WG tests

[Jim O.]

A few questions -
Are the runs you recorded at wide open throttle, or at partial throttle?

By "Boost", I'm assuming you mean intake manifold pressure, right? Is that the number you're getting from the KLR? Something looks wrong - the number doesn't indicate vacuum in the intake manifold, but the dash gauge (I assume that's also being driven from the KLR...) does indicate vacuum.

Is the sensor that you have on the compressor outlet psig or psia? Are you sure that the sensor is defined correctly in your logging/graphing software?

Looking at the top graph in post 101, something doesn't make sense to me - at the starting point, the measurement box says that the compressor outlet is under slight vacuum, but the "Boost" is at zero. If the compressor outlet is in vacuum, I would expect the manifold pressure to also be in vacuum. At the second measurement box, the compressor outlet is reading 2.43 psi, but the "Boost" reading is still zero.

Same thing for the top graph in post 102 - it looks like you're in deceleration, with an afr of 22, rpms were dropping, throttle must be closed. I would expect to see a lot of vacuum in the intake manifold, and close to atmospheric pressure in the compressor outlet. Instead, both numbers are only .5 psi apart. Seems a little strange...

If you could post a graph with the measurement box showing values at idle, it may help clear things up.

[Jim O.]

One last thought - has anyone measured the pressure in the line that goes from the CV to the wastegate? Might be able to see if it's back pressure blowing open the wastegate, or the CV or KLR actually opening the wastegate.

[soxnail]

A few questions -
Are the runs you recorded at wide open throttle, or at partial throttle?

Sox: At wide open…paddle to the metal.

By "Boost", I'm assuming you mean intake manifold pressure, right? Is that the number you're getting from the KLR? Something looks wrong - the number doesn't indicate vacuum in the intake manifold, but the dash gauge (I assume that's also being driven from the KLR...) does indicate vacuum.

Sox: It is from the manifold, the measuring system is set up to only report values above 14.7 psi so it ignores the vacuum.

Is the sensor that you have on the compressor outlet psig or psia? Are you sure that the sensor is defined correctly in your logging/graphing software?

Sox: It is gauge, and I have calibrated it with known pressure source while actually monitoring its output on a dvm and the logging software simultaneously. Both reported the same thing.
Transfer function used

Output (V) = {[0.8 x Vsupply/(Pmax-Pmin)]x (Pressureapplied – Pmin.) }+ (0.10 x Vsupply ) (ABP series sensor)


Looking at the top graph in post 101, something doesn't make sense to me - at the starting point, the measurement box says that the compressor outlet is under slight vacuum, but the "Boost" is at zero. If the compressor outlet is in vacuum, I would expect the manifold pressure to also be in vacuum. At the second measurement box, the compressor outlet is reading 2.43 psi, but the "Boost" reading is still zero.

Sox: Not sure I follow “post 101”
Boost will never show below 0 for my setup as explained above.
The compressor does lead occasionally I believe due to the delay in the pressure build up through the inter-cooler.

Same thing for the top graph in post 102 - it looks like you're in deceleration, with an afr of 22, rpms were dropping, throttle must be closed. I would expect to see a lot of vacuum in the intake manifold, and close to atmospheric pressure in the compressor outlet. Instead, both numbers are only .5 psi apart. Seems a little strange...

Sox: Same as above……

If you could post a graph with the measurement box showing values at idle, it may help clear things up.

Sox: see below

[soxnail]

I have looked at the WG to CV line, I put a gauge on this line and pinched it off to the CV so any pressure build up would be from the valve being pushed up by the exhaust.
At 22 to 25 psi Turbine inlet pressure there is an increase in the trapped air in this line.
Sox

[Jim O.]

How much pressure did you see in the WG to CV line without the line pinched off?

Also, another question about the last graph you posted:


Your note says "Notice the fluctuations in the turbine inlet pressure with the engine power strokes" - each tick on the x-axis looks like 0.1 seconds, which matches the times in the measurement boxes. That makes the peak-to-peak time between the peaks in the turbine inlet pressure line about 0.5 seconds. If the engine is running at 778 RPM, that means that it's at 12.967 revs/second, which means that 1 rev = 0.077 seconds. A four cylinder engine fires two cylinders per revolution, so you should see peaks every 0.038 seconds, if it is matching the power strokes, right? Or is something else going on?

[soxnail]

Yah you are correct...will do some more digging. It (turbine pressure) looks very erratic., maybe caused by the turbine spin up and idle coast in between ignition events? My rpm was steady and also when I took the line of the sensor I could fell the puff, puff coming out of it consistently, with the power strokes.
Will explore and post.
Sox

[soxnail]

Haven't forgotten this thread...
Maddog is down with a clutch replace. Will resume after the car is functional again.
Sox

[soxnail]

I am bored....hunkered down in my impenetrable bunker for 2 weeks or what ever, so here is an updated map showing the theoretical ability of the k26-6 attached to a 2.5 well 2.495 liter engine.
Sox

[soxnail]

Ok back on this project again. Since car is now running and in good shape.
Re-calibrated the 3 pressure measuring sensors, I can't get rid of the slight discrepancy between the car's boost gauge and other pressure sensors.
Anyhoo did a test the pic posted here is for change from 3rd to 4th gear and then WOT till traffic interfered. Line to WG was pinched off as I just wanted to test the WG's spring without the KLR bleeding the boost.
The compression on this page makes the data boxes too small to read so I posted them separately.
Sox

[soxnail]

Another run, with CV enabled and a sensor attached to the CV to WG control line.
Looks like about a 4 PSI difference in boost. The sensor outputs have some smoothing done to get rid of the spikes.
Sox



Data boxes expanded....

[nize]

that looks way more complicated than this one:

[soxnail]

Well my WG/turbo is characterized now, so on to my target of 13 to 15 psi of boost.
Sox

[soxnail]

Quick note, what I tried to do so far is to change the gain of the amplifier that the pressure sensor drives. So the signal coming out to the KLR processor is lower than the actual value of the boost.
This causes the KLR to delay bypassing the boost to the WG till a higher number is achieved. In my case the KLR thinks 13 psi is 10 psi. So it lets the pressure build up to 13 plus psi before actuating the WG. Of course at 13 psi the exhaust is already blowing past the WG valve. (See last couple of posts).
This works as far as getting higher boost from the car..There are two issues with this approach.
-1 the car's boost gauge reads low, no biggie as it is not very accurate anyway.
-2 the high boost cut off from the KLR becomes harder to set, because the sensor runs out of usable range at about 14.5 psi.
The low boost side continues to work, with the low boost fault getting set at a 3 psi higher point than before.
The rest of the KLR functionality is retained.
The above reads pretty convoluted...short version trick the KLR into thinking the boost is 3 psi lower than it really is.
Opinions, pitfalls etc. That I am not seeing ?
Sox

[Mike Goebel]

Correct knock retard for 3 psi higher boost pressure?

Mike G.

[soxnail]

Knock retard will happen at the same place. The retard for high boost will shift by 3 psi.
Sox

[nize]

i don't recall anyone trying to trick the KLR into thinking the boost is 3psi lower than actual, but if someone would know this trick it would be josh at rogue tuning because he basically reverse-engineered the whole thing. here we are 30+ years later and it's not a common hack so i'd be super surprised if it actually worked.
Rogue Tuning - Tuning freeware:

[soxnail]

The 3 psi increase does happen. It's probably a path less taken as it required a redesign of the electronics in the pressure sensor.
If one has already de compiled the code and knows where to modify it...that would definitely be the better approach.
As I said this works, but to get the over boost protection I am going to use a sensor with a larger range, so it does not run out of "headroom".
Finally the KLR's native control of the CV is very jittery due to the processor getting reset at each trigger event. This causes spikes in the boost signal and erratic control of the CV.
I am designing a software based filter to read the output of the KLR and then use it to continuously modulate the boost control. The existing CV is not meant to be used for precise PWM.
This thread seems to getting tangled up in my KLR thread
Anyhoo will keep y'all up dated.
Sox

[nize]

another issue contributing to inaccurate control of the wastegate is that the factory wastegate is a single port. converting to a dual-port wastegate improves control.

[nize]

also, why not add some sort of physical restriction/control to the KLR boost sensor line to consistently reduce the pressure by 3psi? that seems easier than re-engineering the electronics.

[soxnail]

It would be hard to do a consistent job..as the line to KLR does not have an outlet to be able to maintain a 3 psi differential.
I am moving this discussion to my KLR thread as the WG part is done.
Sox