Ed's 1991 CQ: Progress!

[Tom]

anything new,Ed?

[Wheeljack]

If you are just trying to get the tach to work, you can go directly to the yellow T26 connector in the back of your cluster. Pin 22 (grn/blue) is the tach signal from the ecu and the ground for it should be Pin 23 (tied to the rest of the grounds, so you probably don't need to worry about this one). You can also access it in the aux. relay panel in the lower black T10 connector. This is where the signal splits to the tach and the oil warning unit.

The reason the tach signal wire splits and also goes through the oil pressure unit is because the system is set up to alarm if the pressure is too low when the RPM is greater than 2000 or something.

[Mcstiff]

I'm feeding the gn/bl with VEMS RPM output, works fine for both RPM and the oil system.

Ummmm, updates? Tune Tune Tune Pretty boring to post about as I have not had time to install my MAC valve and I am focusing on a good emissions tune (kpamax=112 Lambda=1). Other than that, I have some maintenance that I need to do and I'll install the MAC for some more boost. No mechanical changes are planned till spring-ish, pending collection of parts.

That said, the car is 100% better to drive even with the WG basically open all the time lol.

Edit: Almost forgot, somebody decided that a rock should not be on an overpass...

[pilihp2]

Yay for progress!
Boo for rocks. good god that's gnarly.

[Mcstiff]

After driving some new Porsches (Panamera GTS, Boxster S, Carrera, Carrera S) at High Plains Raceway, I have decided that over the winter I need to refresh anything squishy. I daily drive the car but want to firm things up for canyon drives, track days, etc. I am hoping to get some kw v1s before next summer.

My "plan":

FRONT
Strut top mounts - 2 in total - Promounts or 034 Track Density or adjustable pillow ball (KW club sport style)
Wishbone bushes - 4 in total - I have delrin from 034
FARB bushes - Not running a front bar ATM
Subframe bushes - 4 at the front - solid alloy from 034


REAR
Strut top mounts - same as front
Wishbone bushes - same as front
RARB bushes - OEM for now
Rear track rods - OEM for now
Subframe bushes - 4 at the rear - same as front


ENGINE
Front snub mount - OEM
Left mount - Poly Mustang Transmission mount
Right mount - as above


GEARBOX- 911 engine mounts
Propshaft bearing - OEM


REAR DIFF
Three mounts here - 911 engine mounts & Powerflex rear mount

[yodasfro]

You want 911 cabrio/club sport mounts they are more better(stiffer) 91137504307 I used the rein brand those ebay HAMBURG TECHNIC brand are reportedly crummy.

[Mcstiff]

Cool! I was also looking at Wevo's black mounts but the jump in price is more that I'd prefer.

034 Track Density are also a reasonable option.

[yodasfro]

Yeah there's plenty of aftermarket options too from solid to poly to just poly inserts for the stock mounts. But to be honest the stockers are plenty stiff read almost solid as it is.

[Mcstiff]

I see a few 91137504307 manufacturers.

One benefit I see in the 034 stuff is that I know all the durometers match.

[Mcstiff]

Anyone else have comments on Track Density vs ?

[audifreakjim]

I run a poly front snub, poly mustang mounts up front and stock trans mounts. I tried stiffer transmission mounts and the NVH was maddening with my setup around 2500 rpms.

[Mcstiff]

So I have been messing with boost control a bit; still need more than 10mins to really get it working. Regardless, I found this description on an E30 forum and thought it was really good:

PID -
a link that makes it sound complicated - http://en.wikipedia.org/wiki/PID_controller

Basically in simple terms it´s the back bone of almost all closed loop engine control systems.

First you need a base duty cycle value/curve or a table. As duty cycle must have a starting point however simple or complicated it is.

Now if we say that you do a pull and your target was 200kpa but you only got 180kpa, technically you should re-do your duty cycle but lets imagine that you don´t.

Now what happens is that there is a 20kpa error in the target and actual. You then start increasing the P value and repeat, the P value represents an addition of duty cycle or reduction based on the error amount. So you´d think that since your off that loads of P would snap it back quickly. but the problem is that the addition of duty cycle is related to the error amount and P.

So with large amount of P you end up with to high of addition even while the error becomes smaller and smaller and will overshoot the target, upon the overshoot the ecu will apply a negative duty cycle addition. This will then overshoot below the target and so it will oscilliate uncontrollably.

So by starting off with a small P you stay away from the overshooting.
Now we enter a problem most ecu´s have is that they have a single value P, which means that at all error values the same proportional DC% is added.

i.e at 20kpa you´d get say 10% more DC%, at 10kpa you get 5% and so on. This can be cause problems with stability close to the target as in some cases even small DC% changes have X effect, but at large errors you need the large DC% changes. This is solved with a P value that is non linear so you can lower the P at low errors but make it very high at very high errors (VEMS doesn´t have this and most ecu´s dont)

So lets assume that after P tuning you reach 195kpa and we´ll imagine that the P´s effect is so small that it doesn´t change the DC% anymore and you are in effect stuck at 195kpa and can´t reach 200kpa. This is where the I value comes in.

The I represents a rolling DC% addition based on the error amount. So to cut this short you must not have too much I windup allowance (limits) otherwise the I might have wound up to far and cause very slow reaction as it needs to wind back down again (I´d like to have a kpa value that basically holds the I at 0 until you are X kpa away from the target and then it can wind, this will remove all windup problems and it could be a bit stronger as well).

The I with it´s windup will windup some DC% addition value that adds on to the base duty cycle and the DC% added by the P term.

i.e

Base duty + Pterm + Iterm = total duty cycle.
The I will now lift the DC% that last little bit to reach 200kpa.

This could be like

35% + 4% + 2.5% = 41.5%

The D acts as the damper based on the rate of change of the error. I.e
if the error is going down very fast then the ecu will lower DC% faster and faster to avoid overshooting.

I personally like to set the base duty cycle either at the least amount required to reach the least ever planned target and then have the PID add on all the DC% required to reach higher targets. Or sometimes I put in DC% values that are half way between the lowest and highest required DC% to reach whatever boost targets I may have. If I plan on running on the wastegate then sometimes I put in the values that are the highest without increasing boost (usually around 15-20%).

The boost control is useless until you set the base values
duty cycle vs rpm and frequency and then reach as close to your target as possible. First of course you need to do this as well as tune the fuel and ignition so that when you play with boost control all possible kpa points are already tuned.

After you get the PID values right then by testing using the Boost target curve you can add on the anytrim table that adjusts boost target based on
a 0-5v , this can be used to multiply the target or add on to the target, I use addition so at 0v I add on nothing, at 5v I might be adding another bar on top. So when you wind the knob back the target is usually wastegate pressure , say 150kpa, at the other end the target is 250kpa.

You can then fine tune the PID values further to improve behaviour with large boost target changes. I find using the I and allowing that to wind usually keeps overshooting down but always reaches the target.

[Mcstiff]

Also, found some clarity on just using Crank VE, Cranking threshold RPM, and enrichment table curve (without prime pulses or afterstart) to clean up starting. Fired right up (~3sec of cranking) with E85 <0ºC B)

[loxxrider]

Do yourself a favor and try to get some specs on the MAC valve. You wont get ANYWHERE without figuring out what pulse width that thing likes to work with... they seem to be pickier than the N75.

[audifreakjim]

Ed, I switched to a mini pressure regulator that is hooked to the top WG chamber. I still use a boost solenoid to enable and disable it, but I have never had a more stable and flat boost curve. The best part is you can adjust it on the fly and still use the boost valve to reduce boost in an rpm range or lower gear if you would like.

[Mcstiff]

It should not quite be rocket science, I've just barely had time to set it in the car. It is this valve so the documentation is out there: http://shop.vems.hu/catalog/high-capaci ... p-171.html

Part of my problem may be that I have it defaulted to divert 100% of MAP to WG with 0 power to the valve. I was thinking that this is safe but it seems it requires more PW. Of course, the wiki conflicts with the catalog http://www.vems.hu/wiki/index.php?page=BoostControl" onclick="window.open(this.href);return false;.

Jim, it sounds like a good plan; my top port is pretty tight to the chassis so I have been trying to ignore it

[loxxrider]

I did some quick research for you... not really that much info out there. I saw figures of 25 and 31 Hz. The stock VEMS settings isn't far off of that, so changing it probably wont do much. I wish I was there to help figure out what the problem is quickly for you. I'd definitely suggest trying out my boost settings as long as it is hooked up the "normal" way (valve connected to bottom of wastegate).

[Mcstiff]

Thanks for confirming! I plan on trying a few things this afternoon, updates this evening.

[Mcstiff]

Got it to hit my fuel cut (130 right now ); refDC=90 does not do much but 100 hits fuel cut (as it should). I should not need such a high cycle to get above 120 kpa (PID zeroed at the moment) so I am thinking another setting is off.

[pkw]

MBC

[loxxrider]

lol turn that boost cut up man. Turn that up and keep tweaking. There isn't much else you can change other than pulsewidth and duty cycle... and wastegate/valve vacuum configuration.

[Mcstiff]

Yeah, I turned it up to 150 and hit fuelcut with a higher pulse width. Slowly buy surely.

[Mcstiff]

So I just exceeded Colorado's HC limit by 0. 64 GRAMS PER MILE :frustrated: Such a poor standard to hold a 21 year old car to.

Regardless, any tuning wizards have tips? I'm targeting 0.99λ. My CO is well below the limit (<50%) and my NOx is about 15% below the limit.

My biggest gripe with the testers is that they shift at ~4k so they are not even really getting into the powerband.

I wonder if cutting my accel enrichment would get me done. I'm also wondering how much HC is getting pulled through my CCV (need a catch can).

I could throw a low pro cat in but that's ~$200 by the time I add vbands.

[Mcstiff]

I'm sitting at emissions and about to loose it since these idiots cannot drive. I was really hoping to fast pass with a new cat but it seems to not be in the cards. Hopefully it fiddlesticking passes!

I went back to E85 from 92, but made some quick adjustments on the way over here...


Edit: So the problem was the driver not the car! Passed after the second try (happens when they can't follow the speed curve).



Now on to figure out why I cannot get any real boost (just rerouted the lines so that the WGFV sends pressure to the top port when on).

[Audilard]

Be sure to save the file that passed before you start messing with tuning! Then you can reload the file.