MegaSquirt II A/C Idle Up Modification

[jangell]

If anyone else has an MSII and wants to add A/C idle up, I documented what I did here:

http://www.tmproductions.com/repairs...p-modification

Short(er) version: I did both A/C sense (so that MS can adjust the idle) and A/C control (so MS can turn the A/C on and off). I built the standard "going high" circuit from the hardware manual, but I found the relay output circuit confusing (I still don't know what the diode to 12v is for), so I just replaced the whole thing with a P-channel MOSFET. I also used a 5v relay board instead of an automotive relay to control the A/C compressor clutch. This required fewer parts, which was nice. To connect to the MegaSquirt I used the DB15 connector, which is pre-wired for 5v and ground out. I soldered header pins to the pads in MegaSquirt and use breadboard jumpers to connect everything together. They fit pretty tightly, so I'm not worried about anything popping off while driving. I used JS5 for input and JS11 for output.

Wiring in the car is just cutting the K/O wire under the passenger side dash, running the mode switch side to the common terminal of the relay board in and the DB15 to the MegaSquirt, and the other end to the normally open terminal of the relay board. I ran a 6' DB15 extension cable from the MegaSquirt behind the driver's seat to passenger side of the center stack, and plugged my DB15 connector into it there. I also used my new 3D printer to make a case that holds the relay and the MOSFET.

And that's it -- I turned on A/C idle up in TunerStudio, everything works just like it should.

I'm tempted to add another relay to turn off the A/C on WOT. Unfortunately, you can't just use the Programable I/O features of TunerStudio with the existing relay, because A/C idle up is doing that, and it will complain loudly if you try -- you need to use another pin, which means another relay. I also considered a "panic open IAC" button to kludge around low idle/near stall situations by forcing the IAC open by another 30 steps, but that's just silly and I should just fix the stalling issues instead.

-- Joe

[Spittybug]

Nice. I always considered doing this until I realized that the AC in Houston is ALWAYS on! Now the car is gone I can be a pure academic on the subject....

What's going with your stalling at low idle? I ran mine at ~850 IIRC and had no issues at idle, even with AC coming on or off. I'm sure you've done the obvious and bumped up the timing and fueling in your lowest RPM bins. Are you running the stock throttle body? I know I used to have an issue with my LT1 throttle body being too big and dropping incoming air velocity too low. This would bog me down on acceleration. The solution was a simple blanking plate to reduce the opening size.

[jangell]

I have another thread talking about my low idle issues. A lot of my settings had gotten a bit off target due to trying to compensate for the stalling, such as increasing the idle to 900 (instead of 790) to avoid stalling issues, and screwing with the ignition table to try to keep the idle up.

FABombJoy has been trying to help me figure it out. He very quickly found some problems in my settings due to all that tinkering. I've dropped the idle back down to 790 and I haven't stalled yet, but I have dropped to the low 500s at times, which is right at the edge of stalling. My idle is seeking a little now, but I turned off closed loop idle advance while doing these tests, so I wonder if turning that back on will help, and fix the current RPM drops.

A probably unrelated issue I have is that on some hot starts the car will start, then immediately stall. It always starts up the second time. The RPMs don't fly up above 1200 or whatever it does on a normal start, but rather it gets up to 700-1000, then stalls out. Starting it a second time works fine. I usually go from "off" to "start" without pausing for fuel pump priming in "run", but I don't know if that's relevant, and honestly it shouldn't matter.

A lot of my parts are from Josh when he updated his 3.0L to the LS engine. That includes his a throttle body he got off of a Mustang, although I don't think he ever ran with this -- a box of parts got lost in the mail and he went out and sourced me all new parts (luckily most of the fabricated parts arrived in the box that actually made it here, and a year later the lost box showed up). It is definitely larger than stock, at just under 2.5" inner diameter. I had been wondering if the extra air was affecting the idle, but I don't really know.

-- Joe

[Spittybug]

Go online and use a throttle body size calculator for your engine. I was going crazy playing with the software..... A small decrease on the diameter by adding a thin metal restrictor (aluminum flashing) MATERIALLY improved things. Whenever I opened the throttle from idle it would bog; the VE would climb fast but the RPMs went down. Too much incoming air at too low a velocity. The PRV seems to be very susceptible to this. Now, it could be because I wasn't using the stock intake with its longer runners, but I'd check one of those calculators if I were you.

[FABombjoy]

I don't think velocity was the issue. Putting a 300HP engine throttle body (208.2 CFM) on a 130HP (90.22 CFM) engine and tuning based on MAP caused loss of finite torque control.

An accel enrichment strategy could tune around an LT1 throttle but the end result would always be a very nonlinear throttle, and with standard injectors you'd likely need x-tau tuned to make it work under all conditions. Converting to MAF would probably have made it easier.

jangell's log shows stepper is 100+ at idle, hinting at insufficient base opening combined with setting 0 steps on shifts. The stepper valve likely can't react fast enough. The low RPM zones are too lean as you can watch the load target circle around them and only stabilize at higher loads/RPMs.

[jangell]

I don't think velocity was the issue. Putting a 300HP engine throttle body (208.2 CFM) on a 130HP (90.22 CFM) engine and tuning based on MAP caused loss of finite torque control.

An accel enrichment strategy could tune around an LT1 throttle but the end result would always be a very nonlinear throttle, and with standard injectors you'd likely need x-tau tuned to make it work under all conditions. Converting to MAF would probably have made it easier.

I'm guessing you mean for Owen here.

I did just find a calculator, and it suggests that 2.02" inches is the ideal size for 182 CID @ 5000 RPM; I'm a little under 2.5" with the Mustang one. Not sure how much that half inch matters. I do have the original throttle body, but I'm not really interested in spending the time to change everything over to it at this point if I don't have to. I'd just put in a one of Owen's baffles.

jangell's log shows stepper is 100+ at idle, hinting at insufficient base opening combined with setting 0 steps on shifts. The stepper valve likely can't react fast enough. The low RPM zones are too lean as you can watch the load target circle around them and only stabilize at higher loads/RPMs.

Cool... Not sure what that means.

Base opening (throttle plate position) is very likely wrong, since the plates are completely closed at idle at the moment. "0 steps on shifts" has me lost, though. And are you suggesting I need more fuel in the low RPM cells?

Thanks

-- Joe

[Spittybug]

2.02" TB vs. 2.5" = 3.21" area vs. 4.91" area or oversized by 52%. HUGE. And I assume that's a single barrel TB, not dual, right?

I don't dismiss any of FABomjoy's other findings, but this has material impact on things as I have first hand knowledge of. You want to stick with MAP if you can. This allows it.

My TB was damned near 100% closed at warm idle with my IAC stepper valve open something like 10 - 15 steps. I was using a GM pintle style that goes right into the LT1 TB.
What kind of VE are you getting at warm idle? I would hope around 30kPa.

[FABombjoy]

I'm guessing you mean for Owen here.

Yes, not advocating you switch to MAF

A too large throttle body would emerge in a average plot of TPS vs MAP. If your logs show that a minor change in TPS results in a huge shift upward in MAP then you'd likely have too large a throttle body.

Base opening (throttle plate position) is very likely wrong, since the plates are completely closed at idle at the moment. "0 steps on shifts" has me lost, though. And are you suggesting I need more fuel in the low RPM cells?

Your engine is telling you that. Increases in step position at low load/RPM causes a decrease in RPM. Lean.
The close on shift (0 steps) is all in the idle setup menus.

I think you need to create a view in MLV that shows you all of these players regarding idle control
RPM
MAP
TPS (even if it's broken)
Stepper idle position
Spark: Idle advance correction
AFR
CLT

Then load your recent log into MLV along with your tune and watch the conditions when pointer drops below 700. You can see AFRs jump to 15/16:1 which will decrease RPM.

If the fuel isn't right for a given RPM and load, the engine will move elsewhere. If you see RPMs "orbiting" a zone, the settings in that zone aren't correct.

[jangell]

(replying to Owen here)

I should have done the area math. My friend and I had pizza once at a place called Galaxy 500, where they had a 500 sq in pizza. By diameter it wasn't much bigger than the normal large, but while we were waiting we did the math and realized just how much bigger that was. We took the vast majority of it home.

Tell me more about this baffle you built.

I am using MAP, mostly because my TPS binds after around 5 turns of the throttle for no reason. Taking the TPS off un-binds it, and adjusting the tightness of the screws or adding an extra gasket doesn't help. For now it's disconnected. I'm taking of building one that connects to the gas pedal just so I can have something for logging purposes.

My TB is currently 100% closed, relying entirely on the IAC to provide air for idle. I'm willing to open it up a bit and bring the IAC down to 0 steps. It sounds like the IAC I'm using is similar to yours -- it mounts in the throttle body via a housing that is bolted directly to it. In fact, I took some pictures when I did my adjustments to it:

Open: https://images.squarespace-cdn.com/c...g?format=1000w

Closed: https://images.squarespace-cdn.com/c...g?format=1000w

The page where I document getting the IAC settings right: http://www.tmproductions.com/repairs...of-small-fixes

I did break off one of the screws that holds the IAC housing to the the throttle body, but I'm skeptical that this missing screw is causing any kind of vacuum leak. I mean, it's on there really tight, and the other three screws are not far apart. Someday I'll get around to replacing it...

For the MAP at warm idle, I want to say I'm in the 35-40 kPa range with the A/C off, but I'll have to go for a drive to be sure.

-- Joe

[jangell]

Yes, not advocating you switch to MAF

A too large throttle body would emerge in a average plot of TPS vs MAP. If your logs show that a minor change in TPS results in a huge shift upward in MAP then you'd likely have too large a throttle body.

So I should get some kind of TPS working, even if it's just for logging.

Your engine is telling you that.

It's telling me I have too big of a throttle body, you mean?

Increases in step position at low load/RPM causes a decrease in RPM. Lean.
The close on shift (0 steps) is all in the idle setup menus.

Ah, I see the "Shift Settings" now in the Closed Loop Idle panel; I kept looking at the Idle Control panel. I guess I wasn't sure how it knew there was a shift, and assumed it needed some extra input for that. Does it just detect load changes and assume that a sudden change means that a shift has occurred? I'd been leaving it off, mostly because I didn't understand it.

I think you need to create a view in MLV that shows you all of these players regarding idle control
RPM
MAP
TPS (even if it's broken)
Stepper idle position
Spark: Idle advance correction
AFR
CLT

Then load your recent log into MLV along with your tune and watch the conditions when pointer drops below 700. You can see AFRs jump to 15/16:1 which will decrease RPM.

If the fuel isn't right for a given RPM and load, the engine will move elsewhere. If you see RPMs "orbiting" a zone, the settings in that zone aren't correct.

With regards to the TPS, it's actually disabled -- I unplugged it and jumped it to ground so MS won't try to do anything with the input. It getting stuck at odd values could cause idle not to kick in or it to think I want it to flood clear or something, and they spread the TPS-dependant settings around enough that I wasn't sure I'd be able to find them all and disable them properly.

I checked my last log and I see around record 6607 that it dropped to 640 RPM, and the AFR went up to 15.7 soon after. I see how the IAC is opening to try to recover from the low idle as well. Based on what you're saying above, I need more fuel in the ~650 cells to compensate for that? Or are there other settings that may need adjustment?

I'll get the car warmed up later and do my R-N-D cycle, which is usually the easiest way to get the RPMs to drop to low and will give me more samples to compare against. Presumably I can also watch the AFR spike live and adjust the fuel in real time, then shift from between R/D and N to see how it affects the AFR, just like tuning any other cell.

I guess the problem I always had with the low cells is that the only time I would get into that position was by mistake, not intentionally, although the R-N-D cycle seems to make it happen reliably enough that I might be able to adjust the fuel properly.

Thanks!

-- Joe