Is anyone here doing boost control using the MAP sensor signal as shown in the Sloppy video or know of someone who can help figure out an issue with it? I set it up on a buddy's car using the PID settings in Sloppy's video as a starting point but not having success. No matter how I change the PID and Target Rate Limiter settings, it seems to make no difference, making me suspect it might be a boost control setup issue.

I have seen a couple of videos on YouTube discussing this but do not know how to contact any of those folks directly. So, I'd like to hear from anyone using this successfully or know of someone/tuner they can refer me to or how to get in direct contact with Matt Happel so he can refer me to someone.

Thanks.

 

youre very vague.
i set mine up similar to that video and it showed promise and was better than trying to use the dual solenoid setup. happy to answer more specific questions.
tuning pid loops is not easy.

 

I set mine up using the Sloppy way, using a 3 port mac valve, map as dome. Works pretty good, and I dont know what a pid is, lol.

 

You can also just run a PWM output and have Terminator control it just like an AEM or other type of basic electronic boost controller does. I use the PWM output table with a $23 amazon mac solenoid, works great as you can do PWM vs RPM, MPH, etc. Freq setting in the 30-35hz range gets ya going.

 

Is this a new build? Are you able to control boost with only a reference line to the bottom port of the wastegate?

 

Thanks for the comments, guys.

The build is not new. It's the typical LS fox body with a 370, FI billet 80mm T4, TH400 and has a best of 8.26 @ 165 with a 1.20. This was done with a MS3X, FIC 127s maxed out with E70 and a little alky injection as a crutch. Upgraded to Bosch 210s so he could run E85 and ditch the alky injection crutch but had nothing but fueling issues. I suspected the MS3 injector driver not handling the Bosch 210s well as the injectors would basically quit working as they approached 50% DC. Not the only reporting similar issues but never able to properly test and prove it, thus the switch to Terminator X.

As I said, I used the PID settings from the video as a starting point. No matter what launch RPM or launch boost I try. It will only spool to 6psi (WG spring pressure) on the t-brake. On the launch, it does not follow the ramp defined in the Boost/Time table. The max target boost in the table is 11.5psi and just trying to see what that actually gets me as it has never been stated if the resulting boost would be table max + WG spring pressure. Matt never stated that but he got 24psi from a 19psi max target with a 5psi WG spring. With that said, I assumed I would see somewhere around 17+psi max. What I got was a steep ramp, overshoot to 19.2psi and consistent oscillation that does not decay throughout the pass. Oscillation will go from about 16+ to 11.5. Needless to say, the timing and fueling follows the oscillation and it sounds and runs not so good.

Because the PID values are so small, I tried adjusting the P down and I up in small increments with no difference. Only difference I saw was when I drastically reduced the Traget Rate Limiter and that dropped the initial overshoot down to 18.2psi. I could continue fiddling with it but I've never been able to fully wrap my head arouns PID loop tuning but hoped that the results I see from PID setting changes would help my understanding. I cannot discount the fact that something else might be wrong in the GCF, so what I am actually looking for is someone who is familiar with Terminator X and this boost strategy to go through the GCF along with the datalog, to see if there is another setting that might be causing issues. I know there are other methods that I could try, but I'm determined to make this one work.

BTW, this is not my car.

 

As I stated, because the PID values are so low for this strategy, I have been making adjustments in small increments. However, here is something from the Holley manual:

"- increase P to the point of oscillation and then immediately divide it is half. This P Term (half the value when oscillation ensues) represents the maximum recommended P Term to insure an adequate stability margin."

That would make the P = .75, which didn't look right to me (DUH, what do I know), so next time time out I'll try that value but cautious me will get there in 2 steps.

 

a log would help.
you can check the target boost, bcs dc, boost error to make sure its all doing what it thinks it is doing. the duty cycle should be going up as long as the dome pres is below target boost.

 

the pcm is going to target the pressure source you gave it, so if you tell it 11psi it will try to make the manifold have 11psi.
if its oscillating like you say, i would try turning the P down a little.
P is proportional so it is an adjustment factor based directly on how far away from the target you are. I is integral so its an error over time.

Going off memory, if your actual boost curve overlaid on the BCS DC and the ups and downs mostly line up, adjust your P. if they are inverted like two sine waves 180 out, try adjusting the I

 

OK gurus, not able test the car mentioned above but I was able to do a little testing with mine. Both engines are 370ci but the mustang has a ProFlo intake and I have a LS1 intake. The Mustang has a FI billet S480 T4 with the biggest exh wheel/housing available and I have a VSR billet 7875 with the 1.25 housing. I figure whatever I come up with could be applicable or be a good starting point for the mustang.

I'm testing from a roll in 2nd going full throttle at about 4000 with white knuckles, (LOL, for those who know me) spinning the 235 DRs and wanting to drift to the centerline. This is the 3rd test hit. Started out with the Sloppy PID values of 1.5, 2.5, .5 and a Target Rate Limiter of 28. First hit had a massive overshoot and pretty bad oscillation. This 3rd pass the PID values were .75. 3.0, .5 and a Target Rate Limiter of 8! Oscillation is under control and some fine tuning will clean it up but the overshoot has only a slight reduction. Once I floor it, it goes from 0psi to 15psi in 1.04 secs, recovers fairly quickly and slowly comes down to about 10psi. That is with a target of 8psi and I don't know if target 8 should get me 8 or 8 + WG spring (7psi) which is 15psi, exactly what it overshoots to. The overshoot is what I would like to fix. I know it will probably not be fixed 100% as the turbo is on the small side for the engine combo.

Anyway, I dodged a bullet today. After the pull, I slowed down to make a u-turn and the engine stumbled but recovered once I got going. Started to smell fuel and noticed that the fuel level was sitting at E. Wasn't far from home and when I shut it down and got out, there was a massive puddle under the left front. Popped the hood and there were puddles in every crevice on the drivers-side of the intake. Had the wife turn the key on while I checked for the leak and luckily I was standing at the LF corner because the top o-ring in #5 must be completely gone because fuel shot out of there and would have gone in my eyes. How the hell the car didn't catch on fire is beyond me. Tank was just about empty so had to have sprayed at least 5+ gals in the engine bay! The MAN above was looking out for me.

Anyway, here are pics of the datalog. The one with just 1 panel is more focused on boost.

 

weird, the same thing in the same spot just happen to my buddies car

 

I was trying to fix an over shoot on my buddie's car this weekend. Dropping the P below 6 caused some serious oscillation with the boost curve. Ended around 6, 30, 0 and 100 target rate. Still had the overshoot at the start of the ramp but boost was fairly stable after that.

 

Never heard that one before but mine are obviously inverted. Are we looking at increasing or decreasing the I term? I will keep decreasing the Target Rate Limiter to address the overshoot but going from 28 to 8 seemed to have little effect. I've not tried adjusting the D term yet, do you think think that might help with the overshoot.

So, I'll this question again, more explicitly. Does the boost that you guys run match the target in your table or is it target+WG or something in between?

I'd love to see someone come up with a cheat sheet for PID tuning. Not what the actual values should be, but a list of actions based on results, like what you posted above. If you see this, try adjusting that up or down.

 

I always assumed it should try to target the boost in the table, but its looking like its not able to accomplish that from your graphs. target of 8 but it sure looks like its aiming for 10. mine actually acts similar but i keep fragging transmissions so i havent been able to really think too hard on boost control more than a little here and there.

i just made my own DBW throttle body controller from scratch so i had to tune the PID parameters and ended up going down quite the rabbit hole. This was a few months ago so i hope its still accurate info in my head, but the general rules of PID (not specific to this application) is stuff that is slow changing like an oven or a fluid heater will usually have a P term that is numerically higher than the I term. in applications where everything moves a lot faster (like a DBW tb) the I term ends up larger than the P term. in my case i have nested loops similar to how the dominator-only setting the holley can do to control dome using manifold pressure as a target. on my DBW TB my I terms were about double my P terms. I also have had the best luck zeroing out the D term.
now that ive gone completely off topic.... based on my ramblings and completely uneducated guess, i think you should try a little more P and roughly double that for I and zero the D. You can see as the boost is coming up it starts pulling bcdc but not fast enough. its not overshooting after the first spike so i think you can raise up the P to get the bcdc out of there quicker to help catch the overshoot in time. the higher I term should tighten up the delta between the target and actual after the initial spike quicker.

in case there was much doubt,

 

Man, I certainly understand what you're trying to say, but sometimes there are contradictions that make it all so confusing.

Here is an excerpt from the Holley manual: "- increase P to the point of oscillation and then immediately divide it in half. This P Term (half the value when oscillation ensues) represents the maximum recommended P Term to insure an adequate stability margin."

That got my attention when I was dealing with oscillation and it certainly addressed it, but jayyyw also said dropping the P below a certain number caused serious oscillation for him. The issue I was most focused on was the overshoot and nothing I've tried has any effect. I've always suspected that part of the problem was the 7875 turbine wheel being too small as the boost is instantaneous and on the verge of being too violent. So, I spoke to Vren Friday to see if there was a larger turbine wheel option for the 7875 and before I could finish, he said the 7875 is way too small for a built 370. He strongly recommended a T6 but that is not an option I want to deal with. He has a T4 billet 80mm with a 96mm trubine wheel and 1.25a/r housing that is not the optimum solution but should greatly reduce the back pressure and help address the overshoot. Would be going from a 83/75 wheel to a 96/88! He actually spent the time to bust out his caliper and took some measurements of both turbos and said it looks like my existing plumbing should line up without any mods. MAJOR KUDOS to Vren for that. That's the kind of customer service that keeps customers coming back. Selling the almost new 7875 should help cover most of the cost, so gonna make that turbo switch to avoid chasing my tail (and scaring myself on the road) before I go back to tuning this thing.

 

just like any tuning, hardware can drive the tune as well as software. it sounds like youre going down the right path of fixing some hardware issues that appeared to be causing irregular pid feedback/outcome.
good luck and let us know how things work out.

 

I've given up on the MAP sensor signal strategy. The Mustang is going back to on-board air with dual 3-port MAC valves, dome sensor and conventional closed-loop.

For the Volvo, I finished up the VSR 80mm install and what a difference in the power delivery! 2nd gear pull and the boost comes in much smoother even though it is definitely pulling harder. Funny thing is, at 10psi I probably haven't even gotten the turbo's attention yet.

 

Nice! I agree, the 80/96 is a much better fit for a 370 vs the 7875, if you went T6 his 102 turbine would have been even better.

 

I've wondered why people splice the map sensor rather than just putting a "dome" sensor in the intake? Wouldn't that accomplish the same thing without splicing into the map?

 

that requires an additional sensor and full wiring for it. i spliced mine at the ecu so it was one pin and a few inches of wire and done. your way would work also.