Cheap Electronic FCD - Revised

That spinning thing that makes all of the cool noises. OE and Aftermarket.

Moderators: Helpinators, Moderators

vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

If you're selling them to ricers, I think you have to make the adjustment accessible to the end user.

I guess it makes the most sense not to make it super-rugged but to just use a regular ABS plastic case with a hole in it that you can use to adjust the trimpot. Then just say it should be installed inside the passenger compartment, near the ECU.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
Do It Sidewayz
In Neutral
Posts: 21
Joined: Fri Mar 07, 2003 6:10 pm

Post by Do It Sidewayz »

i'll let you know how it works.

We are building one for our Legacy Turbo Race Car.

We are tuning the thing to run on 100 octane fuel. and lots of boost
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

How are you tuning it for race gas? If you're using a programmable ECU then I wouldn't expect you to need a fuel cut defender.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
Do It Sidewayz
In Neutral
Posts: 21
Joined: Fri Mar 07, 2003 6:10 pm

Post by Do It Sidewayz »

nope no fancy crap.

We are playing around with a potentiometer in the Water Temperature Sensor line to richen the fuel mixture.

Along with a cheap Manual boost controller, Cone Filter, and an adjustable fuel pressure regulator.

Plus we normally just run race gas (98 or 100 octane) as "cheap" insurance
Do It Sidewayz
In Neutral
Posts: 21
Joined: Fri Mar 07, 2003 6:10 pm

Post by Do It Sidewayz »

actually. i just talked to him.

He built the circuit last night.

Parts cost him around 3 bucks CDN.

Car now pulls 15 PSI no problem and pulls like a bear
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

Awesome! Victory jig! Image
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
ciper
Knowledgeable
Knowledgeable
Posts: 4387
Joined: Tue Oct 15, 2002 8:16 pm
Location: SFCA

Post by ciper »

Someone double check this for me. Seems that the lowest output voltage would be about 2 volts when you are at 10k feet and the vehicle is not running but the key is in the on position(Doubt you will be any higher).

vrg3 Shouldnt you modify the forumula you posted so that the maximum output voltage will never exceed the input?
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

ciper wrote:Someone double check this for me. Seems that the lowest output voltage would be about 2 volts when you are at 10k feet and the vehicle is not running but the key is in the on position(Doubt you will be any higher).
Yes, the pressure sensor outputs approximately 2 volts when input pressure is approximately 20 inHg, which is what you'd see at 10,000 feet elevation.

Why do you bring it up?
vrg3 Shouldnt you modify the forumula you posted so that the maximum output voltage will never exceed the input?
I don't understand what you mean. You're talking about the equation on the web page I posted? The relationship only holds as long as pressure remains within the sensor's valid input range...
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
ciper
Knowledgeable
Knowledgeable
Posts: 4387
Joined: Tue Oct 15, 2002 8:16 pm
Location: SFCA

Post by ciper »

"The relationship only holds as long as pressure remains within the sensor's valid input range"

Yes but valid as long as pressure doesnt exceed 14 psi. Many plan to run higher pressure than 14 when using an FCD and if the equation was used the output would exceed the input. It doesnt really matter though.

"Why do you bring it up?"
Two friends of mine are knowledged in electronics so I first gave them thefull requirements then showed them the diagram. I didnt want any assumptions made on the operating condition of the curcuit so that changes made wont effect operation. Basically I said I wanted 330 degrees of rotation resulting in a variance of 2.7-4.7 volts output when the power supply is 5 volts and the input voltage is between 2 volts and 5 volts.

Im also trying to keep cost low where possible so cost savings can be past to the end user. Most likely a flat piece of metal bent at 90 degree with a hole drilled in the top will be included with the curcuit inside the epoxy case to allow easy mounting.
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

ciper wrote:Yes but valid as long as pressure doesnt exceed 14 psi.
Right... cuz the sensor can't read higher than that. That page does say "The Subaru pressure sensor can read up to approximately 29 psia."
Many plan to run higher pressure than 14 when using an FCD and if the equation was used the output would exceed the input. It doesnt really matter though.
The output would exceed the input? I guess I don't understand what that means.

If you feed the sensor a pressure higher than 29 psia, it will read 29 psia.

On the first attempt to repair boostjunkie's failed pressure sensor, we put in a 1.5-bar sensor from a naturally aspirated Legacy. If it had worked (it didn't presumably due to his broken ECU or wiring), he wouldn't have needed an FCD since the ECU would never see more than half a bar of boost.
I didnt want any assumptions made on the operating condition of the curcuit so that changes made wont effect operation. Basically I said I wanted 330 degrees of rotation resulting in a variance of 2.7-4.7 volts output when the power supply is 5 volts and the input voltage is between 2 volts and 5 volts.
I see what you're trying to do with it... you want to pick a minimum sensible level so you get good resolution. You'd just need to put resistors in series with the pot, one on each side, and you can compute their values. But it seems silly to add a whole other component to the circuit just to get it to limit at 4.7v; you may as well have the threshold adjustable from 2.7 to 5.0 volts.

But why?

Every manufacturer's pressure sensors have different transfer functions. The equation on that page applies only to Subaru part number 22012AA020, used (as far as I know) only on Subaru turbos built from 1989 to 1996.

For all US-market Legacy Turbos, the circuit as described on that page (non-adjustable) will work. If your only other market is overseas Legacy and Impreza turbos, this circuit would still probably work okay, although it'd be better to increase the 47K resistor a bit.

Your target market (I thought) was all factory turbocharged cars. Since they'll all have different pressure sensor transfer functions, you can't use that equation at all. Just make it adjustable from 2.5 to 5 volts and call it a day; pretty much all factory turbo cars' pressure sensors will have their cut somewhere in that range. I've added such a circuit to the web page.
Im also trying to keep cost low where possible so cost savings can be past to the end user. Most likely a flat piece of metal bent at 90 degree with a hole drilled in the top will be included with the curcuit inside the epoxy case to allow easy mounting.
That's pretty clever. Me likes. :)

Or you could even just include any random TO-220-type component on the board, just to use its mounting tab, if that happens to be cheap. It wouldn't have to connect to anything.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
ciper
Knowledgeable
Knowledgeable
Posts: 4387
Joined: Tue Oct 15, 2002 8:16 pm
Location: SFCA

Post by ciper »

"2.5 to 5 volts and call it a day"
You got it, that was the goal if reducing the resolution farther increased cost. I proposed the idea to them and asked for options, cost being the deciding factor on one.

"The output would exceed the input?"
If I put 32 psi in the equation (14 sea level plus 18 boost) the equation would give a higher number. Doesnt matter.

"1.5-bar sensor from a naturally aspirated Legacy"
Seems like this would work except wouldnt the Turbo ECU think its getting higher elevation than it actually was (9 psi greater range on the turbo sensor right?)?
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

ciper wrote:"2.5 to 5 volts and call it a day"
You got it, that was the goal if reducing the resolution farther increased cost. I proposed the idea to them and asked for options, cost being the deciding factor on one.
Right, but what I'm saying is that irrespective of cost, you shouldn't reduce the resolution too much because some cars' sensors might need the clamping voltage to be outside the range you choose.
"1.5-bar sensor from a naturally aspirated Legacy"
Seems like this would work except wouldnt the Turbo ECU think its getting higher elevation than it actually was (9 psi greater range on the turbo sensor right?)?
Well, we used a conditioning circuit to transform the signals, so the ECU would see the correct numbers up through 6psi. If it had worked, that is.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
ciper
Knowledgeable
Knowledgeable
Posts: 4387
Joined: Tue Oct 15, 2002 8:16 pm
Location: SFCA

Post by ciper »

I dont know what you mean
"you shouldn't reduce the resolution too much because some cars' sensors might need the clamping voltage to be outside the range you choose. "
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

ciper wrote:I dont know what you mean
"you shouldn't reduce the resolution too much because some cars' sensors might need the clamping voltage to be outside the range you choose. "
I'm just saying that there may be some cars that'll need to clamp at 4.71 volts. Or maybe (though much less likely) at 2.69 volts... While 2.7 is probably a safe lower limit, your upper limit really ought to be the full 5 volts.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
Do It Sidewayz
In Neutral
Posts: 21
Joined: Fri Mar 07, 2003 6:10 pm

Post by Do It Sidewayz »

yeah....it's a little quick

I actually drove the car today.

runs 15PSI spike and drops down to about 12-13 which is perfect.

I was able to spin all 4 in 1st, and a little in 2nd.

Alot faster than Inzomniac's Legacy with minor mods.
boostjunkie
Knowledgeable
Knowledgeable
Posts: 1770
Joined: Mon Dec 16, 2002 3:20 pm
Location: Owings Mills, MD
Contact:

Post by boostjunkie »

Do It Sidewayz wrote:runs 15PSI spike and drops down to about 12-13 which is perfect.
You should see what 16psi throughout the powerband feels like!

What mods are done to your race car? :twisted:
[url=http://www.angelfire.com/md3/91turbolegacy/images/On_the_Lawn.jpg]1991 Legacy Turbo (RIP)[/url]

[url=http://www.angelfire.com/md3/91turbolegacy/images/Summer_Car_Wash3.jpg]2000 Celica GT-S[/url]
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

Just to see how hard it was, I decided to build one free-standing:

http://www.graphics.cornell.edu/~v/fcd/ ... anding.jpg

It was pretty easy.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
ciper
Knowledgeable
Knowledgeable
Posts: 4387
Joined: Tue Oct 15, 2002 8:16 pm
Location: SFCA

Post by ciper »

Now make a box out of masking tape and fill with epoxy!
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

Way ahead of you. :)
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
Do It Sidewayz
In Neutral
Posts: 21
Joined: Fri Mar 07, 2003 6:10 pm

Post by Do It Sidewayz »

lets see.

We are running full 2.25 inch cat-less exhaust (it's a little loud)

We have done away with the stock air-box, and replaced it with about a 10 inch long K&N Cone filter. With a custom Air Flow straigtener placed infront of the Mass Air Flow Sensor.

We also removed all the PVC System, and other emissions control devices.

We are installing a Potentiometer in the water temperature line so that we can trick the computer into running richer for the higher boost.

That is about it.

Lettme tell you though...it's quick...I thought Inzomniacs car was quick. but it's nothing compared with this thing
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

I sealed it with epoxy and now I'm selling it in the Parts Shed since I don't need it... http://bbs.legacycentral.org/viewtopic.php?t=14489
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
ciper
Knowledgeable
Knowledgeable
Posts: 4387
Joined: Tue Oct 15, 2002 8:16 pm
Location: SFCA

Post by ciper »

Does the ECU have a single 5 volt rail that all the devices share? If so does anyone know the most power that can be drawn from it?
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

Yes, the ECU does have a single 5-volt rail it uses as a power supply for logic components and as a reference for the sensors. However, none of the external pins connect directly to this rail.

The TPS power supply pin is connected to this rail through two 51-ohm resistors, making it a fairly high impedance source. I'm not sure why, but I think it has to do with fault detection. If you draw any useful amount of current from this pin you'll usually trip an idle swith CEL.

The pressure sensor power supply goes through what appears to be some kind of buffer fed by the 5-volt rail. That would imply that it's a low-impedance source but I haven't verified it.

The safest thing to do if you wanted to draw a significant amount of power from it would be to buffer it yourself.

Why? Are you thinking of eliminating the requirement for the +12v power to the FCD by using a rail-to-rail op amp? That would probably be okay. I suggest the TLV2374. It would drop right into this circuit except the power supply could connect to +5v instead of to +12v.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
ciper
Knowledgeable
Knowledgeable
Posts: 4387
Joined: Tue Oct 15, 2002 8:16 pm
Location: SFCA

Post by ciper »

Why not get rid of the +5 volt for the curcuit and just run it from only 12 volts?

Im looking at options to make it as simple install as possible. The fewer ECU leads needing cut the better. Reducing it from 5 wires to 4 would also help.

As an update, Im going to include a switch that allows pass through operation. In one position the FCD is active, in the second it passes full voltage and you dont have to retune it. I also will probably stick in in a plain black box without using epoxy.
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

ciper wrote:Why not get rid of the +5 volt for the curcuit and just run it from only 12 volts?
Like I said above, the ECU uses the 5-volt rail as a reference. The actual voltage on the rail is actually a little above or below 5 volts, and all sensor readings are in terms of this voltage. So you really should define your clamping threshold in terms of this reference voltage as well.

You could get away with making your own regulated reference, and just keep the clamping voltage low enough that even with a relatively "low" ECU reference it would still work, but... that's a lot of complexity. But you could.
Im looking at options to make it as simple install as possible. The fewer ECU leads needing cut the better. Reducing it from 5 wires to 4 would also help.
Then I highly recommend using a TLV2374 instead of an LM224. That way you only splice it into +5v and ground, and put it inline with the signal wire. This could be done easily at either the ECU or the sensor.
As an update, Im going to include a switch that allows pass through operation. In one position the FCD is active, in the second it passes full voltage and you dont have to retune it. I also will probably stick in in a plain black box without using epoxy.
Sounds cool... The one I made for boostjunkie is in a plain black plastic box.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
Post Reply