Landon303

iTrader: (25)

Hey guys! First build venturing out of the stock computer controll. Going w/ an MS2 V3. Anyways- I'm using an FD UIM on a T2 LIM. Got the manifold free, however, sans BAC. I like, do not want to spend $100 on an FD BAC when I've got the know how to make an S4 BAC work. Here's what I've got.



1" thick aluminum stock, which will have one port at the "feed" coming from the throttle body. Port turns down 90 & enters a 1/4" NPT barb fitting, w/ a 3/8" ID.

From here, I go to the feed pipe on the S4 BAC, and out to the 2nd port on my adapter block. Holes are offset but overlap enough I should have good, non restricted flow through them. From here it enters back into the manifold and carries on.

So the mechanical design is done, but before I go & spend the couple hours making it, I wanted to know about duty cycle limits & dwell time limits on the BAC.

There's a reduction in area of port size. 3/8" is 9.5mm diameter. That'll be my smallest opening. The stock feed pipe to the BAC is approx 14mm in diameter. So, a few mm's of port area lost.

Question is- will this cause me to have to hold the BAC open for a longer period of time, or will there be any other repercussions from the smaller size'd hose?

Thanks!!

Aaron Cake

iTrader: (1)

FD BACs must be pretty cheap on the forum or eBay...

Anyway, you're over thinking it. You don't tell the MegaSquirt how much to hold the BAC open. It figures this out itself based on a PID algorithm. You will have to tell it the idle open duty (normally I set this at 100%) and then idle closed DC (normally I set this at zero...or occasionally if the BAC is being a bit sluggish, around 10%). Then you are tuning the PID directly and the MS figures out duty cycle.

Why V3? It doesn't have PWM idle built in. Run 3.57 which has that "mod" already.

I doubt you'll see any noticeable effect with your adapter restriction. You may not be able to BAC the car up to 5,000 RPM, but why would you want to?

arran

My experience with a stock FC S5 BAC setup with an MS3/MS3X is that the BAC needs to operate between 65% and 100% duty and 204Hz

Lower the frequency and the valve becomes very audible, the next lowest frequency of around 158Hz iirc is quite audible and it just gets worse from there.

The valve does not start to affect the engine RPM until around 65%, and maxes out at about 3000rpm at 100% duty.

I needed to investigate this in some detail on my car because despite spending substantial time on tuning the closed loop idle it continued to fall through the idle setpoint. There is a small period of time as the rpmdot maxes out on engine decel with the clutch in for the PID to ramp up to over 65% to start affecting RPM.

Aaron Cake

iTrader: (1)

I'm going to guess you don't have a diode on your BAC. That's exactly the behavior I'd expect if there isn't a flyback diode installed.

I forget what setting is, but there is a setting called something like "last value" which will cause the PID to start the valve off at opening it with the last value known to hit the RPM target. It will save that time it takes to ramp up to whatever duty cycle your idle is regulated at.

But, first add that missing diode. You'll see duty drop to around 30% or so.

arran

I have no flyback diode installed. Which is irritating. I have just gone back and had a look at the wiring diagram pages at MSExtra here:
Megasquirt-3 MS3 Hardware Manual

In the two wire idle valve section it clearly shows an idle valve with diode and text in the notes discussing it (which is still a little confusing to be honest, what constitutes high/medium/low freq?). So I wonder how the hell could I have missed that? Curious I checked out the Web archive site:
Megasquirt-3 MS3 Hardware Manual

It looks like around April 2013 the Note appeared and in June 2013 the diagram for 2 wire was up-dated with the diode. I assembled my MS3 around January 2013 so none of those directions about flyback were present at that time.

Thanks for the tip on last value. Hijacking thread somewhat, sorry about that, I have idle control rock solid now (pre-diode!) but since I commissioned the Air Con a couple of months ago i can't stop it stalling almost every time ith AC on, under the same conditions mentioned in my first post in this thread, max negative rpmdot with compressor on and clutch in.

Arran

muythaibxr

There is actually a flyback circuit installed in the ms3x but it's only sufficient to protect the circuit if the BAC is the only thing it's driving. You don't HAVE to put a diode on there unless it bothers you that you need to operate it at an odd duty range to make it work.

arran

I just installed a flyback, 1n4004 power diode.

Not yet driven the car, but started and warmed it up. Upon starting it immediately idled at 3000rpm and stayed there. If I blatted the throttle it would settle down steadily over 5 secs. Turning the AC on resulted in a 3000rpm idle again but there was no slowing it down with a throttle blat.

All sorts of hunting and strange behavior resulted from soldering in the diode.

I pulled duty out of the cranking duty table (more than 4 bins would be an improvement) down from the 70% range to 40%. Also being able to watch the idleVE fueling in the MLV tuning console would be great.

I pulled duty out of the Closed loop idle initial value table (which I am using rather than "last value" which was mentioned above - please advise if this is a bad idea!), from 70% range to 20-40% range.

When the car is idling now the actual minimum duty will go below the value of "Idle valve closed duty %" which was at 68 and is now at 30, I saw it as low as 1% so the code seems to be behaving differently. Up until now all my data logs had the observed min duty the same as "idle valve closed duty" = 68%

I also bumped up the fueling in the idle VE table lowest rpm column (800) to be richer than the next column (950) to try to stop the lean condition as the engine goes into negative rpmdot with AC on (which I also talk about here: Megasquirt MSEXTRA / MS3EFI • Idle, Air Con and AFR wierdness (View topic)).

Fingers crossed that now the ecu seems to have better control of the BAC, and I've added fuel, I have resolved the stall problem with AC on.

Aaron Cake

iTrader: (1)

Now that the diode installed, the PID settings will have to be readjusted. Like minimum duty, etc. The BAC now has a much broader range.

It also sounds like your VE table is a bit lumpy. Areas of the table around BAC control need to be very smooth or by themselves, they will cause oscillation.

Are you not running the AC compressor control through the MegaSquirt? If not, I would highly recommend doing so. That way the MegaSquirt can idle up and enable the fan(s), then engage the compressor. Otherwise you have a very large unanticipated load that the MegaSquirt must instantly correct for with the BAC and there will always be some dropping.

arran

Thanks for the reply,

The duty settings have been adjusted in the initial value table, the cranking duty, the closed loop idle settings and EXCEPT for when I turn on the AC everything is great.

The Megasquirt is connected to the car's AC system via the Launch in digital input so it knows when the AC is active. I've added an Indicator Bulb to TS that goes green/red when the AC cycles on/off. I have not connected up the Megasquirt output that can switch the AC on or off but I don't think that is relevant here. It gets to see when the AC is on/off which is the main thing.

I went for a drive last night and it stalled on decel every time with the AC active. I disabled idle VE and auto tuned the main fuel table and thought I had it sorted, I was without stall with AC on for a couple of dozen 3k decel clutch in tests, but as I got back home and turned into the drive way it did it again. I have attached the MSL and MSQ to the link to MSEXTRA I posted above,

Thanks for taking the time to read!,

Arran

Aaron Cake

iTrader: (1)

Whoa! That VE table is a mess.

It's not surprising you are seeing stalling. Bins 30 - 44 @ 500RPM are way too low. Same with many of the 30 bins across the bottom of the table. The bins at the bottom should be very lean but don't have to be in the 18s because you're running overrun fuel cut most likely in those bins anyway.

AC idle up will help a lot but in order to really see the benefit, you should activate the compressor with the MS. That way, the MS can idle up and then engage the compressor. Otherwise the compressor clicks on (that a 10HP load on the engine...very significant) just as the MS is raising the idle. Not to mention the fan. That's 3 transients that PID has to deal with and I'm sure it sends the ECU into one of those bad bins on the VE table.

Your timing seems a bit low at idle. Bump it to 10 degrees then bump the bins to the left and below the idle to about 13 degrees. That will cause a self-stabilization. You can also run a lot more timing in cruise unless you are seeing hot EGTs. Should be in the 35 range.

AFR table is also way too rich in most areas. You can run 14.7 at all light load up to about 85 KPA and above 4000 RPM (make the transition to the richer ratios smooth). Then 15.5 or so in cruise as long as EGTs agree.

I would recommend sticking with simple pump based AE until you have a workable VE table, then play with EAE when the care is running correctly. Yeah, you'll have to revisit the VE table after you do, but it will make the initial process way easier.

arran

A mess, that's a bit harsh !

A bit to think about there

Aaron Cake

iTrader: (1)

That VE table needs to be cleaned up a lot before things like closed loop idle and A/C are considered though. It's not a mess per say, but very lumpy which is going to cause all kinds of issues during transients.