would splatter my brains all over. Don't want to waste that memory like that.
:banana:

Congrats Fragment. Good to hear that someone is making an alternative controller. Though I have driven a car with the GEMS a bunch of times, I couldn't go back to a manual control or viscous center diff. Having the capabilities of the center diff being "active" , with percentage lock-up on decel/ accel , speed/rpm , makes so much difference in the world. I can only partially understand how Wrc cars feel with full active front,center, rear diffs.

Ver.III can you expand on what changes under what conditions? ie, under lift throttle conditions, does the diff go all the way open? What frequency is commonly used with GEMS? Anything could be helpful.

The best person who knows the GEMS and how it works would be Pat. He understands it alot more than I do. The way I look at it, The 04/05 Sti has a good "all around" map program. The GEMS is fully mapable as like a stand alone engine management. You can change any parameters to suit your driving style or specific conditions.

For snow , using engine rpm or speed input, at low speed/rpm you can map out the center diff to reduce percentage lock-up (ie 20% or full open) to allow the front wheels to gain traction and allow better turn-in. This is an avantage compared to a viscous/plated center diff because in this case, you would have the deal with the diff preload. On exit of the turn, you can program full lock up when you go 70-100% throttle. Likewise, for left foot breaking you can program xx% diff lock-up .

just as a tip, the STi service manual has a description of its operation in various conditions, it even has a sort of "map" of how it works.

Ben

I made one of these a couple weeks ago using a different schematic I found online for PWM controller. Works great.

I'm pretty sure 1khz is way too fast for a coil you are probably saturating it and its basically fully locked whenever you have it turned up even a little bit.

I'd turn down your frequency as close to 100hz as possible.

I used a Techtronics TDS210 scope to measure the frequency of the PWM.

http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?ViewItem&category=33733&item=7956600413&rd=1&ssPageName=WDVW

Here's a link to ebay for buying one already made up.

Spiider, from what I've read, the controller you measured was the ver6. PHATsuby measured a ver8 controller at 1kHz. patr told me that he uses different frequencies to adjust the responsiveness of the diff for different conditions.

Interesting find on the ebay item. I guess I just saved myself $349. :)

This is cool. I wonder if something like this can be adapted to the 4EAT....

spiider what transmission did you scope? I would be interested to see what yours looks like as well. Have you had any sales yet? You can actually just buy a factory dccd controller used(yes it is hard to find) for like 200 and have the same ability and be sure that it works.

I would love to see your scope results, maybe you can answer one question, did you use 12v as well?

thanks
Ben

This is cool. I wonder if something like this can be adapted to the 4EAT....

No. Totally different.

Does anyone knows if that Ebay controler is any good?
Who makes it?

Looks like Spiider makes the ebay controller. Unless there are two guys in Guelph running '02's with ver8 swaps and handmade DCCD controllers :lol:

Yep, I'm making them and selling them. They cost not too bad for parts but they take a couple hours to make up and properly test, etc. So the cost of production gets up there pretty quickly.

Based on demand so far and what they cost me to produce $350 is pretty decent for a nicely built and tested product that works, if you can get a complete v5 or v6 manual controller for $200 I'll take a couple for people I know too. The manual Rallispec unit is $700 USD.

They have been selling well at $350 but I'll lower my price to $249.99 to be a little more fair to everyone and then maybe more people would have access to them that need them.

Luckily due to Planetmotorports here in Guelph I have access to lots of trannies to scope, from Ver5 - Ver8, I'll make some more measurements of different versions and see what comes up. I can tune my PWM output to whatever hz is required for a specific application but the 100hz works fine on v8.

I still think 12v 1khz PWM is going to saturate the coil even with a very low on time.

so what voltage are you running to the coil then spiider?

I am sure you have done some testing but you say that it is a tested product that works, does that mean it works correctly? Again, not trying to put you down, since I myself havent tried it out, I would just be weary to buy a DCCD controller that has only been tested on your car, and if it truly is running proper voltage and frequency to make sure no damage is done. How long have you been running it on your car and do you just leave it 50/50 all the time?

thanks
Ben

The voltages I measured were straight off the coil. The signals are 12-13v PWM 100hz, when you turn the thumbwheel, the time duration is what changes, not the voltage. I don't know where you got less than 12v from or the 1khz, but it doesn't sound right to me.

Additionally, if you are trying to use a multimeter then you won't be able to measure the voltage of PWM. I have a $700 Fluke 87 III that can measure the frequency in hz very easily but just doesn't have the speed to pick up the PWM signal. For that you have to use an oscilloscope and as I said before, I used a Techtronics TDS210.

I measured from pins 20 and 40, which are the wires going directly from the dccd controller into the center diff. Also if you look at the factory manual, which i would assume you have, it even lists that full lock is 6-7v, the highest i read was just over 5v at full lock while the STi was parked. I would assume there is error in the frequencies taken, i am not denying that, but it should have least been reading correct voltage.

I used a digital oscillioscope for mine, but at the time had just got it, i didnt know the settings it was on, and when i received a frequency(which seemed to be somewhat close to what people expected, remember no one had done it up to that point) i just recorded those.

I am just trying to pool info here to make sure the correct controllers are made, i know you have a product to sell, which you want to somewhat safeguard, just trying to get all the info straight and procedures used so i can do it as well.

thanks
Ben

Hey what scope are you using? Maybe I can help with setup and reading everything properly. I intend to scope a few more models and see what I get.

The current ciruit design I'm using allows me to vary voltage or frequency if I need to for other versions.

As far as further development maybe Fragment, you and I can work together. I would like to add some more advanced functions such as throttle inputs, and maybe an aftermarket lateral gforce sensor.

email me: jeff_reid@sympatico.ca

The nice thing about the circuit I'm using is that I can change these values if I want. Changing the frequency is a simple pot turn, and changing the max duty cycle is also do-able. I just need a stinking scope :( I'm thinking about buying the Velleman scope kit for a couple hundred. Its pretty basic and requires the use of a PC, but the specs will more than cover whats needed for this project. But I'm tapped out these days. Paypal donations accepted :) stage4wrx at hotmail dot com

For the greater good of course!

I still think the max duty cycle would be close to 50%, not 100 or 95.

spiider, thanks for the offer, i will have to check a little later, it is in my car. I would be able to check with the manual but, unfortunately i left that at my friends as well, LOL. I guess that gives me a reason to go back over and get it and take measurements, now i just have to go do it.

Other inputs would be sweet, we would need programming and hardware knowledge, which i do not have but know a couple people that do know a good amount.

Ben

here is what i got, who knows if it is sufficient, i hope it is, haha.

http://www.professionalequipment.com/xq/ASP/ProductID.911/id.448/subID.449/qx/default.htm

Ben

It looks a bit small, but I think that should do it.

Here's my scope.

http://www.tek.com/Measurement/cgi-bin/framed.pl?Document=/Measurement/Products/catalog/tds200/eng/&FrameSet=oscilloscopes

You should have your channel 1 scope lead attached to the coil positive, and the ground for that lead hooked onto the coil negative or car ground.

I found a 2 axis lateral g-force sensor to add to it, but its about $250 just for the sensor, and then there's the cost of integrating it to the unit properly to take all the inputs and have some programmability.

how do those work? since its 60 or 100hz does it only pick up those frequencies? I would assume no since that doesnt make sense, but was curious about that when i searched before.

Yea, with the combo of adding the price of that sensor, plus the time/money to program something may be a waste of time or a bad investment if you are doing it for that aspect.

Ben

I think a g-sensor would need to be in version 3 or 4 of a controller. Getting something to accept analog inputs and apply logic / table lookups and then translating that to a pulse width would be a large undertaking in itself. The brake / handbrake thing could be done with a couple of diodes tying into the "soft start" line on the IC. Then the next level of complexity would be having the unit programmable.

how do those work? since its 60 or 100hz does it only pick up those frequencies? I would assume no since that doesnt make sense, but was curious about that when i searched before.

Yea, with the combo of adding the price of that sensor, plus the time/money to program something may be a waste of time or a bad investment if you are doing it for that aspect.

Ben

60 "M"hz. (mega-hertz)

A scope is just like a very high speed multimeter, with a "time graphing" function if you want to put it in simple terms. You can slow your readings down enough to see what happens in a short time period.

I think a g-sensor would need to be in version 3 or 4 of a controller. Getting something to accept analog inputs and apply logic / table lookups and then translating that to a pulse width would be a large undertaking in itself. The brake / handbrake thing could be done with a couple of diodes tying into the "soft start" line on the IC. Then the next level of complexity would be having the unit programmable.

Yes, we'd have to do what we can in a simple circuit first before we could even think of added programmability.

yea, my bad i did not see the "M", haha.

Ben

Well, just to be on the safe side, I've dialed my freq down to ~100 Hz and reduced my max duty cycle to ~50%. The nice thing about having a well documented IC is that it gives detailed specs on what control voltages = what outcome. So my little DMM can get me there just fine :)

Just as a data point, my inline 2A fuse held ok till about 70% duty cycle at 1 kHz before letting go. So overall its a pretty effecient box.

fragment did you notice any lockup difference by changing those parameters?

thanks
Ben

Sorry, no. I've had dry roads for a while now and I haven't been able to get meaningful feedback on the traction differences. I'm not about to drive hard enough to break the back end loose and drift the twisties on public roads in the dry. As much as I want to :devil:

I think I've noticed a little less "bind" in tight slow speed turns though.

Subscribing

lol...didn't read all of the thread...when talking with one of the guys at competition he said he had designed the one for the car and was making one for his subaru with sti swap...so I put 2&2 together :o

HomemadeWRX; Umm, huh?

Yep, I'm making them and selling them. They cost not too bad for parts but they take a couple hours to make up and properly test, etc. So the cost of production gets up there pretty quickly.

Based on demand so far and what they cost me to produce $350 is pretty decent for a nicely built and tested product that works, if you can get a complete v5 or v6 manual controller for $200 I'll take a couple for people I know too. The manual Rallispec unit is $700 USD.

They have been selling well at $350 but I'll lower my price to $249.99 to be a little more fair to everyone and then maybe more people would have access to them that need them.

Luckily due to Planetmotorports here in Guelph I have access to lots of trannies to scope, from Ver5 - Ver8, I'll make some more measurements of different versions and see what comes up. I can tune my PWM output to whatever hz is required for a specific application but the 100hz works fine on v8.

I still think 12v 1khz PWM is going to saturate the coil even with a very low on time.

Sorry, no. I've had dry roads for a while now and I haven't been able to get meaningful feedback on the traction differences. I'm not about to drive hard enough to break the back end loose and drift the twisties on public roads in the dry. As much as I want to :devil:




nothing an empty parking lot and some oil on your tires won't fix :devil:

p.s. just watch for flare ups :)

nothing an empty parking lot and some oil on your tires won't fix :devil:

p.s. just watch for flare ups :)

If I didn't know better, I'd say that sounds like the voice of experience :lol:

If I didn't know better, I'd say that sounds like the voice of experience :lol:


only in fwd POS cars...never had the guts to do it in suby.

I'm interested in helping on this project with any programming that might be involved. I have a bit of experience with programming microcontrollers and talking to the subaru ecus. I have a v8 six speed with an open dccd diff and it's already got me in trouble once in the rain :)

Anyone know offhand how many sensor inputs would be ideal for this? I can easily make a device that's as smart as I am about differentials but that's not saying much.

Help with programming would be great! But first we need to compile a list of components / strategies for the hardware design. I'm open to suggestions on all aspects of it, but I was thinking of some variation on the 8051/8052 with integrated I/O capability, digipot for digital control of the PWM chip, a/d convertors for things like TPS, and maybe the ADXL202E for a two axis accelerometer

If we wanted something VERY simple, we could use the 4 ABS wheel sensors as inputs, and just have it change split if it has fronts or rears going faster than the others. Like progressively lock it up(or unlock) proportional to wheelspin maybe?

Making an in depth map with many inputs would be sweet, but we would need someone very educated to make it work smoothly.

Ben

I think we have to go back to basics first and consider that all we are doing is tossing control from the rear tires only to the rears and the fronts together.

I think we need a 2 axis accelerometer/g-force sensor (analog), and a hookup to the TPS (analog) for comparing whats happening in real life with what the driver is trying to do.

Also, maybe compare the rear wheelspin with the front wheelspin via abs sensors and see if we can control wheelspin at the rear by increasing the lock to the front.

It may be advantageous to check the 2 axis accelerometer to see if the driver is turning, then we could partially/fully unlock the front to help the car oversteer/understeer, we could just sample the car yaw angle and do a scaled/mapped response to lateral g-forces.

TO do all this sensor-wise we are talking about;

Analog's:
1. accelerometer longitudinal axis
2. accelerometer transverse axis
3. TPS

Digital:
1. abs fr
2. abs fl
3. abs rr
4. abs rl

I agree. In fact, isn't that pretty much what I said except for the ABS? ;)

I think we need to do hardware first, then worry about logic etc. Are you suggesting a microcontroller isn't needed?

Cool keep up the good work!

Actually, there is a company in Toronto that makes the controller for the DCCD. Their basic one takes inputs from few sources ( I think- brake, h.brake, tps and speed), and is programmable. I will try to find a picture of it, or a link to it.

Cool keep up the good work!
ditto

interesting project...

good luck!

I got all excited because it snowed Saturday night, but by the time I got outside it all turned to water.

So much for that testing opp.

Ok, I found the web address for the other DCCD controller.
here is the link www.neetronics.com it is in the newsflash section.

Its amazing.

fragment, phatsuby and I have been experimenting with our own homemade ones that do almost everything this one does for a couple months, then this one comes out and its groundbreaking earthshattering news because someone's making one.

Holy crap.

You can make our homemade ones for a hundred bucks, plus the cost of the thumbwheel.

Yeah, amen to that. Also, I don't think that kit includes the thumbwheel. I looked at all the pics and the description of the kit and its not anywhere to be seen.

I'm glad someone is making one, but for $300 US, I'm glad I made my own.

If I was racing, the programmability would be very nice but I'm not.

Why would it saturate at a faster switching frequency? Lower switching frequency = more chance of inductor core saturation.

Why would it saturate at a faster switching frequency? Lower switching frequency = more chance of inductor core saturation.

You just gave me a reason to make an account on this forum. At a low switching frequency you run basically DC through the coil, saturating it. Identical to "clipping" in speakers. This is bad in ANY system that contains a coil... unless it's the heating element in your hairdryer.

That was always my understanding as well.

You just gave me a reason to make an account on this forum. At a low switching frequency you run basically DC through the coil, saturating it. Identical to "clipping" in speakers. This is bad in ANY system that contains a coil... unless it's the heating element in your hairdryer.



So how do we find out the proper amount of power to pump through this thing? Can we figure it out just by looking at an OEM STi box in action or what?

I'm ready to sit down with a microcontroller and make a super cool DCCD controller with digital controls and a display... I'm just weak on the hardware side.

there's a difference between saturating a coil (which drives the inductance down sharply, generally increasing current through it), and saturating the device which the coil is driving. ie. you can attempt to actuate the coil at such a frequency that its mechanical inertia prevents it from changing state rapidly enough to track the electrical state.

I actually posted something like that last night when the board shut down at 5pm, and it wouldn't take my post.

I'm not thinking along the lines of inductance we have that covered with the gate and anti-inductance diodes in the circuit to keep it from being damaged from reverse emf, the coil may heat up if the rate is too low though.

When I posted this weeks ago, it was because my oscilloscope measurements on a newer STi showed 100hz, PWM range 5%-95%, ~13.6v. So, thats what I produced for a controller, a copy of the STi's measurements.

I'm thinking along the lines of allowing the coil time to modulate the mechanical side of things properly. I'm not sure that the coil would have time to energise/de-energise fully that quickly and it would probably just act as if fully locked. fragment said he made one at 1khz, and I just threw out that "saturation" statement as a guessimate.

now i want a dccd to play with too... ;)

Thanks for clearing that up Spiider. I'm amaized that they'd use such a low frequency though. Normally a frequency over 20khz (above audible) is used in such applications to keep the mechanical and electrical resonance frequencies as far apart as possible. Does anyone know what RPM range the center diff might encounter from 0 to 150mph? I wonder if that might have decided the PWM frequency to reduce resonance.

Also, does anyone know the difference in the electromagnetic clutch used to lock the center diff between what's in the normal WRX and the STi? Is it the same?
I'm an electronics guy, just getting into cars.

We need to be clear that we have two frequencies that have been measured. Spiider measured 100 Hz on a older manual DCCD controller (V6?) and PHATsuby measured 1 kHz on a USDM STi. Although PHATsuby isn't 100% certain his measurement is right, I think it raises a valid question about what the new controllers are using. We need more data.

Anyone out there with a scope and STi? I'd love to see actual traces from a PC based scope with the time base and amplitude at different wheel settings.

It's 100Hz for everything.

Unless you are driving in the snow or on marbles or super super agro I doubt you'll notice anything between the various lock-ups - aside from turning a tight circle and scrubbing the wheels.

That's awesome, fragment...how much did it all cost you?

My costs in Canuck bucks:

Thumbwheel: $125
Velleman Kit: $35
Misc: $25
Labor: $800 (me, but I'm worth it :) )

So hard costs were $185 CDN or about $150 USD

It's 100Hz for everything.

Unless you are driving in the snow or on marbles or super super agro I doubt you'll notice anything between the various lock-ups - aside from turning a tight circle and scrubbing the wheels.

At Race City road course last fall, with the center diff open and an R160 rear, I was spinning the inside rear tire on the exit of a 4th gear turn for almost 40 yards. Thats with fairly big rubber too: Falken ST115's in 235/45/17

Are you saying that I woulnd't notice 50% lock vs. full lock?

I'm not thinking along the lines of inductance we have that covered with the gate and anti-inductance diodes in the circuit to keep it from being damaged from reverse emf, the coil may heat up if the rate is too low though.

I'm thinking along the lines of allowing the coil time to modulate the mechanical side of things properly. I'm not sure that the coil would have time to energise/de-energise fully that quickly and it would probably just act as if fully locked. fragment said he made one at 1khz, and I just threw out that "saturation" statement as a guessimate.

The proper term is freewheeling diode. Not anti-inductance....

At 100Hz the ripple on the inductor is very very low. Like way less than 1/4 amp. This means that the inductance is quite big... well let's calculate:

V = L di/dt so L = V/(di/dt) = ~ 14V/(.25A/.005sec) = 280mH.

There is little need to switch faster in this case.

At Race City road course last fall, with the center diff open and an R160 rear, I was spinning the inside rear tire on the exit of a 4th gear turn for almost 40 yards. Thats with fairly big rubber too: Falken ST115's in 235/45/17

Are you saying that I woulnd't notice 50% lock vs. full lock?

I think that qualifies as pretty agro driving :) You'd notice I difference. I think. That's cool. Where's your rear LSD?

At 100Hz the ripple on the inductor is very very low. Like way less than 1/4 amp. This means that the inductance is quite big...

I was wondering why you're guessing 1/4th amp? Is this an assumption?

I'm curious, does the STi DCCD have the power driver in it? (a MOSFET) Can an anyone answer this? If so, put the part number of the device up so we can check the maximum possible current going through it.
... or..
if someone with an STi wouldn't mind using a curent meter to check actual current that would be great.

PHATsuby used a current meter when he did his testing. Less than 1 amp is all I remember, and I'm too lazy to search right now.

Where's your rear LSD?

At the moment its bleeding RedLine Heavyweight Shockproof all over my garage floor. I've got a complete R180 in now.

Most of this would be fairly easy to convert from a fairly dumb PWM generator over to something like a PIC microcontroller, or for the assembly challenged like myself, a PICAXE chip. I'm a complete amature electronics hobbiest, but I can easily implement the picaxe, reprogram quickly and easily without having to remove the chip for flashing, via serial cable. Easy LCD display interfacing, datalogging, plenty of input/outputs, dedicated PWM timers, etc. And CHEAP!



http://subaruwagon.com/albums/album03/small_DSC02118.sized.jpg

I was wondering why you're guessing 1/4th amp? Is this an assumption?

I'm curious, does the STi DCCD have the power driver in it? (a MOSFET) Can an anyone answer this? If so, put the part number of the device up so we can check the maximum possible current going through it.
... or..
if someone with an STi wouldn't mind using a curent meter to check actual current that would be great.

I measured it. 4A with a 1/4A ripple at 95% dutycyle. 2A at 50% etc. etc. You don't need the part number to select a 4A FET.

I measured it. 4A with a 1/4A ripple at 95% dutycyle. 2A at 50% etc. etc. You don't need the part number to select a 4A FET.


So generating a 5v pwm signal from a microcontroller is cake. We shouldn't need a custom pwm chip just a circuit to convert TTL to 12-13v 4amp signal...

So generating a 5v pwm signal from a microcontroller is cake. We shouldn't need a custom pwm chip just a circuit to convert TTL to 12-13v 4amp signal...

John, we miss you in #mwsoc! Have you ever toyed around with the picaxe chips, or are you too cool for a BASIC programed microcontroller? :p I think they are the cats meow compared to Basic Stamps and similar multi-chip designs.

PWM output is literally a single line of code in Picaxe basic. Easily specify the period and the duty cycle. Timing is all done via internal resonator, so its not uber accurate but for something like this is would work well.


Now all I need is a DCCD tranny to play with...in my dreams :lol:

An update. We've had snow conditions for the last few days, and I've been playing around with settings. First, I know for sure this is working. With the thumbwheel set to "off" I can induce biiig slides in easy corners, and stepping off from a stop can spin the back with just mild acceleration. By moving the thumbwheel up to mid-point (should equate to 25% duty cycle) the car still feels "rear wheel drivish" but the slip and spin behavior is greatly reduced. Full up on the thumbwheel (should be 50% D.C.) means I have to go to boost to induce tire spin, and it "seems" to be all four at once. Moving from a stop it just grips and goes. All good.

But...

Understeer is now in play; big time. At half mast, I get push right after turn in but I can control my line with more steering lock. At full twist, it just wants to go straight. Power on, power off, whatever. It just wants to plow a line straight ahead. I can't figure out why this behaviour is so pronounced from a tight centre diff, but it is.

So next up is a modification for brake and handbrake defeat and if I can figure it out, zero throttle defeat as well.

Just need to wire the handbrake light, the brake light, and TPS into the controller (I believe it's 5V for zero throttle) that will make the controller go to 0% Duty Cycle

Understeer is now in play; big time. At half mast, I get push right after turn in but I can control my line with more steering lock. At full twist, it just wants to go straight. Power on, power off, whatever. It just wants to plow a line straight ahead. I can't figure out why this behaviour is so pronounced from a tight centre diff, but it is.

i'm not surprised.

the thing is that the front wheels, as a whole, travel FURTHER than do the back wheels while turning. to see this just drive through a puddle and watch the tracks.

by locking up the center diff, you're forcing the rears to push the fronts, which are going "too slow." since you're also simultaneously asking the fronts to provide lateral force to push the front of the car around a corner, they end up losing tracting more quickly. the result is classic understeer.

this is also the shortfall of the stiffer VC center locker, ie the 20kgf/100rpm units. stock wrx is 5kgf.

hth
ken

That's kind of what I had in my head, but a very good explanation.

fragment, if you're interested in a TPS controlled binary output, check out my www page. i have a TPS switch on there that taps into the oem tps wire, and is a simple comparator. when the tps voltage exceeds a setpoint, the output state changes. as it stands it is a 12v output. obviously you can have it either NO or NC below the setpoint by changing the relay connections.

here's a direct link: http://ken-gilbert.com/wrx/TPS.pdf

hth
ken

Man this was something to read from the begining :eek: So I will mark the spot were I left with this post :p

Thanks, I just might use that. The real question in my mind is what would be a good setpoint? I would say something like 30% throttle, but what about those situations where you want to maintain neutral throttle position while in-corner? I can see where working out the logic could be a full time job. Guess that's why Subaru has all those engineers.

fragment, if you're interested in a TPS controlled binary output, check out my www page. i have a TPS switch on there that taps into the oem tps wire, and is a simple comparator. when the tps voltage exceeds a setpoint, the output state changes. as it stands it is a 12v output. obviously you can have it either NO or NC below the setpoint by changing the relay connections.

here's a direct link: http://ken-gilbert.com/wrx/TPS.pdf

hth
ken

any update on your progress I am really interested.

Sorry, no real change this winter. I have been using the manual mode for really deep snow conditions and it works fine, but I really really need the controller to open up the DCCD under braking and no/light throttle to be really useful.

I'm still developing the new microcontroller based DCCD controller. I have a working prototype done, I just need some time (and money) to take it a little further.

I should have something more complete by summer.

so what do you guys think about the neetronics controller is it worth the money.


thanks

Personally, I like it.
I have seen it on a couple cars, and they have done a good job with it.
It is a little pricey, but not ridiculously expensive like a few others and has a good mix of features, fair value for what you pay.

cool thanks

just found this thread and will be sure to follow it as i'm all about DCCD but am more than broke right now so the oem setup is completely out of the question...

Just an update for the many people who have emailed me asking about the progress on the automatic version...I'm running the microprocessor driven version in my car and actively developing and tweaking the firmware for my microcontroller based design.

I have also worked out interfacing to the stock ABS G-Sensor, which should lower the costs dramatically for an "active" system. (Previously I was planning to add an expensive external sensor in order to do this.)

FYI, I'm still producing the manual "version 1" homemade controller for current and future orders, and I will be offering a second product the automatic "version 2". (Many customers have only been interested in a simple and effective manual unit so I will continue to support and sell those as well.)

I think it will be great to offer a second more sophisticated level of product.

Expect to see it up in my ebay store in the next few days/week. I still have to add up the economics of production costs, but I'm sure it will not exceed $250 USD.

thanks for the update...
how does this respond to the abs g-sensor? and are you tying in any other sensors to the automated version?
just interested to get an understanding of the differences...
bryan

thanks for the update...
how does this respond to the abs g-sensor? and are you tying in any other sensors to the automated version?
just interested to get an understanding of the differences...
bryan

The stock ABS G-Sensor is a fore/aft unit, I sample its output using a 10bit A/D convertor, and do some proprietary math for offset and scaling in order to input it into the calculation of the appropriate duty cycle for the DCCD. (It isn't lookup table based, I developed my own formulae to calculate it on the fly.)

What it works out to is a smoothed and scaled response to acceleration/deceleration. Acceleration causes lockup to be increased, deceleration causes the DCCD to go open. There is a manual override setting to lock it 100% if you are stuck or something. I wanted to keep the interface simple and intuitive so it is still based on one single adjustment knob.

Currently, I have achieved a good portion of what I wanted to do. The only other input I am interested in adding in the future is TPS. (I'd like to do a comparison between TPS and acceleration.)

I have plenty of digital I/O and A/D left so if anyone has input at this early stage feel free.

With the current firmware in a street application, the handbrake cutout, and the low speed cutout are definitely not needed. The DCCD isn't doing much in my firmware at low accel/deccel anyway.

I may need to tweak things to be more aggressive for rally or AutoX, but I will work with those people to develop firmware programs more specific to them.

so you currently do have a handbrake and low speed cutout? or you're saying you dont have them and dont feel they are necessary either?
as for the abs sensor, couldnt the same thing be achieved through the tps? or does something vary between the two inputs that i'm not aware of, and combining the two will be more beneficial than just one?
hope you dont mind the questions, if anything it'll help explain it to everyone...
bryan

The handbrake, brake peddle and vehicle speed sensor just aren't necessary with the way this system is working.

With the latest firmware I currently have no trouble with the car fighting me during parking or low speed maneuvering. The handbrake affects the car in terms of decelleration, as soon as the car see's a certain amount of decelleration it begins rapidly reducing power to the DCCD.

However I am interested in adding TPS, if only to determine lag between the driver input and the g-forces generated by the car. If the lag gets big I can try to help by ramping up the DCCD more aggressively.

I did some testing/tweaking yesterday in the cold wet weather. I had no problems spinning the rear tires unlimitedly with the system off, with it on the fronts easily maintained control and consequently I was able to accelerate much quicker.

I will probably test this out over the next couple days, it is a easy addition electrically speaking and some firmware changes to work out. Part of the beauty of this system is that I can change firmware very easily for testing or for future customer upgrades if I add later functionality.

nice info i'll be making one soon.

jerry

free bump because it sucks in the winter with a rwd suby.

as soon as the car see's a certain amount of decelleration it begins rapidly reducing power to the DCCD.

spiider, would it be advantageous to keep the dccd close to locked during high g deceleration? i know that stiff diffs actually HELP braking (at least in a straight line) because they prevent any one wheel from locking up.

on the other hand, this could be disasterous if done in a decelerating turn.

just some thoughts. keep up the excellent work!

ken

did you get one of the auto dccd controllers and if so how does it work.

Ideally, I'd prefer to have two knobs (one for brake and one for throttle). Each knob would have it's own set of 'curves' which can be selected for % lock. I think even 5 points of adjustment per knob would be fine. This would let me set up the car's attitude during corner entry and exit and adjust it on the fly. 0% brake = 0% lockup, 100% brake = 50% lock up etc.


Brake% 20 40 60 80 100
Set1 0 10 20 30 35
Set2 0 20 30 40 50
Set3 10 20 40 50 50
Set4 20 30 40 50 50
Set5 20 40 50 50 50



TPS% 20 40 60 80 100
Set1 0 10 20 30 35
Set2 0 20 30 40 50
Set3 10 20 40 50 50
Set4 20 30 40 50 50
Set5 20 40 50 50 50


And for situations of left-foot braking, the lockup % will read the values between the two maps for brake and throttle and apply the following logic. If you're applying enough brake that the map calls for 45% lock up while you're applying enough throttle that the map calls for 25% lock up, the final lock up will be the absolute value of (throttle - brake). In this case it will be |25-45| = 20% lockup. Let me know if I should explain this better because I'm not sure I did.

Also I vaguely remember reading somewhere that anything above 50% duty cycle means full lock so is there a need to run 100% duty cycle?

I don't run up to 100% duty cycle in my units for keeping the temperature down more than anything.

If you run 100% you might as well just have a switch to connect power directly to the dccd coil. I have recommended this for temporary measure to people such as for short term duration of dyno pulls, immediately to be shut off after the pull.

I go as high as 75% duty cycle in my unit, which makes for a much stronger locking force than 50%. My dccd controller is a bit "wider" than the stock STi model controller as I also go all the way down to 0% duty cycle so you can effectively run open if desired.

I like the idea of an adjustable second input so as to adjust the dccd response during braking but I'm not using lookup tables for setting the duty cycle, I'm taking a unique "vector" based mathematical approach so that the microcontroller calculates the required lock on the fly. The control knob is used to scale that mathematical response curve for a more or less aggressive traction based on throttle position and g-sensor readings. The microcontroller sampling is very fast so there is no lag between throttle position and center diff reaction.

My thought process on the braking response was to let the dccd go open based on the level of negative g's sensed during braking. High negative g's would open the diff fully, lower negative g's would not open it as much. Under heavy braking the diff should go open to allow the abs to do its job, whereas under moderate to low braking g's traction could be maintained by keeping some locking force.

The auto dccd mode does a constant comparison between throttle position and g reading's to see if you are holding the throttle down and if the car is or is not accelerating in response to throttle.

I'm away for about 6 weeks overseas, so I am delayed once again with the final release. It's a part time development process with me right now as it is the most time I can currently dedicate to the project. Too much of my primary work getting in the way.

My thought process on the braking response was to let the dccd go open based on the level of negative g's sensed during braking. High negative g's would open the diff fully, lower negative g's would not open it as much. Under heavy braking the diff should go open to allow the abs to do its job, whereas under moderate to low braking g's traction could be maintained by keeping some locking force.


I'm not fond of that. I'd prefer the exact inverse of that. Where light braking results in an open diff and allows the car to be trail-braked. And a tighter setting when you're slowing down from high speed where you want more stability than light braking. Can you incorporate vehicle speed into yoru formula somehow? I think it's pretty important for the car to get more stable as the speeds increase.


temp full locked mode - Can i run any 12V to the solenoid? Does it matter which pin is 'in' or 'out'? What about an amp rating? Can you recommend me an existing 12V wire in the car which I can splice and loop the DCCD solenoid with an on/off switch?

^^^^^^ that is a great question ^^^^^^

When I said that you could run 12v direct to the dccd coil what I meant was as a last resort for a couple minutes at a time to do dyno pulls only.

If you are stuck in this rare situation of having dyno time booked and no dccd controller installed I would hook into the power mirror/heated seats circuit, there are lots of amps available there and nothing else on the circuit breaker.

Your dccd coil is going to get HOT from this, hence why I don't recommend it for that 2 hour long winter commute home from work unless you like to burn up center diff's just for kicks.

There are only two wires going into the dccd coil, it doesn't matter which is positive or negative (neither wire is grounded).

The dccd coil has very low ohm's, usually ~2 ohms, so at 12-14 volts you are going to be pulling 6-7 amps from the circuit perhaps climbing higher as the coil heats up and its resistance increases or decreases.

The dccd coil has very low ohm's, usually ~2 ohms, so at 12-14 volts you are going to be pulling 6-7 amps from the circuit perhaps climbing higher as the coil heats up and its resistance increases or decreases.

the R would increase with T, so current flow would drop over time as the coil heated up.

this may be a long shot,
but is there any way to incorperate a steering position sensor and add an additional Yaw sensor to help further contorl the DCCD?

Spiider - Got it about the coil heating up. This temp mode will only be used during short stints in open snow filled parking lots for a couple of minutes at a time or while executing a launch in the rain. I tried launching last night in the rain with the DCCD open and I don't think I got off the line any faster than a RWD car.

And the following question is for Spiider or anyone else than can answer it. I am going splice a wire from the power mirror/heated circuit. I should look for the 12V lead and splice a wire from that to one of the pins on the DCCD solenoid. I will put a switch on this wire. But what do I do with the other pin? Run a wire from it back into the 12V lead of the heated seats circuit?

Sorry if these questions seem retarded but I'm having a difficult time picturing this in my head. Anyone want to illustrate it with a friendly MS paint file? I can read wiring diagrams but not text. :lol:

I think that by that time resistance increases enough to limit current you will be experiencing dccd "China syndrome". :) Personally I'm not interested in testing that theory.

One wire of the dccd plug would go to power, and the other wire to ground. I would recommend to put the switch on the ground/neutral side of the circuit, or even better use an additional relay to switch the high current to the dccd and use a low current circuit for switching the relay on and off.

I'm really against just hooking a switch up to the dccd for any type of short duration but permanent instasllation, let me put in a warning not to blame me if you leave the switch on by accident, or your wife or kids turn it on by accident or otherwise burn up your center diff as a result of this. I don't want anyone to harm their car because I gave them advice that even I'm not 100% comfortable with for someone to use on a regular basis. The PWM signal required by the dccd would never get to 100% duty cycle on a normal basis, so this is a little above and beyond what the normal input should ever be. More than a few minutes at 100% could be too much.

The new cars have a steering position sensor and the STi's have a yaw sensor, but all the older cars only have a fore/aft g-sensor for the abs. To add just a g-sensor for yaw can be from $20-200 depending on the quality of the sensor, double that for a good 3 axis sensor. I was going to put in a 3d g-sensor but the cheapest was ~$100. (Basically any additional costs directly increase the cost to produce and sell these.)

Alright got it about the way to wire it in. Also appreciate the disclaimer. :)

But worse case scenario if the DCCD coil burns up what damage occurs? And what does it take to fix it? Replace the solenoid and coil. But does the actual center diff crap itself? What are the symptoms of a burnt out DCCD coil? Always stuck in fully open or full lock? Does the gearbox need to be disassembled to fix the broken parts?

Alright got it about the way to wire it in. Also appreciate the disclaimer. :)

But worse case scenario if the DCCD coil burns up what damage occurs? And what does it take to fix it? Replace the solenoid and coil. But does the actual center diff crap itself? What are the symptoms of a burnt out DCCD coil? Always stuck in fully open or full lock? Does the gearbox need to be disassembled to fix the broken parts?

when a solenoid coil typically burns out, the electrical circuit goes open. that would return the solenoid to the rest/unenergized state. the coil would measure infinite resistance when tested with a DMM.

solenoids can also short circuit, in which case the electrical circuit stays closed, but there are insufficient turns to generate the necessary magnetic flux to keep the mechanical portion in the actuated/energized state. the coil would measure less resistance than it originally had when tested with a DMM.

that would mean, in either case, that the CD would fail "open."

hth
ken

I think that by that time resistance increases enough to limit current you will be experiencing dccd "China syndrome". :) Personally I'm not interested in testing that theory.

One wire of the dccd plug would go to power, and the other wire to ground. I would recommend to put the switch on the ground/neutral side of the circuit, or even better use an additional relay to switch the high current to the dccd and use a low current circuit for switching the relay on and off.

I'm really against just hooking a switch up to the dccd for any type of short duration but permanent instasllation, let me put in a warning not to blame me if you leave the switch on by accident, or your wife or kids turn it on by accident or otherwise burn up your center diff as a result of this. I don't want anyone to harm their car because I gave them advice that even I'm not 100% comfortable with for someone to use on a regular basis. The PWM signal required by the dccd would never get to 100% duty cycle on a normal basis, so this is a little above and beyond what the normal input should ever be. More than a few minutes at 100% could be too much.

The new cars have a steering position sensor and the STi's have a yaw sensor, but all the older cars only have a fore/aft g-sensor for the abs. To add just a g-sensor for yaw can be from $20-200 depending on the quality of the sensor, double that for a good 3 axis sensor. I was going to put in a 3d g-sensor but the cheapest was ~$100. (Basically any additional costs directly increase the cost to produce and sell these.)


the direct expense might be more but its much cheaper than the alternitive!

steering colum/abs pump/sensors/wireing.....

it would be nice to have a unit that could plug in and do everything in its own little box, it would be 100x easier to install and much easier to use, in the long run an extra 100.00 would be well worth it. i personally would like the most advanced DCCD controller avalable!

Spiider, any updates? :)

What are the latest specs of the 2nd version? I don't think using only the G-sensor is a good idea because starting off in snow will be painful. The car won't produce any real acceleration so the DCCD will stay open. So low traction situations in general will suffer.

Unless there was a way to progressively increase minimum lockup as traction decreases.

As the decelleration force increases, the CD lock up should also increase IMO.

As TPS increases, the CD lock up should increase IMO. I think a reference to TPS is necessary since it gives the driver some direct control over the behavior of the center diff. In snow/dirt acceleration out of a turn is best with a high lock up rate.

On tarmac, you don't need as much lock or you'll get understeer. So could there be a way to adjust the max lock percentage with a knob?

Is the stock boost control solenoid operated by a 12V PWM signal as well? With my limited knowledge, it seems like there could be a way to control the DCCD with my Wastegate Duty Cycle map from the stock ECU. At the very least it would lock up the center diff based on my throttle input. Educate me.

just an fyi, I finally hooked up the DCCD control Spiider sold me and it works well.

thanks amn

Do you have the manual version or the automatic version? I need to figure out a solution for my open diff soon because corner exits are driving my crazy.

The manual version. I bought it for the same reason. During autocross I felt like I was driving the "General Lee" lol

yeah i autoxed last time with no dccd controller on my 6mt and the car was definitely getting loose.
from what i hear spiider is out of the country for a while, bayley is having trouble sourcing parts for what he's working on, and neetronics has yet to respond to my email (although its been less than a week and its a busy time of year).
does anyone have a good dccd controller they are selling?
or better yet, is there a way to take the factory dccd controller and wire it up just to operate in manual mode? i have an oem dccd controller sitting in the garage.

WOW....just.....W O W

I'm back and I should have ~3 units ready to go by the end of next week.
Please check on ebay for my listing if you need one.
http://stores.ebay.com/CMM-Pros-Online-Store

I also have to get some more work into the automatic version, I was almost ready to sell them and found some improvements I need to make to the input signal processing.

The g-sensor is checked against the TPS in my system so if there is TPS input and no g-sensor input (stuck snow situation described earlier) the DCCD will ramp up the lock signal more aggressively to give you the traction you need.
(It see's that you are applying throttle and not getting anywhere.)

Is the auto version in-car adjustable? What settings can you change?

The auto version still isn't ready yet, sorry. I really wanted it to be ready 6 months ago but I've had little time for the development.

So far on the prototype I already have TPS, brakes, and the abs g-sensor interfaced.

The only "control" per se was the adjustable knob, but it had some "intelligence" programmed in for the various settings. ie. 3 different knob settings ran different calculations for lock percentage scaling and manual mode override.

This prototype is very intuitive, predictable and was incredibly fast at sensing input conditions and changing the action to just get you the traction you need when you needed it. I had to try a lot of formula variations vs. electrical measurements to get it to feel just right.

sounds good, keep us updated and thanks for your hard work.

even though i have no dccd, i gotta say: spiider, you rock! kick ass DIY project.

Thank you ride5000,

Also I have some controllers in my ebay store, new stock available if anyone needs one.

:confused: < Double checks to see who's thread this is > :rolleyes:

i have nothing but love for you too, fragment. :)

let's just say "go canada!"

-ken

fragment is a good guy, we have talked a bit in emails in the past about this project. Sorry for hijacking your thread.

any updates on this?

Installed Spiider's manual DCCD controller this weekend. Thanks Jeff.

- Short Review -
Well 'manufactured', easy to follow directions, comes with all necessary parts.

Huge improvement in traction while accelerating. Visited an unplowed (snow) stretch of road at a housing community under development. There was a nice sweeping uphill longish left hander perfect for testing. With the controller at it's min position, the car drove like a RWD car and constant countersteer was needed to maintain travel in the intended direction. With the controller at it's max position, the car experienced heavy understeer unless a 'drift' was initiated with a very mild scandanavian flick. I noticed I was able to apply considerably more throttle in the turn at this setting. I doubt I'll ever be using the max position on the potentiometer however.

I installed the controller so the min setting is at the 6 o'clock postion while the max setting is somewhere in the 4-5 o'clock range. The knob is turned clockwise to inrease lock. I seemed to like the way the car behaved at the 9 o'clock position but on days when I'm feeling a bit less adventerous, I may bump it up to the 12 o'clock position. Anything past that is pretty useless in my opinion since the car understeers too much for my liking.

This weekend's testing has me convinced that an open CD is 'worthless' in the snow. Now if only my rear 'LSD' diff was worth a damn.

Jeff, what is the duty cycle value when the knob is at it's max position? And does the duty cycle rise linearly wrt the angle on the potentiometer?

I had a chance to play with the Spiider DCCD controler in the rain but I got nervous driving around with it set . Will driving around with voltage going to the coils over heat them?

Hey MRF582 thanks for the nice review!

The manual DCCD controller runs up to max ~75% duty cycle. (That gives the coil plenty of "off" time to reduce the heating effects.)

I have driven with the knob set to 100% for several hours in really poor backroads winter commuting without any problems.

With it at 100% the car almost unmanageably understeers. In 99% of situations something around 50% will be more than enough. I would use the 100% setting if you are stuck in deep snow, or driving in really harsh conditions, or some customers have told me they are using it set 100% for drag racing as well.

In other news, I ordered a nice 2 axis g-sensor to test with in the auto version prototype. The cost is reasonable so I'm thinking I will change up the final design to include the new sensor. I will be able to change the reaction of the controller based not only on braking and acceleration, but also based on vehicle dynamics during turning.

I would like some input into what the desired behavior would be under different driving conditions.

1. Straight line acceleration is obvious, based on TPS and g-sense differential increase the lockup.

2. Under braking? Go open, or maintain some center diff lock?

3. Acceleration while under different g rates of turning?

4. Deceleration while under different g rates of turning?

5. Mixed situations like trail braking?

6. Any other conditions?

Could someone post a pic of Spiider's controller? I would just like to see what it looks like.

how close are you to the production of the "active" controllers?

will you be making any more from the first production model?

Prices? (there are none in your store)

i'd be happy to post a pic, but isnt this fragment's thread? does spiider have a thread somewhere where we can further discuss his product?

perhaps you can show your install in "interior/exterior modification...."

whats wrong with the factory DCCD controllers on th sti?

whats wrong with the factory DCCD controllers on th sti?

Nothing. Just try putting one in a WRX with a six speed.

^yea it's a freaking nightmare to say the least

how close are you to the production of the "active" controllers?

will you be making any more from the first production model?

Prices? (there are none in your store)


The active controllers are about 8 months behind my original release date, lol.
Sometimes other aspects of life gets in the way...

The active controller is in its 2nd or 3rd iteration, hopefully this will be the last redesign now that I added a 2 axis g-sensor to the final design.

I have basically stopped producing the manual version except by popular demand. I'll make you one if you need one in the meantime, but I'm not making new stock for the eBay store. (The production and sale of the manual units takes much needed development time away from finally getting the active controller done and ready for production.)

I'm still trying to keep the target price at $250 shipped for the active unit. It really depends on component prices, but looks very do-able at this time.

Got any pictures of the manual controller? How much would that cost if I wanted one?

You can send me an email, I have a picture I can email you.

reid_jeff@rogers.com

wow.. i never even use this in my sti. dont know why youd really want it this bad.. but hey.. good job!!

Lol, you are "using it" even if your STi controller is turned all the way down but you probably don't notice. (The stock STi controller is always putting out a small minimum voltage to the dccd.)

The reason these controllers exist is not for people who already have an STi, these are used predominantly in swapped cars that have no controller.

If you have swapped your tranny to an STi 6 speed and don't have a controller you very quickly find out that your car is now 100% rwd.

Subscribe
And I hope that home made DCCD controllers can be as functional as OEM
Keep all 6 speed swapper's happy
Waiting for good news
Thanks

Just bought the kit and I am going to follow the instructions best I can... I will let u know how it works on a 6sp SVX

Tom

well, my ver 3 sti RA tranny showed up...time to make my own controller :D

Well with a basic multimeter and a fear of destroying the coil... I got my controller hooked up and running... I am not sure I can read the frequency with any of my instruments so I can only read voltage which I have turned down... 1.3v or so at the lowest setting on my "dimmer switch" (yes I wired it to the dimmer switch on the blinker arm) and about 3.3v at max (not positive on the #'s right now) I also haven't had any "play" time with it, just checking for binding in slow turns at 3.3v and none at 1.3v. I did not make this controller for much more than using it on the dyno and for 1/4 mile runs. I will have some fun with it at AutoX to see what is best for my needs... No one has ever had a DCCD controller in an SVX so the car's characteristics are still unknown as of right now. If someone could tell me what i need to do to adjust the freq so I don't kill my coil it would be appreciated

Tom

Not exactly true about no other SVX swap's having working DCCD. Iain at Planetmotorsport has done this in the past, I know because he bought one of my manual controllers specifically for a swapped SVX this spring.

You can't use a (cheap) multi meter to properly check the PWM signal, you need an oscilloscope to view the waveform. I use a very expensive scope to adjust and calibrate my controllers. The manual and automatic controllers I make output the correct hz and waveform for the dccd. The dccd coil isn't only about voltage, it is about frequency and PWM duty cycle.

If you bought a "kit", you will more than likely need to beef up several of the components to hold up.

I got the velleman kjt that was mentioned in the thread... I am only speculating on the SVX thing... many owners and modifiers are on my other site and to be honest, I have done 4 out of the 7 6mt swaps that I know of, so it veyr well could have been one of my customers... Where were they located?? I need to see if I can borrow a scope from one of my friends cuz the reason I didn't buy a controller was for the simple fact that I couldn't afford it right now. I have a HUGE engine build in the process and has drained the wallet pretty well. Either way, my questions applied to the kit mentioned in this thread because this thread is meant for the velleman kit.

Tom

Iain at Planetmotorsport is located here in Guelph, Ontario Canada.

He has done several of these SVX swaps too. He is undoubtedly (IMO) the best swap shop in Canada.

FYI, the velleman kit is a good start for a manual only controller but won't hold up long term without some mods to the design and some upgraded components.

Spiider, you still selling the dccd controllers? I'd be happy with just a manual one.

You can't use a (cheap) multi meter to properly check the PWM signal, you need an oscilloscope to view the waveform. I use a very expensive scope to adjust and calibrate my controllers. The manual and automatic controllers I make output the correct hz and waveform for the dccd. The dccd coil isn't only about voltage, it is about frequency and PWM duty cycle.

Actually, some of the newer hand-held multimeters have the ability to read PWM square waves. My Fluke meter as this feature and I use it all the time. It's a lot more cost effective and practical than carrying a large O-scope with you into your car. :)

is there a way to wire in a disconnect signal? basically to make the diff open or rwd?
Thinking about heating up tires ;)

I have a fluke 87 that can measure the duty cycle as well as my tec scope.

It can only measure voltage and duty cycle though, you can't see the waveform properly (obviously) without the proper display.

Most people, unless they are electronics guys like us don't have $500 meters. :)

Spiider, you still selling the dccd controllers? I'd be happy with just a manual one.

I have just one remaining manual controller right now for sale on eBay, it was from a customer that never installed it and wanted to upgrade to the auto version.

The auto version is available as regular production I have a batch pre-selling right now.

Jeff,
Just a couple of questions on your Auto controller:
- Can I still adjust the diff manually, or is it auto all the time?
- Is there a link somewhere to the installation instructions? I'd like to read them in advance.
Thanks,
Rob

Jeff,
Just a couple of questions on your Auto controller:
- Can I still adjust the diff manually, or is it auto all the time?
- Is there a link somewhere to the installation instructions? I'd like to read them in advance.
Thanks,
Rob

The current version of the auto controller adjusts itself automatically all the time. There are no user controls. I am also developing another option that shloudn't take to too long that will allow people to adjust the magnitude for acceleration, braking and TPS. It will require a small control panel to be installed.
It will retain the auto functions with user adjust-ability.

You can check out the install instructions etc at my new website, or on eBay.

i need a center diff for a wrx transmission asap

Spiider. dibs on one from the next batch

Spiider,

Can you please post your website address and ebay page or seller name.

Thanks,
Mitch

when will the next batch be up and what systems are required? I'm imagining wheel speed sensors?
so ABS is necessary and tied into?

have a ver 3 dccd with no controller ;)

Sorry, I am unable to post direct links to my store or eBay due to Nasioc vendor regulations. (I can't quite afford official vendor status yet, but it is a goal.)

This batch is all spoken for and will be shipped out Tuesday (Monday is Canada day here), the next batch will be built immediately afterwards if there are enough people needing them so if you figure another 2-3 weeks for parts ordering, build and test time...

That will probably be the last batch before I go on vacation at the end of July so if anyone needs one the next batch of controllers will be the last before end of Aug.

VERY interested in auto controller! I'd be more interested in a product that offered some type of adjustment or remapping though, as the settings I would like for autox are not going to be the same as the settings I would like for prorally.

^^^True, true..^^^

I've already added the aforementioned knob/control for user adjustment.
It allows the user to fine tune the DCCD response level across the board. The new input is used in all 4 formula's as a magnifier of center diff "persistence" or "aggressiveness". Think of it as a "volume" knob for your center diff, still working in full auto mode but "quieter" or "louder" adjustment levels so you can tune it to your preference.

Don't forget that I personally build these in small batches by hand. I write all the firmware so we can do literally anything with the programming. The programs I'm using now I developed for tarmac, but if you need different response curves I'm more than happy to work with you to develop something personalized to your needs.

Ps. Everyone who ordered from the last batch got the latest and greatest available. (With the adjustment knob.) Next batch will be up soon...

I would give you some good instructions about mapping for mine then...too bad I've never driven a car with dccd before :D

I'm going to be driving this car tarmac style for the next year or two and then it's going to switch over to gravel/snow...any chance your units can be "reflashed" or have their IC's swapped to gain new programming in the future?

I would give you some good instructions about mapping for mine then...too bad I've never driven a car with dccd before :D

I'm going to be driving this car tarmac style for the next year or two and then it's going to switch over to gravel/snow...any chance your units can be "reflashed" or have their IC's swapped to gain new programming in the future?

Absolutely, it is as simple as changing a rom chip. There is actually enough ram/rom to allow for me to switch between at least 2-3 multiple programs on the fly with a simple selector switch. Maybe if I get enough serious "input" from competitors in different disciplines I could add this in the future, or offer multiple chips for people to swap out.

A map selector would be REALLY cool.

I can add the map selector without any effort, it is the data that I need...I need to know how people want their cars to behave in different types of racing.

Right now the tarmac program is tweaked quite nicely, I guess the next logical program to add would be for snow or gravel (?). The tarmac program already has a built in mode for "stuck" condition, I guess the only difference might be in cornering or braking traction.

P.s. The next batch will be finished ~Monday. (There were only 4 in this batch and 2 are already sold locally.)