john banks

Why not? Rude not to!

[Edited by john banks - 4/1/2002 10:58:06 PM]

C h a z

Paul H
When do I get to have a look at this car of yours?

[Edited by C h a z - 4/1/2002 11:10:39 PM]

Chris.Palmer

Scott,

Hate to state the obvious but why not use a MC78L10ACP 10V regulator?

Chris.

Edited to add - also consider single OP296 instead of 2 LMC7111s

[Edited by Chris.Palmer - 5/1/2002 2:02:18 PM]

Scott.T

iTrader: (2)

Couldn't find a 10V version of the regulator in Farnell, but after my tests today it would be possible to use the 9V version, without the POT.

12V-14V input gives constant 9V and allows the device to be set to run with the step between 4.1V and 4.8V as required (for PPP MY94, for non PPP it will be worth bring the voltage down to 3.9V)

David.... you have mail regarding the thingies (a bit of a ***** up this morning, I'm afraid)

[Edited by Scott.T - 5/1/2002 2:59:44 PM]

Scott.T

iTrader: (2)

John,

Doesn't the ECU adjust the Duty Cycle to bring the boost back down to standard settings, after the actuator has been adjusted.

I'm sure mine did, a few months back, when I wound it in 3-turns.

P.S I know that the Volvo Actuator which is of very similar design i.e bent rod, has a tendancy to snap at levels above 15psi.
Volvo T4 Modifiers fit an uprated rod/actuator.

[Edited by Scott.T - 4/2/2002 8:28:25 AM]

James_PowerMad

If you guys are gonna go this way, and use the very well engineered Subaru boost controller, then perhaps the following MAP re-map may be better... (and simpler)


R1
IN ------/\/\/-------- OUT
|
|
>
< R2
>
|
|
_z_
/\ (poor attempt at zener diode)!
--
|
|
---
- GND

<< *** Sorry, the text came out wrong... The resistor R2 should come down from the OUT side of R1!!!!!!!!!!! *** >>


If you get the zener to cut in at around 3 volts (or whatever is about 10 psi - got no info with me here). And set R1 and R2 to sensible values, then you can get a nice unadulterated signal through until 10 psi, when the slope will decrease in gradient (you effectively compress the pressure into a smaller voltage range).

I'm sure John knows what I am getting at, and I would expect such a circuit to perform better than a discontinuous one (because the ECU can see any changes coming, rather than suddenly banging through to fuel cut teritory).

The resistances will need to be chosen with thought to the internal resistance of the MAP sensor, and ECU (so that you can accurately predict what will happen). Alternatively you can stick a couple of op-amps in to buffer the signal (at IN and OUT).

And another... If a zener diode cannot be found with an appropriate voltage, then either stack up with some normal diodes (0.7v), or use an op-amp / potential divider as a voltage source.

[Edited by James_PowerMad - 4/2/2002 1:27:54 PM]

john banks

Good report Paul, thanks. I take it you are sticking to this method now then?

I used to use 1.35 mm restrictor with the standard actuator with PPP and downpipe and it held 16.5 PSI. Peak and held were the same. Any higher with the peak and it fluctuated and was nasty.

Now I run 1.30 mm restrictor and the actuator tightened one turn. I think on my car this is required otherwise the boost with any methods drops off from 4000 RPM quite quickly. This has not happened to other cars I have been in/driven.

I think just fitting the fuel cut defender without changing actuator or restrictor will only increase the boost if you were running a peak before hand. As I was not doing with the PPP as it overshot then undershot and then stabilised and felt weird, then I don't think it would give me any more boost. However, note my standard restrictor is 1.2mm but even with the actuator as standard with the PPP this results in wild overboost, but still disappointing over 4500RPM.

Any more than one turn of the actuator it does not hold above 4500 RPM any better, but this one turn makes a big difference.

I don't honestly think the ECU actually changes the duty cycle at all unless it detects that you are OVER its intended boost. Hence the reason for defending from just before it. The gains we are getting are from altering the mechanical systems that open the wastegate to make them open at higher manifold pressure. The FCD just stops the ECU interfering. This is how I believe the Subaru boost controller works - it is actually OPEN LOOP unless you overboost its target. It handles this reasonably well within limits and you get a peak, go too far and it oscillates. You delay all this to the new fuel cut point with the defender.

I thought it spooled up slower than the Dawes because it wasn't as ragged and "clonky" but very smooth. Looking at the tachometer and the boost gauge I can still get 18 PSI at 2400 RPM. I am still holding about 14 PSI at 6000 RPM in 3rd or 4th gear. Difficult to honestly say if the spool up is any slower then, I reckon not - it is just so much smoother, and deceptively fast. The boost is hitting the same points as far as I can see. On some favourite stretches of (private) road I could still get to 6000 RPM in 4th gear just before braking for a corner, so the overall performance is as good as the EBC or Dawes.

With the EBC or the Dawes if you had substantial proportional gain or a small bleed hole (effectively the same thing) it just spiked (and with the Dawes was horrendous on part throttle), and I don't think we are losing that much having a duty cycle mapped by RPM by the Subaru boost controller. This does help the low down spool up. It perhaps doesn't work the turbo as hard below 2000RPM, but honestly with the Dawes you could see boost but it didn't shift the car. This way does seem more mechanically sympathetic.

I had a missing sensation on boost with the boost controller and the Dawes which was helped by regapping the plugs but not away entirely. It is now with this method.

Also with the EBC the boost behaved oddly now and again and I wondered if it was the ECU reading atmospheric pressure which upset my controller. It would have been difficult to work around this without a separate pressure sensor or an interface to the switching solenoid. As well as the EBC being a pain to program and to set up. I learned a lot by doing it though.

Some have reported identical acceleration figures using a restrictor/bleed valve rather than a Dawes. As I wondered in another post I think there is actually very little wastegate creep until quite near target, it is just that Subaru mapped in a boost profile that is several PSI below what is possible. Changing the restrictor/actuator seems to solve this very nicely and you are left with the limits of the turbo and wastegate spring common to all boost control devices if they are pushed. It's just that we have now pushed the standard one as far as we were pushing the Dawes.

[Edited by john banks - 4/2/2002 9:56:51 PM]

john banks

"Just a question : what happens to the output when the input voltage reaches the zener voltage? Haven't you got a strong decrease of the output voltage as current goes through the R2 resistor? I believe there is a big negative step when the zener diode reaches its threshold and then, the curve continues from this new origin."

James, Xavier my French EE friend sent me this when I pointed him to your circuit. I thought you might understand? I think your circuit theoretically looks great. Should be a doddle to build.


[Edited by john banks - 4/3/2002 2:55:03 PM]

john banks

A fuel cut lifter using the standard ECU along with a restrictor and/or actuator adjustment makes a very good performance package.

FINDINGS:
PROS:

• Part throttle response much more progressive than Dawes
• On boost smoothness is superior to a Dawes or the EBC design I was working on
• Spools up and holds boost as well as any other method
• Retains ECU functionality for reducing the boost if knock is detected which cannot be corrected by retarding the ignition timing
• Limp home functionality preserved
• Fuel cut still present
• ECU can reduce duty cycle to try and reduce boost before resorting to fuel cut

CONS:

• 2nd gear half boost after holding constant throttle for a few seconds still present for emissions and noise testing

Of note you have to defend the MAP sensor signal a few PSI before fuel cut as otherwise the ECU would reduce the duty cycle to control boost.

I have tried this on MY00 AE802 ECU and PPP ECU with success. Notably no bleed valve is required.

The circuit is as follows (sorry can't find old DIY FCD Results thread).




This is what it does, but do not be fooled by the scale. I am also running the notch from 3.8 to 4.3V now.

The op amps are LMC7111 8 pin DIL chips (two required). Two 4k7 or similar potentiometers (exact value not crucial). 1N4001 diode. A little black box and some board to solder it on. About 7 small wire links are required. Total cost about £5 including the box and board.

You can happily power it off the throttle position sensor power supply (regulated 5.10V) and ground - it only draws about 2-3mA. Then you set one potentiometer from the point where you want the circuit to defend from, and the other potentiometer to set where you want the new fuel cut to be. I have set 3.80 and 4.30V respectively to defend from 15-20 PSI. I run 18-18.5 PSI with a PPP ECU, but have also tested a MY00 AE802 ECU running the same and it fuels nice and rich on both of them and drives very well. I think the PPP ECU does have the edge but we know the AE802 is a very poor ECU. Perhaps an AE800 or AE801 ECU would be much more like a PPP ECU - perhaps indistinguishable at the same boost?

As usual there is no warranty and you need to be confident about your engine and your electronics and restrictor fiddling/actuator adjusting abilities to take on this project, but overall it is only a little more difficult than the DIY AFR project, just a bit more serious if you get it wrong.


http://www.national.com/ds/LM/LMC7111.pdf

You can get the parts from Farnell. The op amp has to be "beyond the rails" input hence the rather odd selection of the 7111.

[Edited by john banks - 4/1/2002 10:54:18 PM]

Paul_H

The DIY FCD thread is here:
http://www.scoobynet.co.uk/bbs/threa...threadid=78214

john banks

Thanks Paul. Damn search! Do you have any results yet?

T-uk

"Perhaps an AE800 or AE801 ECU would be much more like a PPP ECU - perhaps indistinguishable at the same boost?"

want one to test

Paul_H

Actuator tightened by 2 turns;
Original restrictor replaced - Dawes removed;
Fuel cut lifter defend zone set from 3.6V to 4.3V;

Can I mention the EIGHT bolts that hold the heatshield on!! over engineered or what!! and of course as always there is one that is much better removed from below. There are now "only" seven bolts on my heatshield. I left the actuator rod connected (locknut slackened) and unbolted the wastegate body from the turbo, and turned that instead, it's tight but can be done.

Will run it tomorrow & let you know.

john banks

Paul the effect of the actuator tightening seems to vary car to car. If it doesn't lift the boost much you might need a smaller restrictor. I await your results with interest.

Paul_H

OK...and I'll look at defending from 3.4V too.

john banks

No it doesn't as the boost level has not breached its boost target if you defend correctly. Whilst my actuator is 5 turns at present it does work well at about 2 - just need to get the restrictor size about right. At 2 I think mine is effectively only 1 turn tighter than standard as my car did not hold boost over 4000 RPM even with a disconnected wastegate unlike most other cars.

Is it the tightness of the actuator that causes it to snap or the boost? I say this because PPP runs well over 15 PSI and the actuator is not uprated. At 2 turns I can't see it being too much of a problem - I kept turning as I thought the boost had more to climb at the top end, but it was already on optimum - about 1 bar at 6000 RPM.

The ECU seems to be open loop boost control unless the boost target is breached then it drops the duty cycle in a closed loop fashion but usually it overcorrects, and it does not seem to learn each time. If you defend from just before the boost target it holds fine - just like a Superchipped car does.

Scott.T

iTrader: (2)

OK, I understand now.
Might be worth giving it a try then.

With respect to the Volvo Actuator, I'm not sure if it's boost or tightness that is the issue. I'll try and find the links......

john banks

Thanks. You don't need to touch the actuator if your car holds boost really well with just a restrictor. Just that mine didn't. Having said that if you can get the heatshield off (mine only has three bolts as it was trimmed for the DP) then it is actually much easier to adjust than getting just the right sized restrictor. However, it is not as bad as it used to be where a smaller restrictor just makes it oscillate and with a 20 PSI fuel cut I have some leeway for getting in the range 17.5-18.5 held with enough that I won't overboost overtaking a tractor.

HarryBoy

John, you continue to amaze me!

There I was all ready to build your EBC.........

Is there a direct link between the voltage and PSI? Is it linear? IE is it possible to set this up on the bench then finally tweak it on the car?

Harry

john banks

There is a linear relationship. Roughly voltage = 2.3 + boost in PSI/10. This is accurate enough for setting this - you are not setting boost with the fuel cut lifter, just allowing your restrictor to freely influence the boost without interruption up until the fuel cut you choose. A 20 PSI fuel cut seems a sensible figure - about 2 PSI higher than anyone sane would choose to run. The only mechanisms by which you will see this are like on the unmodified car - hose comes off, wastegate stuck shut, leak, restrictor or solenoid gunked up etc etc. There is of course no bleed valve to go dodgy, no Dawes to stick etc etc.

john banks

I clicked edit on your post so I could see what you originally typed. Tried it on the breadboard with a 220K pot with 6 V across it as a simulated MAP sensor. I had handy a few 1K resistors and a 3.6V zener (lowest I had == 13 PSI).

The results were odd because you can't plot one voltage against another as the clamp pulls them all down, but the signal was uncorrupted up to a point and then it fell away compared with what I would expect from the potentiometer position.

Very difficult to set up though to get the behaviour you want? I think the values of the resistors would need to be played with a lot, but with the circuit I have used you can set each point independently although it is a bit more involved to build.

The ECU doesn't seem to mind having a clamp on its MAP signal, and when if it goes over your overboost point it copes very well by reducing the duty cycle.

If someone can find the right resistor values for James' circuit it would be very easy to build, but I think finding them may be a tricky process?

I am now running 18-18.5 PSI with a 1.35mm restrictor and the actuator tightened only one turn from standard and it all works rather well. Will run a 1.4mm restrictor for the track which drops it to about 17 PSI.

Scott.T

iTrader: (2)

John,

Just out of interest, how much higher than the standard MY99/00 PPP is that ?

James_PowerMad

When I have time later, I will work out the design properly... It's just a case of doing a few sums ;-)

The only tricky issue is what point in the range to start compressing the psi.

i.e, you could start compressing quite low - say 8psi, and then don't reduce the slope by much.

Or, you could have the compression start much nearer to the Subaru standard target - say 11 psi, and have a much steeper slope (this becomes more like your system John).

But basically, you want to calculate it so that when the ECU 'sees' its target boost, then it is actually your (slightly higher) target boost.

john banks

Standard PPP holds 16-16.5 on MY99/00 depending on gear so I am running 2 more than this on the road and 1 more on the track. The extra boost from using this method seems to be about 2 PSI across the range from 2000-6500.

Scott.T

iTrader: (2)

Whats the MAP sensor rated at on an MY99 as I belive that the early Pre'96 is only rated at 1 BAR, hence why the PPP on early cars is just under 1 BAR.

john banks

3 bar absolute

Scott.T

iTrader: (2)

Perhaps I should of said 2 BAR absolute then ?????????? as 1 BAR is atmospheric pressure (I think)

john banks

Yes. 1 bar = 1 atmos average.

Paul_H

My car is a MY99 UK Turbo, with AE800 ECU. As standard, boost peaked at 1.1 and held 1.0 Bar. With fuel cut lifter & Dawes Device I have been running it at 1.2 - 1.25 Bar. This of course gave much better performance, but with a few quirks.

With the Dawes fitted, when coming on boost there was a real turbo kick as the boost arrived. I must admit I do quite like this even if it isn't very good for the car However the more boost you run the more violent it is, with a 1.0mm bleed hole in the Dawes, there was quite a shudder as the boost comes in, quite low in the rev range. This probably isn't good for the turbo, gearbox, drivetrain in the long run. A good laugh for a little while though

I've just done a run with the car set up as follows:

Actuator tightened 2 turns
Original restrictor in place
Boost solenoid connected as standard
No Dawes anywhere to be seen
Fuel cut lifter defending 3.6V to 4.3V

The initial impression was that it wasn't quite as fast as with the Dawes - but this isn't the story.

The boost held 1.2 Bar easily - in 5th on the motorway. There was a slight peak, only momentary, above this, then a solid 1.2 Bar held. Remember this is WITHOUT the Dawes and WITH the standard boost solenoid.

The power delivery is moved a little up the rev range, and the Dawes "shudder" from the onset of boost is gone. In its place is a much smoother and more subtle power delivery. It is still has the characteristics of a turbo engine of course, just that the nature of the delivery is more like the standard car - but more of it. Quite a bit more in fact

I think the main reason for the extra power is the fact that the ECU, which is being fed a defended signal, isn't using the standard duty cycle for the solenoid - or rather it is, but the boost is higher than the ECU thinks. If you know what I mean

Consequently the solenoid is allowing more pressure to be bled away from the wastegate actuator than the standard Subaru map intended.

John, if you have a willing victim I think it would be interesting to leave the actuator alone and just defend the MAP signal, and see if this generates extra power. I expect it would if the defend point is low enough, as we would be effectively adjusting the solenoid duty cycle. Then adjust the actuator and measure again, then we could quantify how much of the increase is down to the actuator.

I think this reveals the true main function of the MAP sensor - we all know about fuel cut protection, but I think the main purpose is control of the boost solenoid.

I reset the defend point to 3.4V and did another run; it made no difference at all on my car. Same boost levels.

I've not got my AFR connected at the moment, but fuelling was fine with the Dawes and I expect the same again.

So the conclusion?

- The car is probably just as fast. It certainly develops plenty of boost - 1.2 Bar held on my car with 2 turns on the actuator.
- The sudden Dawes "shudder" when coming on boost is gone. I think this is better for reliability.
- The power delivery characteristics are changed to a more subtle and sophisticated delivery. Power delivery characteristics are more like the standard car.
- Except for more power
- I found myself using more revs. Power comes in slightly later than with the Dawes. Maybe I was just trying harder?
- It is effectively a shift away from a mechanical boost controller, to an electronic one. The one Subaru gave us, just tweaked a bit and a lot more discreet too.
- It is arguably nicer to drive
- Recommended!

ustolemyname??stevieturbo

iTrader: (1)

SILLY QUESTION... and are you getting safe amounts of fuel to go with the extra boost, and what are your charge air temps?? Its easy to increase the boost pressure, any dimwit can do that and fool the boost cut too, no matter what way you do it. Little chips dont make it any safer than valves etc.