Computers "think" in binary--On or off (0 or 1), technically +3V or ground on many chips. Everything is in powers of 2. 255 is actually out of 256 technically (255 + the value 0 ), or 2^8 power. 2^8 is 256. Some of the other settings in the PFC are out of 255/256, such as INJ vs accel TPS 1 in settings 2.

Remember that the more air bled from the actuator (higher duty value), the HIGHER the boost will be, and the opposite is also true: lower duty correlates with lower boost. Don't get the relationship confused. To some extent the PFC will drop duty in an overboost situation. Your target secondary duty value is 76%. The PFC dropped it to 65% (and in some cases 60% in this log) in an attempt to correct the overboost. One thing i've noticed lately is that in some cases the PFC will "hone in" on a boost value between .05 and .10 kg/cm^2 higher than what you set it at for reasons unknown to me. There are times when you need to set the target boost lower. It's not a perfect system unfortunately.

Also, any closed loop control logic will have what's called a "controller authority," or the amount the self-correcting logic is allowed to deviate from the set duty value given certain conditions. This "controller authority" is actually adjustable in the fuel correction closed-loop circuit of the Megasquirt standalones for example. We have no control over the PFC's controller authority for boost control. I usually see it deviate up to 5-7% or so from the set secondary duty value. The fact that in your case it is dropping 11-16% from the set value indicates that it is doing the best it can and it is probably functioning within its design parameters. If your restrictor pill is removed but your wastegate doesn't flow enough, you need to correct that mechanical aspect of the system, or if there is some other sequential problem you need to fix that.

Before the "high" transition point in the PFC, the wastegate is held at 95% duty (or 243/255). This is exactly what the FSM specifies for pin 4U, page F-162. Once the transition point has been reached and the turbo control actuator ("TCN" switch) is engaged, the max allowable duty appears to be 225/255 (89% or so) for whatever reason. If you set your WG duty (Sc duty) to 95% in the PFC it will still be capped at 89% once the TCN has been switched. I've seen this during dyno runs where a mechanical problem kept the PFC from holding boost up top. I tried to correct the boost drop with more boost control duty but it capped at 89% anyway.

Right after transition the PFC may keep the wastegate duty at 89% in order to improve spool, and then drop down when its control logic dictates that the wastegate should be opened. Remember that the lower the duty, the more the wastegate is opening and the less pressure is vented from the actuator (lower boost). The more pressure an internal wastegate actuator sees (less duty, less venting), the more it will stay closed. FYI, I tend to disregard the primary duty value. If it does have a purpose, I can't figure out how it is used yet.

I hope that helps some. I'm not claiming the PFC is perfect, nor do I claim to understand everything about it. I am still testing it in my custom single turbo boost control project which I have discussed in another thread ( https://www.rx7club.com/3rd-generation-specific-1993-2002-16/free-power-fc-single-turbo-probably-nonsequential-boost-control-846883/ for anyone interested ).

Thank you very much for your exhaustive explanation.

Clearly, I got confused regarding the WG solenoid duty cycle.
Now, it appears then that the "inability" of the PFC to deviate much from the "initial" duty cycle set (76% in my case for the secondary) is of no help in mitigating boost creep, but actually would make it worse.
I will try lowering the setting of the initial duty and see what happen under those conditions. I still have the pills, so maybe that even with lower duty at the bleeding valve, the WG would not open quickly on boost build up, we'll see it.

Now the question is, could an AVCR be more effective? I have one still in the box, which I bought after reading Dale Clark's notes a couple of years ago. I never installed it because the boost control system in the PFC appeared to be sufficient (although - as indicated in my previous post - I still have to investigate the reason for the large boost drop while transitioning from primary to parallel turbos operation). In case of overboost (like my boost creep), would the AVCR operate the feed solenoid valve at its max duty (AVCR solenoid valve in the feed line and bleed line plugged)? If that is the case, I would install it for this reason alone. If the WG could be forced full open by the control system, I might perhaps not experience boost creep at all, or would certainly mitigate it more effectively than with the PFC

Thanks,

Sandro

https://www.rx7club.com/attachment.p...1&d=1245422564

It can drastically deviate from the set value in some cases. Here is a log from my other thread of me controlling a single turbo with the PFC and a factory FD wastegate solenoid. Duty was set to 20%, far too low. The PFC made drastic corrections to slow the boost drop, but it can only do so much. If the duty value actually needs massive oscillations to keep the boost consistent, something in the system isn't right.

If your wastegate doesn't flow enough, your wastegate doesn't flow enough. Exclusively electronic solutions can't fix mechanical problems. Try this: remove the restrictor pill if you haven't already. The lines that go from the actuators to the precontrol and wastegate solenoids, cap off both ends of those. You should now be running spring pressure on both actuators and the wastegate is effectively "full open." Do a few WOT pulls and post the results. In my experience with internal wastegates (series 5 single turbo, internal wastegate), it will creep under higher gears even at just spring pressure. So put that bitch in 3rd and 4th on the highway and log the boost curve. Factory spring pressure is like 7psi right?

Boost creep is very common on the factory 2nd gen turbos. A boost controller cannot increase the flow rate of a wastegate. When people buy aftermarket boost controllers they are removing the restrictor pills and introducing new control logic to the system. But if the wastegate is undersized it's undersized. For example, I tried a lot of different things on my old 40mm HKS external wastegate, but it just didn't flow enough. The boost pattern was inconsistent with the weather, with or without a boost controller installed.

https://www.rx7club.com/attachment.p...1&d=1245512600

Here is an old log of me with that undersized wastegate setup. All I had was pressure going from the compressor housing directly to the side port of the wastegate to force it open as early as possible with no boost controllers involved. In purple I have drawn a line showing where the boost "should" reach, with maybe a few slight oscillations around that point. But because the wastegate was undersized, it opened completely but boost kept climbing, albeit slower.

The "inflection point" in the boost curve is where the boost climbs much more slowly due to the valve completely opening (second derivative is negative if you know anything about calculus). FYI velocity is the first derivative (derivative = rate of change) of an object's position at a particular point in time, and acceleration is the second derivative of an object's position--a measurement of how fast the velocity is changing. But I digress.

The area between our achieved peak boost value and our spring pressure is what I will call our "boost creep integral," the net amount of overboost over a given period of time. An integral is just calculus terminology for the area under a curve.

I hope that clarifies a bit the difference between improper boost control settings and insufficient wastegate flow. Always remember that boost control is a complete system, not just a discrete device.

However, my situation appears to be the reverse of what shown in the chart you posted, if i understand it right. Although the the boost pressure is well above the set pressure during boost creep, the bleed solenoid duty is low. My rationale/question is that, in addition to the "physical" constrains represented by the unported WG, any bleed duty present would certainly be of no help and likely make things worse. If the control system of the PFC were able keep the bleed line shut, all the the pressure in the WG actuator chamber would force the WG full open. If bleed duty is present (like my 65%), it may be that the WG would only be partially opened. That is what I meant when I hypotized that with a different control system (AVCR) I may not experience boost creep at all. Keep also in mind that my runs are in 2nd gear, so the load is lower than 4th or 5th, where normally boost creep may become unavoidable, as I understand.

That prompted my question if an AVCR (i.e. just a device) which utilizes a different control system (by controlling the feed, nor the bleed) could work better against boost creep.

Clearly, the test you suggested would prove whether physical constrain (too small WG) is present, which could not be solved by any control system. For me it is a little difficult to perform it because I only use my car for autox. I'll see if I can do it in some test and tune session one day.

Thanks again.

- Sandro

The diagram I posted was to illustrate that the PFC can make big adjustments, but it shouldn't have to. In most cases the results will likely be disappointing. One reason is because there is always a time delay when you make a drastic change to boost control solenoid duty. Remember that there are hoses and pressure waves involved, unlike fuel injectors (which are duty controlled) and which sprays right into the intake charge.

The second reason is this: if the correction logic is sensitive enough (on other EBC's, gain setting would be too high) in the system, there will be constant overcompensation and undesireable oscillations. It's the same reason why cruise control computers tend to undershoot the target speed when the car is climbing a hill. The factory engineers could have given the cruise control system more authority to respond more quickly (more gain) and apply more throttle (more controller authority), but there is too much of a risk of cyclical overcompensation (sinusoidal curve). There is nothing more unpleasant to ordinary Joe Driver than a perception of hunting or surging in any aspect of vehicle operation. But even performance oriented drivers don't like hunting either.

That's why engineers build systems that don't require too much second-by-second closed-loop correction. On modern cars they have layers of fuel correction logic and self-learning. There is the immediate correction to the fuel mixture (short term fuel trim). When that level of immediate correction is significant (usually over 5% richening up or leaning out), the ECU builds long term fuel trim. The long term correction tells the ECU to make corrections before it even starts the whole second-by-second adjustments. Subaru is using this type of logic on their ignition advance tables and boost control systems.

Again, closed loop systems are not designed for drastic changes. Why do you think these cars have restrictor pills installed from the factory? If the amount of pressure entering the wastegate is predictable within a tight range, drastic changes in solenoid duty will not be required.

From the AVC-R instruction manual:

https://www.rx7club.com/attachment.p...1&d=1245543855

Even the AVC-R is not designed for making massive changes.

I'll put up Sandro's log to illustrate a few things. He goes through transition one time and then stays non sequential for the rest of the run. First, take a look at this description of how the solenoids should operate from the factory:

https://www.rx7club.com/attachment.p...1&d=1245600017

Now here's an excerpt from Sandro's log. Sorry there's so much crap on the screen, I'm trying to get a lot of points across with one screenshot. I am assuming he is using default turbo transition settings.

https://www.rx7club.com/attachment.p...1&d=1245600017

symbols:

TCN ---> Turbo Control solenoid. A value of "1" means it is energized, "0" means it is not.

CCN ---> Charge control solenoid. A value of "1" means it is energized, "0" means it is not.

PIM ---> Manifold pressure value. 10000 is atmospheric pressure, 20000 is 1.0 kg/cm^2 boost, and any values in between

PC% ---> Precontrol solenoid duty. More duty = actuator more closed (higher boost)

WG% ---> Wastegate solenoid duty. More duty = actuator more closed (higher boost)

TPS V ---> Throttle position sensor voltage. Technically 100% throttle is somewhere between 4.2 and 4.6 volts

Raymond, my settings were slightly different, although they did not impact the explanation of the operational steps that you summarized so clearly. For your reference, relevant settings in the run were:

- turbo transition

TPS(%) Low High RPM
60 2800 3600
40 2800 3760
20 2800 5120

I am using 2800 Low to make sure the turbo operation remains non-sequential for the entire run (after the first transition at the start)


- Boost control

Boost(Kg/cm2) Duty
Primary 0.90 62%
Secondary 0.90 76%

Thanks,

Sandro

another log to illustrate turbo transition:

https://www.rx7club.com/attachment.p...1&d=1245787926

settings for that test run:

https://www.rx7club.com/attachment.p...1&d=1245787926

Raymond,

please take a look at these two charts.

The first chart "high duty" is with these settings
Primary 0.90 Kg/cm2 62% initial duty
Secondary 0.90 Kg/cm2 76% initial duty

The second carts "low duty" is with the same settings but the initial duty
Primary 30%
Secondary 30%

Note my map is scaled as follows
P18 18930 PIM
P19 19589
P20 19600

The P20 row is a safety row I use to stop boost creep by pulling the IGL to zero deg

As you can see, with the "high duty" setting, I overshoot boost setting and WG% stays high at 62.5% - which confirms your observation that the PFC is keeping the closed loop operation of the bleed solenoid withing a relatively close range of the initial duty setting. Therefore, under this conditions, given the high bleed duty, I suspect that the WG is just prevented from opening fully.

On the other hand, with the "low duty" setting, WG% is 27% (given the low initial 30% setting) and PIM is limited to 17445.

However, the problem with 30% is that the boost builds up very slowly and during the run never reaches the desired setting of 0.90 Kg/cm2 (approximately 18900 PIM). The max I got during the run was 17200 during the initial transition primary/secondary turbo with the WG "closed", i.e. WG=243 and PC=75

- Sandro

FYI, try viewing these logs with a bunch of separately scaled charts so you can see the numerical range of each trace better.

I'm not sure if you looked through my thread about using the sequential turbo control system to actually control a single turbo: https://www.rx7club.com/showthread.php?t=846883. In that setup the factory WG solenoid is controlling flow to an external wastegate diaphragm which pushes the gate shut. So more duty results in higher boost just as in the stock plumbing for a sequential setup.

I have also tested using low duty cycle numbers and have seen somewhat similar results:

https://www.rx7club.com/attachment.p...1&d=1245422564

But you see here that raising the duty up eventually resulting in a boost curve that is almost flat:

https://www.rx7club.com/attachment.p...1&d=1245422564

Since I made those logs I have still been playing around with the car and I hope to have more logs soon, where the boost doesn't really drop off at all. Assuming your system can mechanically control the boost well enough (plumbing and wastegate flow capability), you should try to start with low duty and gradually increase. Crank up both solenoid duty (primary and secondary duty values) working up from that 20% value, first in 10% increments and then maybe less than that as you get closer to an acceptable boost curve.

I'd ramp up both at first to a point, then adjust primary duty only until it's pretty good. Then adjust secondary, but you may have to jump back to the primary value again after that. Once you start getting close you will have a lot of fiddling back and forth among all the settings, including the boost value. You may have to increase or decrease the boost value by a small increment to get near where you want. The duty curve we see is a function of the target boost, the initial duty, and the feedback logic that we are grappling with right now which is based on the PFC observing the boost curve.

Since those logs I posted here, I now run .95 boost and 50% duty on my single turbo setup. I'm still fiddling with it some, but with a 13psi spring pressure (.91 kg/cm^2), I get right at .95 in first and then a value of about 1.05-1.09 in 2nd and 3rd, which is actually what I want anyway. I haven't logged my boost curve recently, I've just been looking at peak values in the Commander. The turbo I am running is a T04R (bigger than a GT35) which is going to be sensitive to the load differences in each gear.

This may seem odd... but here is an everyday application of a restricter pill:

https://www.rx7club.com/attachment.p...1&d=1249842017

Yes, that's a bottle of barbecue sauce. Now think about it for a second. When you squeeze a bottle of sauce with a restricter in it, it comes out in a much more predictable manner than just taking off the whole lid. There's only so much sauce that can come out of that orifice no matter what's in the bottle or how hard you squeeze. In the same way, restricter pills on OEM boost control systems allow the factory engineers to always have a pretty good idea of how much air is flowing into a wastegate actuator.

Raymond, I eventually was able to put the car on the road and do some logs last weekend.
I will post the results once I'll have the time to sort them out - which will take some time.
Pulls were mostly by sharply applying full throttle in 1st and 2nd gear, to simulate autox conditions. A few points after taking a first look at them:
1. Confirm that PC opens at 225 and stays there for a while and WG opens at 243 and stays there until transition
2. After opening at 225 PC duty undershoots then tapes off at the Primary setting duty while still within the 255 WP duty "envelop". It appears therefore that - at least for these kind of sharp transients - the PC duty rapidly approaches the Primary setting duty.
3. At the start of the transition, PC duty goes to zero and WG% goes to the "initial" duty value as set in Secondary duty
4. Not clear from my sharp and quick pulls how does the P and S boost settings play; anyhow duty goes up and down like trying to hunt a target (boost setting or % duty, or a combination?) while the boost builds up
5. Now, about my boost creeping: If I use low S duty setting I generally don't have any because the WG solenoid has the time to close enough to compensate - it appears there is an exception to this in the 8500-9000 rpm range but those are quite extreme conditions. On the other hand, as I had noticed already, if the S duty setting is high, the control system appears just incapable to open the WG as it can move only within a relative small % from the "initial" duty.
6. Another "extreme" case (but I only have one or two logs about this) is that if I pull sharply and full throttle in 3rd gear (higher load) I get overboosted because the WG duty is still in the 243 "initial" duty period (independently from the duty % setting).

In the end, by focusing on the 2nd gear pulls only, I seem to have found a workable compromise by setting the P and S duties at 67% and 73%. That takes me quickly to around 13 psi without hitting my safety row I set at 19600 PIM. But it appears though that if I wished to run - say - 14 or 15 psi the PFC control system could not handle it with my open exhaust (catless and N1 dual). And, also clearly, this has noting to do with my unported wastegate; it's just that the PFC does not open the WG completely and quickly enough.

Question: how to copy the PFC charts and post the full image as you do? In the past I made screenshots and pasted them onto Paint, saved as jpeg then attached to the message. Is there a more efficient way of doing it and how to make appear the chart images in full size within the message rather than just attachments?

Thanks,

Sandro

RYes, I too have confirmed that the Primary duty value is used to control boost before "Turbo Transition high" rpm is reached.

Yes, if it doesn't reach the exact secondary duty value it gets pretty close.

I have been doing some testing on this. I still can't figure out what the primary boost value does, except that it might have a minor effect on when the precontrol duty drops from 225 to near the target primary value. Some level of boost oscillation is inevitable; that is the nature of PID (proportional, integral, derivative) control or whatever variation Apex'i uses. It's more a matter of getting the system to operate within an acceptable tolerance than trying to eliminate fluctuation altogether.

If you post the logs I can offer further opinion/analysis.

This is a limitation in any closed loop system. There is only so much variation that the system can correct for before overshoot or instability occurs. On any boost controller, if the duty values are set too high some kind of overboost will likely occur. Factory boost control systems are always tuned so that the base duty cycle value without any kind of closed loop feedback would result in underboost. Then the control logic/coefficients raise the boost closer as close to the target as possible, but the sensitivity of that control is set such that the duty and observed boost will not reach an unstable state. Consider these factory boost control tables from a 2005 Subaru Legacy 2.5GT, which is basically an STi engine with a smaller turbo:

https://www.rx7club.com/attachment.p...1&d=1254781239

3 dimensional table of target boost (measured in bar) vs rpm and accelerator pedal position

3 dimensional table of base wastegate duty cycle (before closed loop correction) vs rpm and accelerator pedal position

All the Subaru engines use single port internal wastegate actuators with a two-way bleeder solenoid and a restrictor pill, functionally the same as the FD's. So here we have a lookup tables of what boost the ECU is trying to achieve, and the base duty value that will be used to get us there.

https://www.rx7club.com/attachment.p...1&d=1254781239

Here are several two dimensional tables which adjust the boost control duty. The first is air temperature correction in celcius vs the amount the duty cycle will change. The second is some kind of duty table vs rpm, to reduce the chance of overboosting or underboosting because of changes in rpm. The turbo dynamics (Fine Gain) is the proportional coefficient for the closed loop control. The turbo dynamics (Coarse Gain) is the integral coefficient. Derivative coefficients are usually set to 0 in most boost control setups.

https://www.rx7club.com/attachment.p...1&d=1254781960

This graph I modified off a thread in the Haltech forum indicates which PID coefficient predominantly affects which areas of the boost curve. These types of gains would be set in any other ECU's boost control system (Haltech, AEM, etc). The PFC picks these numbers for us, sacrificing adjustability for an easier learning curve.

One method to determine if there is a mechanical limitation to your boost control would be to cap off the line from the WG and PC actuators to their solenoids, then remove the restrictor pills. It should be able to hold spring pressure. I have not seen any logs of the creep/overboost you were experiencing so I will have to withhold judgment on that.

What I will say is this: At any given target boost level setting, the duty values appear to correspond to different ranges of peak boost. I am currently controlling a single turbo with the sequential boost control logic. A target boost of .95 with duty of 50 yields observed peak boost values of 1.07 +/- .03 . A target boost of 1.00 with duty of 48 yields observed peak boost of 1.15 +/- .02 . I need to take some more logs to observe the duty curve a little more. You may be able to try lowering your target secondary boost by .05 or .10 kg/cm^2 and raising the duty by some percentage.

http://download.cnet.com/Screen-Prin...-10135610.html

Screen printing utility, I think that's what I'm using. If you set it up right it will let you select an area of the screen to capture. Then after you attach it, preview the post. Right click and click either "copy image location" or "properties" and copy the location from that.


Thanks,

Sandro[/QUOTE]

Thanks Raymond. Here is a log.

This is from a combination that works:
P 0.95/65% S 0.95/73
2nd gear pull
throttle pedal controlled not to exceed 8,000 rpm
Note:
To start with, WG stays at 149 @ 95% for the entire period before transition
PC starts at 225 @ 88% duty for less than one sec, then undershoots a bit but rapidly recovers to reach the set Primary duty in less than one sec (not bad!)
PC closes at 168 @ 65.6% right on the money with the P set duty of 65%
after transition WG is 189 @ 73.8% right on the money with the "initial" S duty
Primary max boost before transition is 18597 or about 12.5+ psi
Boost after transition is 18500-19000 or about 13 psi
Note when I tape off the throttle (at 7 sec on) the WG duty tops 225 (one of those "fixed" numbers) or 88% duty, attempting to keep the boost up
https://www.rx7club.com/attachment.p...1&d=1254789358

- Sandro

attachment doesn't work

I must be retarded but this is the best I could do. When I copy "location" it doesn't let me paste it into the message...

https://www.rx7club.com/attachment.p...1&d=1254827156

Anyhow, the thumbnail of the image has loaded (I replaced the previous .pdf you could not open with a .jpeg). I am also attaching the source log.

Copied below are my previous comments with a minor correction on the WG value before transaction, 243 not 149 (typo)

This is from a combination that works:
P 0.95/65% S 0.95/73
2nd gear pull
throttle pedal controlled not to exceed 8,000 rpm
Note:
To start with, WG stays at 243 @ 95% for the entire period before transition
PC starts at 225 @ 88% duty for less than one sec, then undershoots a bit but rapidly recovers to reach the set Primary duty in less than one sec (not bad!)
PC closes at 168 @ 65.6% right on the money with the P set duty of 65%
after transition WG is 189 @ 73.8% right on the money with the "initial" S duty
Primary max boost before transition is 18597 or about 12.5+ psi
Boost after transition is 18500-19000 or about 13 psi
Note when I tape off the throttle (7 sec on) the WG duty tops 225 (one of those "fixed" numbers) or 88% duty, attempting to keep the boost up



Thanks

- Sandro

I took a look at the logs. That first dip in boost after transition is something that every sequential car has to deal with. It's a mechanical limitation of the system. The turbo control and charge control actuators were not optimized for people cranking up the boost, and there is no easy way to adjust them. They come on abruptly at the transition point (turbo transition high setting). I noticed that you were at part throttle in this log. I think your Turbo Transition High - 40% was at play. You could attempt to play with that setting a little bit to see if switching the charge control earlier or later affects the boost dip.

Based on this log I don't see anything major here that could be optimized. If you were using an aftermarket EBC and tuning it with an external boost gauge, a similar behavior would likely emerge after transition. There are two differences though:

1) an aftermarket EBC does not let you easily log duty cycle in a readable way. So you'd have no idea that the duty had been fluctuating in an effort to stabilize boost.

2) external boost gauges usually smooth the boost signal a little bit so that minor fluctuations are harder to notice. In contrast, logging an appropriately calibrated pressure sensor through the Datalogit shows you all the fluctuations. I also see that you have a pretty good sample rate here (around 25 samples/second). That's going to make all the minor fluctuations more visible.


Also, about the screenshots. In internet explorer:

https://www.rx7club.com/attachment.p...1&d=1254840286

here I clicked on the thumbnail (do it in the preview post screen). Then I went to properties and copied the link.

In firefox:

https://www.rx7club.com/attachment.p...1&d=1254840286

Here I opened the thumbnail into a new tab (the middle button/scroller thingie on my mouse does that automatically). Then I right clicked and used "copy image location."

Thanks Raymond. As in the premises, this combination looks like a good compromise - and I do not plan on increasing the boost any further right now, so the PFC boost control system seems OK for my current targets. With higher boost settings, i.e. higher secondary duties, I am faced with the "boost creep" issue I described earlier already - for example see please my previous post #25; the sec. setting on those runs were at 76%

I have been thinking more at the PFC approach of controlling the actuators by acting on the bleed side. What puzzles me is that to achieve a high rate of boost build up, the PFC "locks" the actuators "close" by bleeding air from the WG actuator - for the entire time before transition, and for a good portion of the time from the PC actuator as well. Incidentally, with this type of control, qualitatively it seems to me "better" keeping the pills (which acts like a "I" in your PID comparison). With no pills the pressure would build up faster on the PC and WG actuators and "push" the actuators to open. Anyhow, this is just from a qualitatively point of view. Maybe the "imposed" 95% or 88% initial duties are high enough to bleed the excess air even after the pills are removed, and the resulting pressure on the actuators would still be insufficient to overcome the actuator springs, don't know...

I was just curious though, what is your comparative opinion between PFC control (bleed) and "feeed" type control, like EBC or check valves, Hallman type? Based upon what I have been logging and learning, it seems to me that if I ever want to run at - say - 15 psi, an upgrade to a "feed" control type might be necessary.

Thanks.

- Sandro

Since I mentioned my previous post #25, here is an image of the chart. Sorry I still can't show the image. I followed religiously your detailed and clear instructions - with Firefox - but when I "Copy Image Location" onto the message are I only get the link transferred, not the image. The only way I seem able to do it is to import it into my Album and then copy and paste the link which gets then bracketed by [IMG]....[/IMG] too many passages...

https://www.rx7club.com/attachment.p...4&d=1254851120

In any event, please refer to the attached thumbnail image.
Run is in 2nd gear with Secondary set at 0.90 and 76% duty
Between 328.0 and 328.5 sec I get boosted to my safety row and the IGL gets pulled.
Max PIM is 20165 when I start easing on the throttle (starting with TPS V 4.248); my safety row 20 is set at 19600 and row 19 (normal IGL) is at 19589. WG is at 185 (before starting decreasing) after the safety row is hit. My theory is that although the boost increased beyond the .90 target, the duty on the WG bleed solenoid did not drop quickly or widely enough to compensate for it.
Nothing to do with my un-ported WG, just that the WG does not open enough because the bleed duty is too high.

Note that the run I posted earlier at 73% set duty shows no problem despite a higher sec. boost target of 0.95 and a higher rpm (8,000). Which shows how sensitive this control system is even to such small variations to the duty cycles set.

I have other runs I can post later with low duties, which show there are no mechanical limitations (a kind of equivalent to the procedure you suggested to cap off the bleed lines and check that the 7 psi spring rate holds), unless perhaps for very high rpm, beyond 8,000 (more later).

- Sandro

This seems a bit abrupt to me. Maybe you should give the boost a little more leeway to fluctuate. It's your call though. My overboost protection is the PFC overboost fuel cut (I didn't want to rescale P rows), and I have everything set up such that if boost rises about .05 kg/cm^2 more than I want, it will fuel cut. The PFC is supposed to cut fuel when boost is .25 kg/cm^2 more than the target boost value, but I think it's more like .22 . My boost is set to 1.00 kg/cm^2 and I actually run about 1.16 with my current duty settings. So if boost hits around 1.22 it will cut fuel just like the overrev cut.

Hit the insert image button above ^ and paste it into there, that'll populate the [IMG] tags. That's what I do. Or you can type them in manually.


Running a high duty cycle during spool is the proper approach regardless of plumbing, and that control strategy is utilized by all the OEM's. Look at the OEM Subaru target boost and duty cycle table again.

https://www.rx7club.com/attachment.p...1&d=1254781239

The wastegate is completely shut (max pressure bleed on the internal wastegate actuator, or it would be full pressure applied to top port of am external wastegate) until right before the target of .94 bar is reached. This is the same as the "START BOOST" function in the Greddy Profec Spec II, the "spring" function in the AEM Tru Boost, and the START DUTY function in the Apex'i AVC-R. The "I" portion of the PID control graph I posted (which I didn't make originally) is kind of an oversimplification of how those systems work. Both the proportional and integral coefficients work together with the base duty value tables. We just can't adjust them in the PFC.

I'm not completely sure what the answer is to this, but I suspect the restricter pill is usually needed for higher boost, although "higher" is a relative term and depends on the design of the rest of hte system. What I will do though is post several OEM boost control routing diagrams for you to look at.

https://www.rx7club.com/attachment.p...1&d=1254862363
Rx-7 S5 Turbo II, bleeder valve tee'd off without a restrictor pill

https://www.rx7club.com/attachment.p...1&d=1254862363
Factory STi/Evo boost control plumbing (top) and then a 3 way solenoid installed (bottom) with the NC port rerouted to the intake to eliminate noise.

https://www.rx7club.com/attachment.p...1&d=1254862363

Porsche 944 with factory 3-way interrupt type boost control solenoid. The solenoid interrupts the signal to a single port external wastegate. The single port external wastegate is controlled with a single diaphragm chamber just like an internal gate

https://www.rx7club.com/attachment.p...1&d=1254862363
Vr-4/Stealth . The restrictor pill is actually on the inlet to the bleeder solenoid to keep boost down. The pill was removed in the 94+ models which explains the increase in rated horsepower from 300 to 320.

https://www.rx7club.com/attachment.p...1&d=1254862363
Nissan S14 SR20DET motor, notice two restrictor pills ("orifice"). These combine the Subaru design (restrictor to wastegate inlet to raise boost) with the Mitsu design (restrictor to solenoid inlet to lower boost)

If it were me, I would experiment with running a 3-way solenoid on the wastegate and keep the bleed type on the precontrol as it seems to be doing a good enough job. Remember that the PFC supplies a duty signal and that's all it does. It can work with whatever plumbing you want. I do know that you will likely need to lower your duty values if you go with a 3-way solenoid. The easiest solenoid to use would be the 3-way solenoid that comes with the boost control kit:
http://www.2kracing.com/images/produ...xi499-x003.jpg
You can buy that separately from the kit here: http://www.2kracing.com/Product/321.html . To install that you would cap off the line that goes from the WG actuator to the WG solenoid. Then you'd run a pressure source to the NC port and run the remaining wastegate line to the COM port. The solenoid should plug right into the factory harness I believe.

Thank you very much for all this valuable information Raymond. I really appreciate it and will study into it.

Just to clarify some of my thoughts/questions/choices

On the protection row at 20. Note my target is to keep the boost within - not at - 19589 (row 19). That is why have been attempting to optimize the boost in 2nd gear just to avoid hitting row 20. My target is actually to move not too much around my row 18 which I set at 18930 (about 13 psi). with my scaling there is about one psi difference between row 18 (my - hopefully - operative row) and row 19 (my limit). My idea of row 20 is actually a safety row as suggested by Chuck, hopefully not to be ever hit under normal conditions. But clearly it remain to be seen if this will actually be possible with the PFC.

On the initial "locking" of the duty cycle. I understand clearly this is what needs to be done with a control system based on controlling the bleeding. My question is more about learning if the control system that control the feed, rather then the bleed, like the EBCs with solenoid valves in the feed line (and no bleed) or the spring/ball check valves may actually work better. Please keep in mind that I have zero experience on those. My car was stock until last year and it is only recently that I have been started playing with the PFC with regard to the boost control. So far I had been focusing on the map and I am OK with that now.

Thank you again for sharing your experience.

- Sandro

Feed vs bleed--the main difference is the required duty cycle to achieve a particular boost curve. The "control system" is the same: a basic duty value plus various corrections. Duty is high during spool, drops to flatten out the boost, then fluctuates to keep the boost within a stable range. That's in any type of boost control system, whether it is bleeding air from an internal wastegate actuator, blocking the pressure signal to an internal wastegate actuator, or feeding pressurized air to the top port of an external wastegate. For what it's worth, Subaru owners frequently ditch the bleed style solenoids when they want to run a bigger turbo. They install a 3-way type solenoid ("feed" solenoid) and plug it into the factory harness. Then they adjust the duty cycle and gain values in the factory ECU using reflasher software like I have already shown.

Plumbing (restrictor pill, actuator type, solenoid type, etc) and control logic (base duty curve, closed loop control) are two separate but related parts of the overall boost control system. That's why aftermarket EBC's all work with one solenoid in different plumbing configurations.

Here is another log - eureka! I was finally able to embed the image...

https://www.rx7club.com/attachment.p...8&d=1255488264

Primary 1.00 / 67%
Secondary 0.90 / 65%

from the top chart you can see 4 legs
- smooth full throttle
- throttle partial lift off
- short full throttle application
- throttle full lift off

Comments
1. primary / sec transition, PC% and WG% are very close to their 67% and 65% settings - as already noticed in my previous logs
2. primary boost quickly reaches and stays around PIM 19500, except for a few msec at around 126.3 sec when my safety row (19600) is very briefly hit (recognizable by the dip in the IGL)
3. at 127 sec, WG% momentary drops to 87 (34%)! Looks like some strong derivative gain comes to play, reacting to the short but steep increase in boost after the sec turbo gets on line
4. however, immediately after, WG% stabilizes at around 162 (63%), and does not react at all to the slow but continuous increase in boost; eventually the safety row is hit at about 19600 - even though the 0.90 Kg/cm2 set-point should be like 18900, I believe; so, while the WG% reacts quickly and effectively to rapid changes in boost, slow changes do not seem to cause any reaction in this phase, and the duty stays close to the set duty
5. it is only after time=128 sec that he WG% starts to slowly decreases, down to 128 (50%), but not quickly enough to stop the "boost creep", causing the PIM to exceed 20000.
6. but when the throttle is momentary lifted off at 128.5 sec then resumed, note how the WG% reacts quickly and effectively to counteract for the derivative of the boost changes

- Sandro

hate to burst your bubble, but I don't see any image. sometimes if you upload an attachment, then walk away from the computer for a while before submitting the post, the attachment and embedded image will disappear.

There must be a curse of some sort... because I am now in the thread and I can see the image of my post #48. I also tried and connected as a guest - without logging in - and still can see the image... Anyhow, I am re-uploading the image to this post and attach to it. Hope this works...