TPI V8 no start

[esp42089]

Hi all,

I've grafted the suspension and drivetrain from a 1991 Corvette into a 1953 chevy. I got an MS2 to run the engine after the qdr drivers in the GM ECM fried (my signal generator to over-ride the VATS fed too much voltage into the ECM when I got the alternator running ...whoops).

Here are the specs:

Small Block 350 Chevy with a TPI multi-port injection setup, running batch fire.
I've got a MS2, v3.57 from DIYAutotune, running 3.4.3 firmware.
Stock GM MAP sensor is wired in to pin5 and I'm trying to use it for MAP (later I'll use the MS2 as Baro for continuous correction)
Stock GM TPS, CLT, 4-wire IAC, and an Innovate LC-2 O2, all wired as expected from the manual.
GM HEI-7distributor with Bypass wired through the DB15 (pin10) for Spark B, Ref and EST come in on the DB37.
I'm hoping to use FIDLE to control the fan relay down the road.
Everything is wired up as suggested in the manual (as far as I can tell)

I stepped through the setting up guide, with all the signals looking good at each step. Injector and coil tests were good. I got to the timing section, and had trouble verifying the timing alone. I know it was set in such a way that the GM ecm fired it right up no problem. I really need a helper, but I'm kinda stuck until this weekend on that front. I decided to try starting it up and check timing once idling and warm. No dice. Cranks, but doesn't fire. Any help would be appreciated, even just looking over my settings to make sure they are ok for when I have help to go back to TDC on cylinder 1 and check where I'm at. Tomorrow I plan to pull a plug and verify I have spark while cranking and I'll hook up my fuel pressure gauge to make sure rail pressure is staying put during cranking too. I checked before and it was all good to go, but maybe something is happening to my fuel pressure during cranking.

Thanks
Evan

[esp42089]

I want to add my progress over the weekend:

The MS2 is activating the relays correctly, and I have fuel pressure. The first 2 second burst doesn't bring me up to full pressure, just ~25psi. Cranking quickly brings it the rest of the way up to 42psi. Should I add in a 0.5s priming pulse delay? NOID lights confirm that on cranking the injectors are firing.

I lost spark. Diagnosis points to a bad ignition coil. I've got a new coil, module, and rotor on order to hopefully rectify my lack of spark. The cap and pickup coil look good. I'll update again when I get these parts swapped out. I would still appreciate someone glancing at my tune and the logs to see if anything looks amiss.

Thanks
Evan

[esp42089]

Heres a picture of it:

20170402_142543.jpg

Engine bay, although the corvette computer is gone, and the MS2 is under the dash inside now:

WP_20170416_15_00_35_Pro.jpg

[ChevelleFan]

Most of it looks pretty good to me -- however:

*) Cranking RPM @ 650 is too high -- set it down to 300. (Startup/Idle --> Cranking/Startup Settings)

*) Fuel Settings --> Injector Dead-time/PWM -- you look to be configured for low-impedance injectors. Most TPI injectors are high-impedance. Use an ohm-meter and measure the impedance of your injectors. 2-4 ohms = low impedance, 12-16 = high. If yours are high, as I suspect, change:
PWM Current Limit %: 100
PWM Time Threshold(ms): 25.6

*) Your MSQ is configured for a Narrowband, not the LC-2 (Fuel Settings --> EGO Control). Once you change that, go to Tools menu --> Calibrate AFR Table, choose the LC-1/LC-2 default (or whatever is appropriate if you changed the settings on your LC-2) and click Write to Controller.

Try those changes first.

-Dave

[esp42089]

Hi Dave,

I appreciate the pointers. I had caught the narrowband mistake yesterday, but not the others.

I do have hi-z injectors. They all sit right around 14.6 Ohms and are the bosch style.

For my edification, where does the 25.6ms come from? Is my injector dead time of 0.9 a good starting point?

My cranking speed is only 158rpm. Does that seem too low? I also want to put a 0.5s delay on the priming pulse since the initial 2s run doesn't bring the rail all the way up to pressure. Is this the correct way to get the rail up to the full pressure?

Now the big big question:

I'm trying to use the GM MAP as the controller, not the onboard MAP. Have I set this up correctly/ is it possible? The readings off the MAP seem good, but I actually can't say for sure that it isn't just the onboard that I'm looking at.

Evan

[gjestico]

Which ADC (6 or7) did you wire in the external MAP signal to ? Got any pics of your built board ?

[esp42089]

No pictures. I bought it assembled from DIY and didn't bother to open it up when it arrived. The GM MAP signal goes in on pin5 and the 5V ref for it is tied in with the TPS and all the other sensors. Gnd is also tied to all the sensor gnds.

The invoice lists:

Special external MAP / baro correction mod. Lift pin #1
of MAP sensor and wire to JS5. Connect pin 1 of MAP
header to SPR3 to bring MAP sensor reading in on DB37
pin 5.

[ChevelleFan]

158 cranking RPM not too low. I think the mini-starter on my 383 cranks around 137rpm.

I don't remember where the 25.6ms comes from. My guess is it's legacy from MS1, which was 8-bit and it was the max-commanded time possible. But dont quote me on that. That's how mine are set.

Dead time of 1.0ms is the default. I'd use that instead of 0.9 unless you have measured yours or have some documentation that they are 0.9. Can always change it later, I'm just trying to help you get running.

MAP readings look ok since you're at ~ 5000ft.

I wouldn't worry too much about the delay on the priming pulse. The moment you turn the key, MS will see RPM, turn on the fuel pump and start injecting fuel. It'll build pressure quick enough. That can be fine-tuned later.

-Dave

[rickb794]

Try and figure out the dead time because changing that will affect the VE table.

[esp42089]

Alrighty. I will start with the good news:
it fired up IMMEDIATELY! barely cranked it before it fired and ran. Really strong fast start!
it died a second or so later and I restarted it, and I blipped the throttle some to try and keep it alive. Idle stabilized when I stopped. Maybe it wasn't needed, but it sounded like it wasn't going to hold. This is where the 2 lost syncs came in. I assume since it didn't happen again that I'm fine.

Logs look wild to me, but everything seems to be there and sensible looking. The sinusoidal running was very obvious but seemed stable and didn't sound horrible so I let it idle. I'm going to re-read the manual and formulate a plan. Any advice or help to fast track through any problems you see would be very appreciated!

I had to manually shorten the log. I didn't leave hardly any time before I cranked it up, and I cut a long tail off the end,about 10 minutes or so.I created a second file and put some of the tail end in it. I tried to get timing but it never quite stabilized, moving between 4 and 5 BTDC.

[esp42089]

Second run. I shortened this one too. I set the fixed timing at 6, since that is what I think my timing should be (still not able to get stable reading, it's somewhere around 5). I was heartened to see the thermostat open. Not so happy to see the temperature start running away and I shut it off. I had a couple box fans pointing at the radiator. I don't have the fan relay configured yet for the electric fan.

[ChevelleFan]

Nice work!

First thing you want to do is get the VE table squared away. If the car is drivable, get it up to temp (WUE off) and run VE Analyze Live while you drive around the block and let it dial in your VE table. If car isn't drivable, you can still just do run it in the driveway in Park/Neutral and let it try to sort about the lower RPM/lower Kpa bins a little bit. Something is better than nothing.

As for your timing, you're set for Fixed Advance, so you're only going to get 6* until you change it to Use Table.

-Dave

[rickb794]

And 6° is not enough spark timing to run the motor it is very late. Probably why the temp was climbing, exhaust was probably close to glowing...

I would set timing to 15° fixed for testing.

Your spark timing should not be moving around (with it set fixed). You need to get that straightened out before trying to tune or drive it.

Once you get the timing straight then switch it over to "use table".

Just to verify -
That is a locked distributor?
No centrifugal or vacuum advance?
And you have checked rotor phasing?

[esp42089]

Thanks for the emotional support and help guys. My heart lifted when it fired right up. I'm at 14 months since initial tear down and somewhere around 1500hrs into it at this point. It's been an emotional roller coaster with this car and I'm so excited to drive it.

It is a factory locked distributor. I set it to 6 deg fixed, thinking that was what I think my engine is set to (with #1 at TDC) and I wanted to check that while it was idling. I was thinking I would verify/ adjust the trigger wizard until I saw 6 deg on the crank wheel with my timing light. Is this not a correct way of thinking about / doing it? Should I set it to 15 fixed instead and then adjust the trigger wizard until I see 15 on the crank?

I do agree 6 is pretty darn late and it was probably running hotter than it should because of it. I was watching the CLT like a hawk, expecting the car to overheat sitting in the garage with ambients in the mid 90s and no engine fan.

This engine has 145,000 miles on it. I'm willing to bet, after thinking on it, that my 1 degree wander could be a loose/stretched out old timing chain or similar. This engine is the testbed for me to learn the MS2 with, as well as sort out all the mechanical issues that may crop up with this new build as I start driving it. Once it's all squared away, and I'm ready for more power, I'll pull the motor and rebuild it.

So my plan for tomorrow is to switch to "use Table" instead of fixed timing and bring it to temp, then I will turn on VE Analyze and drive it around a little while I datalog. I'll check out the logs and see what I get. If there is something interesting happening, I'll trim it out of the log (since there's a 1mb limit) and post it here with the tune for you guys to check out. Seem like a good way to go?

Another question: the idle was noticeably sinusoidal. Is that because of the fixed timing? Should I expect that to go away tomorrow with "use table"?

[rukavina]

Fix fan! At least hardwire on so you don't overheat for now. Set timing @15. Get it to temp. Check timing. Make sure rotor points at #1 @25°. That puts you in the operating range of 12 to 36 deg. Get a area on ve table and in and around where it's running set the values all the same and shoot for 12.5 afr or so. Lean out later. Open loop. That should get you somewhat stable. Then you can start to play. I read these post every night since I started. I read tuning ms2 tuner studio and gpio boards. Get good info. Search for posts how to shape timing table to run just under idle rpm and advance there to help stumble. Be conservative and creep up. Easier than premature motor build.

[ChevelleFan]

I agree -- fix the fan (hardwire if needed for now) and set the Fixed Timing at 15*, and use that for dialing in the Trigger Wizard. With 145k, if you're off a degree, could very well be sloppy timing chain. Once Trigger Wizard looks good, switch to Use Table.

I agree with rukavina's strategy to find the area on the VE table where it's idling and setting them all the same. It should probably be a 4-cell or 6-cell area. But 12.5 afr is too rich, no reason for that much fuel at idle. 13.5 or higher. It's a stock engine. Open loop at idle is a good suggestion for now.

the idle was noticeably sinusoidal. Is that because of the fixed timing? Should I expect that to go away tomorrow with "use table"? -- no, it's more likely your VE table. But set all the timing values in the idle area the same like the VE table.

How close is the car to being drivable? Grafting that Vette drivetrain to the '53 had to be a lot of work.

-Dave

[rickb794]

You dont twist the distributor to set the timing,

You twist the distributor to get the rotor pointed directly at the #1 post at 25°BTDC.

This is to mechanically align the rotor with the post inside your advance curve 10° to 40°.

Timing is set within Tuner Studio When we say fixed timing we are talking about the upper right corner of the ignition options pane.

You set it to "fixed" "15°" and then adjust the tooth #1 angle number to make a standard timing light show 15°.

Now what the MS thinks is 15° is 15°, Now you can switch it back from "fixed" to "use table".

This is all covered in the manual.

[esp42089]

I started it up, everything looked good, and so I changed it to fixed timing. I am at 6 BTDC with the timing light,so I changed the angle offset to 6, turned back to use table, and drove it around a little. Nice and gentle. I logged during and I also ran the VE Tune. Things seem pretty promising! Engine seems to be running well. One thing that was throwing me for a bit of a loop is that the AFR would shoot lean when I came off the throttle and coasted. Is this normal behavior?

[rickb794]

Decell fuel cut?

[esp42089]

decell is off. I've got it pretty well ironed out with some tuning, but now I'm having a high idle issue. I started a new thread for that here:

http://www.msextra.com/forums/viewtopic ... 01&t=66512