[rusEfi] Project EcoBoost Smurf (Miata NA) #19

Your chance to introduce yourself and your vehicle
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: Project EcoBoost Smurf (Miata NA) #19

Post by kb1gtt »

Yeah please keep us posted. Good to hear the LED's are blinking, that pretty much means that part of the board is working. Can you see the fuel LED's?

When it died, what happened, was it like an on and off switch, or did is die slowly? Of course don't forget to check compression. I'll keep my fingers crossed that it's only a fuse and scraped wire some where.
Welcome to the friendlier side of internet crazy :)
Thommm
Posts: 98
Joined: Mon Apr 04, 2016 7:47 am
Location: Netherlands
Contact:

Re: Project EcoBoost Smurf (Miata NA) #19

Post by Thommm »

It's the discovery board for sure, I first went through all of the supply pins to make sure there was no spiking on them which could have killed the board. I currently have the board on my desk but the issue is strange, I'll start at the beginning. Bare with me(in chronological order):
-On the laptop tuner studio froze, unplugged the USB as soon as I noticed (I didn't at first because I was just cruizing around, not minding the gauges constantly)
-Restarted Tuner studio (was doing 10 kmu on a country road so no issues there :p) while I had the USB for rusEFI unplugged
-Plugged the USB back in, windows gave a "device malfunctioned and is not recognized" notice, TS didn't do anything.
-About 10-20 seconds later the car started to stumble and I came to a halt.
-I could keep her running by wildly trashing the throttle and keeping the rev's above 3500, but power was minimal (not enough to even slowly move)
-If you kept the throttle stationary at whatever opening she died
-Parked smurf and called my dad for a tow
-A frighting 20 minutes later I was home (you see, smurf is tiny when towed very close to a massive, new BMW 5-series, I couldn't see a damn thing and it was late at night and dark)

-next morning I started by unplugging the discovery board from the frankenso board and trying it on my desk (bare discovery with only USB (either) plugged in)
-kept getting "malfunction" warnings from windows, so the discovery wasn't recognized by Windows.
-checked the frankenso board and started probing around with my scope (no discovery installed, only 12V from the car battery)
-no weird shit, I had received two new discoveries for upcoming projects 8-) which I installed
-car started right up, very lumpy idle
-stupid me forgot the MAP line, plugged that in and she's pirrin' like a kitten, been driving it for the past 3 days without significant problems, only the usual idle stuff (because it's manually fixed, cold = normal RPM, hot = high RPM idle)


So it's definitively the Disco board, observations made with the "broken" board plugged in (NOT the new one):
-spark worked perfectly
-idle pin led looked like normal brightness
-one of the fuel leds worked sometimes (blinking, but PWM'ed) other fuel channel didn't do anything

So it looks like one/two of the fuel pins in the disco board died, I've thought of something to test it but I'm working long days now :/
But so the weird thing is that the USB interface died, the fuel died but the spark pins work perfectly (otherwise the car would work at all)
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Project EcoBoost Smurf (Miata NA) #19

Post by AndreyB »

Interesting stuff!

So TS was connected, I wonder if we've got a software problem here like maybe some noise or bug have caused very invalid configuration.

1) Before we try to erase this discovery, can you try connecting via TS using the USB on Frankenso? Not sure if you have this part of Frankenso populated.
2) as a different option, can you get a TTL/USB dongle and connect via TTL pins PC10/PC11 without Frankenso?
3) only if everything above fails can you "erase" the chip and re-program it from scratch? if that would help we would confirm invalid configuration in flash somehow.

Looks like you need to have a flashed backup discovery board in the car going forward?
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Thommm
Posts: 98
Joined: Mon Apr 04, 2016 7:47 am
Location: Netherlands
Contact:

Re: Project EcoBoost Smurf (Miata NA) #19

Post by Thommm »

I don't have that part of the frankenso assembled, what I will do is hook up an external USB to serial converter, atleast I might be able to communicate (because that's not possible at the moment)

Edit: this is the device status in device manager:



Windows has stopped this device because it has reported problems. (Code 43)

A request for the USB device descriptor failed.

EDIT 2: if I remove the jumper PP1/JP1 I can connect to the debugger part of the disco board, if I jumper it while connected it malfunctions again and the LEDs that are on start shining brighter, so it might be a regulator. But I'm running out of daylight, will continue tomorrow :p keep you guys posted! @andrey, how is the endurance stuff going? :p haven't been online a lot recently :O
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: Project EcoBoost Smurf (Miata NA) #19

Post by kb1gtt »

On the new fixed board, can you sniff the IO that's broken on the bad discovery, and double check if there are some spikes?

Also I would suggest you program a second discovery and keep it as a spare, just in case this one also dies for some reason.
Welcome to the friendlier side of internet crazy :)
Thommm
Posts: 98
Joined: Mon Apr 04, 2016 7:47 am
Location: Netherlands
Contact:

Re: Project EcoBoost Smurf (Miata NA) #19

Post by Thommm »

Shit just got more interesting :O

Refering to the discovery manual: http://www.st.com/content/ccc/resource/technical/document/user_manual/70/fe/4a/3f/e7/e1/4f/7d/DM00039084.pdf/files/DM00039084.pdf/jcr:content/translations/en.DM00039084.pdf
schematic starts at page 31
So I found that one side of the JP1 jumper had shorted to the 5V out from the USB (after a protection diode), the other side was showing 3V as it should. No freaking way I could trace where it was coming from (couldn't find a near-direct link in the schematic anywhere, except the jumper JP1). I then touched the side that should not be at 5V with the DMM at amp setting, had a small spike of around 0.3A, fell back directly to 0.05A. Remeasured the voltage, bam 2.5V. Reconnected the jumper and the board is working fine, like it has always done.

While it works I realized I haven't found the issue yet, so I won't stop, but that's where I stand at the moment.
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: Project EcoBoost Smurf (Miata NA) #19

Post by kb1gtt »

Hmmm, that almost sounds like a reset-able fuse. I know your gold wires between the chip and the pin can sometimes work that way. Basically you run enough current through them to melt them and re-connect an open circuit or vise versa.
Welcome to the friendlier side of internet crazy :)
Thommm
Posts: 98
Joined: Mon Apr 04, 2016 7:47 am
Location: Netherlands
Contact:

Re: Project EcoBoost Smurf (Miata NA) #19

Post by Thommm »

That means something caused an over-current, right? Please enlighten me on the gold die wires melting, I'm an EE undergraduate (converted from what we have in The Netherlands) and I've not heard of that before so please teach me :). Is it something I can investigate to prevent it from happening in the future or is it one of those one-of-a-kind issues?
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: Project EcoBoost Smurf (Miata NA) #19

Post by kb1gtt »

The ASIC contains the rail clamping diodes. If you over voltage an pin on a chip, and if that pin had protection diodes, the diodes dump pretty much un-restricted current to the rails. Between the ASIC and pin there is typically a gold wire as shown in the below video.



After these wires are installed, they POT the entire thing to make it into that chip you solder to the board. If you mildly over current a pin, generally the very small diameter gold wires will generate heat which gets transferred to the potted materiel. If the heat can transfer to the potted material, then it's common the thermal expansion causes a break in the potted materiel which then allows in air, or allows out gassing of the wire materiel, or some times other materials in the ASIC or potting materiel.

Also it's worth noting that the STM series of chips has a reputation of having large clamping diodes which can absorb significant energy before the diodes break. Now if you have a voltage surge, and the diode dumps to the rail, you'll have a short spike of current which can cause a short spike of heat in the gold bond wire. This can cause the gold wire to melt and become liquid. As well when this happens it take some amount of time for the heat to transfer to the potting materiel. If it's a short surge, you may not heat up the diode or potting materiel to the point of failure. If that happens then the surge goes away, the wire can cool down and the potting materiel will keep its shape. However a surge also generally has magnetic field and other such things which can cause the gold to get compressed in areas resulting in an gap in the wire. In a more controlled method, that's how they turn on bits of FPGA fabric, as well when you program a chip and then protect it's memory it generally does it by blowing a fuse. Now if you again create an over voltage, you can bridge the gap and re-melt the wire. If you do this with a bit less of a surge, you can remove the gap and the wire is again a wire. You could potentially do this many times, and as long as the potting materiel doesn't crack and allow outside contamination or allow for out gassing of internal materials, or doesn't short to nearby circuits, etc, it can happen quite often.
Welcome to the friendlier side of internet crazy :)
Thommm
Posts: 98
Joined: Mon Apr 04, 2016 7:47 am
Location: Netherlands
Contact:

Re: Project EcoBoost Smurf (Miata NA) #19

Post by Thommm »

I know that gold wires connect the die to the pins, didn't know they could be so interesting ;P Thanks for the explanation! :D

So I'm working on getting meth injection and the EBC working, but I'm not getting any data according to my new cheapo logic analyser. Then I noticed the console gives the warning: "no FSIO code" I'm running version 20160913, that's not an old version or something right :?:
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Project EcoBoost Smurf (Miata NA) #19

Post by AndreyB »

Thommm wrote:"no FSIO code"
Looks like somehow you got one of the FSIO pins selected in configuration but you have not set the expression. Please post your configuration or look at the FSIO dialogs
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Thommm
Posts: 98
Joined: Mon Apr 04, 2016 7:47 am
Location: Netherlands
Contact:

Re: Project EcoBoost Smurf (Miata NA) #19

Post by Thommm »

Wealth of information coming! Dunno if it's useful but I took a lot of screenshots with the logic analyzer. Something interesting (may already be known); I don't have (like tomiata) the miss and then double pulse on injection, for me it sometimes misses a fuel pulse and sometimes it misses a spark pulse, but no double pulsing. I've got screenshot attached, some miss fuel, some miss spark. Interestingly, sometimes I noticed the miss is at the exact same RPM (2900) and sometimes it is at a slightly higher value of 3100 RPM (but never something in between, 2900 or 3100). The spark miss seems to be at (calculated based on pulse frequency) 3055 RPM, the fuel miss I don't have exact data on (I get an unreasonably high number).

I've also added startup screenshots, it starts, then it dies for a short period and then it keeps running :O
Attachments
debug pic-scope-movie.rar
(245.65 KiB) Downloaded 310 times
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Project EcoBoost Smurf (Miata NA) #19

Post by AndreyB »

This issue is definitely a work in progress. I am currently focusing on fuel misses, spark misses would be next step.

What version are you running? Can you please update to latest? There were a bunch of fixes within the next two months.
In "MIL Settings" can you please set warningPeriod to zero? that would turn off warning codes throttling, this case I would have more info in the log file. Oh, and the log file :)
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Project EcoBoost Smurf (Miata NA) #19

Post by AndreyB »

I've made yet another fuel miss fix recently but have not yet had a chance to try it, any chance you can compare the before and current firmware versions?
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Project EcoBoost Smurf (Miata NA) #19

Post by AndreyB »

Can you please post a picture of your MAP sensor and hose? Wonder how you've mounted it and where you do you take vacuum.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Thommm
Posts: 98
Joined: Mon Apr 04, 2016 7:47 am
Location: Netherlands
Contact:

Re: Project EcoBoost Smurf (Miata NA) #19

Post by Thommm »

Just did a great engine swap!
IMG_20180908_195415432.jpg
IMG_20180908_195415432.jpg (638.87 KiB) Viewed 11372 times
I'm sorry Russian that I went MIA before I got to answer your question, think you figured it out in the last two years ah?

Update on my Ecoboost Miata: It's been doing too well for the past 2 years, I've only occasionally had to replace the turbo and tighten stuff, otherwise it's been running like a charm. I think I have about 20.000mi on it with RusEFI and boost. Just parked Smurf to fix a little more rust on the body, structurally she's very sound (fixed that 3 years ago) but the body is getting worse. Taking this opportunity to build the engine for more boost, forged rods and such. Will post a few more pictures and such in the coming days.

Also have a new project car running which will switch to standalone, still debating what controller to use :lol:
Attachments
IMG_20180422_102720469_2MP.jpg
IMG_20180422_102720469_2MP.jpg (394.13 KiB) Viewed 11372 times
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Project EcoBoost Smurf (Miata NA) #19

Post by AndreyB »

Thommm wrote:
Sun Sep 30, 2018 9:25 am
it's been running like a charm. I think I have about 20.000mi on it with RusEFI and boost
That's really nice to hear! Myself and that other maniac love hearing these words :)
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Thommm
Posts: 98
Joined: Mon Apr 04, 2016 7:47 am
Location: Netherlands
Contact:

Re: Project EcoBoost Smurf (Miata NA) #19

Post by Thommm »

Yeah, sorry it took so long for you to hear it. I really appreciate the work everybody has put in, so I should've given an update earlier.
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: Project EcoBoost Smurf (Miata NA) #19

Post by kb1gtt »

Yes, really cool to hear it's working well as a longer term project. As well, I like the eco friendly barley horse power. I've heard of alcohol, engines. Looks like that makes every pub a filling station :) Did I see the licence plate saying OPEC-SUCK-IT :)
Welcome to the friendlier side of internet crazy :)
Post Reply