My 87 toyota supra hardware build

All home-built FreeEMS implementations without a forum of their own, usually TA-based.
via
QFP80 - Contributor
Posts: 57
Joined: Mon Dec 02, 2013 5:57 pm

Re: My 87 toyota supra hardware build

Post by via »

Stable sync earlier, not anymore. I don't know how this happens. It looked perfect earlier, then I go actually try to start and its just as bad. Ignition was connected earlier too. It loses sync about 10% of the time while cranking, then almost always immediately loses it again the moment it starts to catch. The longest I've been able to run is about 2 seconds.

I have just noticed that emstudio sometimes claims it lost contact with the board after some intense cranking. The board reset? I assume that should not be happening at all. Maybe I should try some giant filter caps on the power to the lm1815 board or the main power supply. I might hook up an oscope to the power supply and make sure its not crazy noisy. I might be able to do that tonight.

The timing was only a few degrees off, btw. I had forgotten to update the code, and when I reloaded the new code it went back to what I had before.
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: My 87 toyota supra hardware build

Post by Fred »

EMStudio losing contact can happen in several ways, which are you experiencing? USB issues are a common cause in my experience. CPU resets are pretty rare, didn't see any in your logs. OLV detects them, so you can watch that channel to see or just look at clocks manually.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
TonyS
LQFP112 - Up with the play
Posts: 192
Joined: Mon Jun 21, 2010 4:18 pm

Re: My 87 toyota supra hardware build

Post by TonyS »

Hi Via,

Do you have a schematic of the Zeal Engineering Dual VR board?
I took a quick look at the LM1815 and it appears to have a few configuration options and outputs and it might help to know the Zeal board implementation.

Have you taken a scope to the outputs of the Zeal board to look for anything odd (seeing anything other than nice 0 - 5V square waves)?

Do you have access to a logic analyzer to try and capture the timing of the Zeal board outputs?

Thanks,
TonyS
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: My 87 toyota supra hardware build

Post by Fred »

TonyS wrote:Do you have access to a logic analyzer to try and capture the timing of the Zeal board outputs?
+1 you can prove what the issue is in hard terms with this. Such a feature should be built in, but isn't yet. Need to work on that in future.

+1 on schematics, too.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
via
QFP80 - Contributor
Posts: 57
Joined: Mon Dec 02, 2013 5:57 pm

Re: My 87 toyota supra hardware build

Post by via »

I unfortunately don't own a logic analyzer. I'll try messing with an fpga to see if I can make one. I can also try to figure out the schematic, but it'll be from looking at traces.

Here's an example log that includes an odd emstudio reset at the end. Its the very last reset that happens -- the rest are probably me turning the key off.
Attachments
2014.01.16-17.01.53.bin.bz2
(1.07 MiB) Downloaded 854 times
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: My 87 toyota supra hardware build

Post by Fred »

OLV field is called "OLV Temp Resets" and 3 of the indicated glitches are comms resets (-1 to -1.254) and the fourth (third occurance) with value -0.987 is a reset and should be marked value 1 but I can't see the issue in the OLV code at a glance. Sorry for the eroneous result. :-/
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
via
QFP80 - Contributor
Posts: 57
Joined: Mon Dec 02, 2013 5:57 pm

Re: My 87 toyota supra hardware build

Post by via »

I hooked up the max board.... and it runs!
It missed a bit at first, which in the first log I showed appears due to code 5. Once it got warmed up a tiny bit and revving higher it was nearly perfect. I had to cut fuel a bit to make it smooth, but its definetely running now. The only main problem is that my idle is way way too high. Its close to 3500 rpms. During this, freeems reports idling aroudn 2000-2200 rpms, so I assume there is a configuration parameter I have missed somewhere, perhaps number of cylinders? I'll look around. Anyway, I know I can control the idle to a degree with timing and fuel control, but I would love any advice you have. Second attached log is a relatively long run. My timing light is failing me again, so I'll have to try some things there. I probably need to make my plug gaps bigger.

I'm probably not going to run it again until tomorrow -- I want to change the oil after all that dumping of fuel, especially if I'm idling at 3500 rpms.

Needless to say I am excited, and looking forward to tuning.

Addendum:
-I got timing, it is right about where it should be, around 10 degrees btdc. I'll note it kinda flashes around a bit, maybe at most 3-4 degrees... pretty sure this happened before even with the stock ecu -- is it enough to be worried about?
- I drove it onto ramps in prep for oil change.
- I realized num cylinders is irrelevent... so should I assume something is funky with the igniter's tach output?
Attachments
2014.01.17-19.38.46.bin.bz2
(1.19 MiB) Downloaded 868 times
2014.01.17-19.35.14.bin.bz2
(167.2 KiB) Downloaded 857 times
via
QFP80 - Contributor
Posts: 57
Joined: Mon Dec 02, 2013 5:57 pm

Re: My 87 toyota supra hardware build

Post by via »

I've made a lot of progress. I have it free reving at stead state. Transients aren't good, but I probably need to continue working on the 50-100 kpa area. One problem is that I get sync loss under very specific circumstances: holding the rpm steady and adding load by letting out the clutch. I add a tiny bit of gas to stablize, and this almost always causes a sync loss <2000 rpms. The loss occurs with ID of 5. I can sustain a 700 rpm idle indefinetely with no sync loss, but add load and it loses it, usually causing me to stall. I can drive around slipping the clutch at 3000+ rpms perfectly well, just not at low engine speeds.

So thinking about it more, I remembered that I merged in the changes to trigger on rising edge, due to my using of the lm1815 board. I tried reverting that change and it became much much worse (I did re-set the timing). It loses sync about every second, no matter a low rpm idle or holding it up to high rpms (often with a nice loud backfire).

So if the stock code triggers on both falling and rising edges, and the merged code only on rising ... is there code that triggers only on falling edges? Taking into account I am now using a MAX board.

Also, what are the qualifications for getting a number placement for which car I am? I technically did drive up and down my driveway all day.
via
QFP80 - Contributor
Posts: 57
Joined: Mon Dec 02, 2013 5:57 pm

Re: My 87 toyota supra hardware build

Post by via »

I tried switching the branch code to trigger on falling edge, and had similar results as to the vanilla eventeeth code, so its probably safe to say the branch code triggering on falling edge is the best.

Attached is an example of some of my issues: Specifically around olv time mark 3200 is a good one. The ego holds relatively even, but then sync is lost.
Attachments
2014.01.18-18.03.23.bin.bz2
(171.84 KiB) Downloaded 846 times
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: My 87 toyota supra hardware build

Post by Fred »

(08:14:56) Fred: COUNT_OF_EVENTS_IMPOSSIBLY_LOW_NOISE   5
(08:15:28) Fred: most likely you've got a rare bit of noise on your sync pulse, so it arrives early and there haven't been enough primary pulses yet

(10:25:19) Fred: see the two-high step?
(10:25:31) Fred: a little premature too IMO
(10:25:51) Fred: so the fake one came through, lost with code 5, then the real one, then carry on


Image
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
Post Reply