[info] microRusEFI

Hardware inside and outside of the ECU
Post Reply
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: microRusEfi

Post by AndreyB »

Information for the guy behind https://www.homebuiltairplanes.com/forums/threads/lighweight-opposite-2-stroke-engines.29814/

PCB side male connector alone weights 34 grams.
MRE v0.1 assembled in Seattle with connector weights 66 grams.

It could be that running assembled ECU without connector weights 31 grams?
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: microRusEfi

Post by AndreyB »

Progress/test harness/issues

Note that only 4 yellow LEDs are blinking

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
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: microRusEfi

Post by kb1gtt »

Do I see the injector blinking light is wired as noted in this sketch?
Test_Fixture_Sketch.png
Test_Fixture_Sketch.png (9.19 KiB) Viewed 38275 times
Also see this picture for a suggested strain relief method. This works better than hot glue.
PCB_Strain_Relief.jpg
PCB_Strain_Relief.jpg (192.86 KiB) Viewed 38275 times
Also dental floss is very important.
Welcome to the friendlier side of internet crazy :)
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: microRusEfi

Post by AndreyB »

Nope, here is how this is connected. The left resistor was added because LED was pretty bright even in OFF state, barely getting brighter in ON mode. Now with the left resistor, I get proper on/off light.

This style of stitching would require drilling and probably dust :( But yes that was considered.
Attachments
8888_test_harness.jpg
8888_test_harness.jpg (126.85 KiB) Viewed 38266 times
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: microRusEfi

Post by AndreyB »

Some firmware progress!

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
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: microRusEfi

Post by kb1gtt »

Glad to hear you got it working with a leakage resistor. That is better, as the light is more intuitive than my schematic. Mine was inverted.
Welcome to the friendlier side of internet crazy :)
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: microRusEfi

Post by AndreyB »

+12v into each of the connector +5v lines was a success - MRE has survived.

Wiring GND and +12v was a success - MRE has survived.

Looks like we are ready for 0.3

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: microRusEfi

Post by AndreyB »

@Abricos reminded me to run a longevity test. microRusEfi powered from +12 has blinked LEDs overnight just fine!
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
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: microRusEfi

Post by mck1117 »

I've left mine running for several days with the trigger simulator running the whole time without issue! Powered from +12v.
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: microRusEfi

Post by mck1117 »

This got posted on Facebook, but not in this thread!

microRusefi has now run an engine! Last week I successfully wired and started my Volvo b230 on microRusefi serial number 1.



Since the engine was already running on Frankenso, the tune ported over with almost no effort, and the engine appears to behave exactly the same on the new ECU. Success!

The engine is a 2.3 liter 4 cylinder, running sequential fuel and wasted spark ignition (in two-wire mode, so four independent coils).
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: microRusEfi

Post by AndreyB »

Naive thermal test:

Q1 seems to be the hottest component at 20C-25C above ambient. That's with rear side of the shell acting as radiator.

Without the rear plate I've seen Q1 at 45C above ambient - 25C+45C=70C on a bench with no load on the MRE.



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: microRusEfi

Post by AndreyB »

Also first start with case painted orange!



Sweet MAF oscillation at maximum data rate.
Attachments
javaw_3jLnNtZjbC.png
javaw_3jLnNtZjbC.png (130.39 KiB) Viewed 38129 times
2019-08-18_MRE_NA6_starting_running.msl
(4.15 MiB) Downloaded 553 times
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
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: microRusEfi

Post by kb1gtt »

Does 8888 and/or stm32 have internal temperature sensors? Those would tell you better if there is a thermal problem. With only half the case it is not really correct.
Welcome to the friendlier side of internet crazy :)
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: microRusEfi

Post by kb1gtt »

What does your no load claim for volts and amps? Aka power supply numbers.
Welcome to the friendlier side of internet crazy :)
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: microRusEfi

Post by mck1117 »

Stm32 does. 45-50c internal stm32 temperature is consistent with what I've seen running on 12v, sitting on carpet with only the bottom half of the case.

My board pulls just shy of 300mA. So Q1 is dissipating (12v-6v) * 300mA = 1.8W.

The 8888 should be dissipating around 0.3 watt in its regulator, plus whatever other current it's drawing (perhaps another half watt, at most)

The 3.3v regulator and STM32 account for the vast majority of the remaining 1.5 watts.
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: microRusEfi

Post by kb1gtt »

Don't I recall q1 is switching? Aka power does not directly mean voltage by amps. Well instantious yes but not on a longer time basis like a linear. I wonder if we can decrease the watts with a faster q1. Also I understand D1 gets hot. Perhaps that would have some thermal via's.
Welcome to the friendlier side of internet crazy :)
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: microRusEfi

Post by mck1117 »

Q1 doesn't switch. The tle8888 is using Q1 as an external linear regulator to get the dissipation out of the tle8888 itself, then doing the precise 6v -> 5v regulation internally.

Per the tle8888 datasheet, page 58:
A linear pre-regulator with an external logic level power MOSFET is implemented to keep the power dissipation of the TLE8888-1QK low.
Infineon recommends the IPD30N06S2L-23, which we are using.

Besides, if q1 was switching, then where would all the extra power be going? 4 watts is waaaay above the peak dissipation of the stm32, which is the only thing actually doing any work on the board. Even if it isn't operating fully steady state, it's still spending its entire existence within the linear (ohmic) region, so it's just a resistor. Faster or slower won't save you.

I've also scoped the gate pin (and output), and they're dead steady. It really is operating steady state. No fluctuation.
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: microRusEfi

Post by mck1117 »

kb1gtt wrote:
Mon Aug 19, 2019 5:24 am
Also I understand D1 gets hot. Perhaps that would have some thermal via's.
Yes - D1 is also dissipating maybe 1/10 watt, but it has almost no thermal conduction to the case, and almost all the conduction it has is to the already-hot Q1. Probably a better plan to investigate using a pfet instead, since that will cut total dissipation instead of trying to dissipate from D1 to the case.
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: microRusEfi

Post by AndreyB »

One step forward - ST-LINK header is now fixed!

We now have very optional BOOT button.
Attachments
microRusEfi_0_3_assembled.JPEG
microRusEfi_0_3_assembled.JPEG (52.27 KiB) Viewed 37954 times
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: microRusEfi

Post by AndreyB »

There is an opinion that the fat white line is the outline of the back side cover cavity. There is an opinion that C21 and C22 are on the wrong side of that line?

https://github.com/rusefi/hw_microRusEfi/issues/109
Attachments
MRE_0_3_C21_issue.png
MRE_0_3_C21_issue.png (22.88 KiB) Viewed 37912 times
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: microRusEfi

Post by AndreyB »

We have PCB tracing pull request - see https://github.com/rusefi/hw_microRusEfi/pull/110

with PCB it's probably all or nothing and only one person could be editing at a time :( @kb1gtt I need a binary decision on this PR since that's obviously not my domain :(
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
DonaldBecker
Posts: 32
Joined: Mon Aug 19, 2019 10:40 pm
Location: Los Gatos CA USA

Re: microRusEfi

Post by DonaldBecker »

I've done a re-work of the board layout, and am looking for a few reviews.

https://github.com/DonaldBecker/hw_microRusEfi

There are many changes, but the key points are
- Much wider traces for all circuits carrying significant current
- Segregated ground under the analog section
- Directed ground current to the TLE8888
- Only passive components (pull-down resistors and clamp diodes) on the bottom
- High ground fill percentage on the bottom layer
- Careful layout of components to improve access for hand soldering
- Tighter layout to open up free space for additional features
- Two sided layout with 8.5 mils minimum trace width for lower fabrication cost

This re-work makes no modifications to the v0.3 schematic. This led to some challenges, especially obvious in the cramped reversing of traces to the resistor networks, but should hopefully make review easier.
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: microRusEfi

Post by AndreyB »

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
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: microRusEfi

Post by kb1gtt »

Cool to see a PR. I'll have to review later on. I'm on the phone at my sons practice right now. 2 layer is interesting. I'm a bit concerned about thermal. Less copper could push limits. From what I've seen thermally we are close to the limit already. We need to analyze the thermal as it stands now. The 4427 is limited to 0.125A because if the thermal limits.

There is an onshape 3D of the case. That is fairly close to the real world. I didn't model model the mold release tapers. As well it's dims as mostly bases on my calipers. So tolerances could cause varaitions.

Are you Eastern standard? Should we chat about misc design stuff?
Welcome to the friendlier side of internet crazy :)
DonaldBecker
Posts: 32
Joined: Mon Aug 19, 2019 10:40 pm
Location: Los Gatos CA USA

Proposal for reverse voltage protection

Post by DonaldBecker »

The v0.3 MRE appears to have a weakness with reverse voltage protection (when the battery is connected in reverse, applying -12V to the +12V circuit).

Specifically unprotected 12V power is supplied to the e-throttle TLE9201 H-bridge chip, several pins of the TLE8888 (BATPA, BATPB) and the main capacitor. These are not rated for negative voltage.

I propose that we fix this by separating the two 12V supply pins, with one being the existing "ignition switched" power and the other being supplied with "main relay" switched power. The "ignition switched" circuit is unprotected, while the main relay is controlled by MRE to only make contact when the system voltage is safe (over +4.5V).

With the v0.3 design there are two connected "12V switched" pins on the 48 pin external connector, using two of the eight wider pins.

The only function on the board that draws significant power directly from 12V is the TLE9201 H-bridge. It should be supplied main-relay-switched 12V, which will protect it from reverse voltage. Everything else (except perhaps the 12V sense in the analog section) should be powered from unprotected pin, with BATPA, BATPB and the main capacitor moved "behind" the protection diode.

Moving the capacitor behind the diode would also have the advantage that the capacitor wouldn't discharge through the starter with a brief voltage sag during cranking.

This change would also allow fusing the e-throttle (5A peak?) separately from the other parts (well under 1A), and provide cleaner 12V power.
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: Proposal for reverse voltage protection

Post by AndreyB »

DonaldBecker wrote:
Fri Sep 06, 2019 11:54 pm
The v0.3 MRE appears to have a weakness with reverse voltage protection (when the battery is connected in reverse, applying -12V to the +12V circuit).
https://github.com/rusefi/hw_microRusEfi/issues/114
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: microRusEfi

Post by AndreyB »

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
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: microRusEfi

Post by kb1gtt »

See reverse polarity pmosfet noted at the below link. The pmosfet is typically better that a protection diode. https://rusefi.com/wiki/index.php?title=Manual:Hardware
Welcome to the friendlier side of internet crazy :)
DonaldBecker
Posts: 32
Joined: Mon Aug 19, 2019 10:40 pm
Location: Los Gatos CA USA

Re: microRusEfi

Post by DonaldBecker »

kb1gtt wrote:
Sat Sep 07, 2019 1:43 am
See reverse polarity pmosfet noted at the below link. The pmosfet is typically better that a protection diode. https://rusefi.com/wiki/index.php?title=Manual:Hardware
In this case it's only negligibly better. MRE draws only a trivial current directly from IGN (BAT, BATPA, BATPB). Probably single digit mA when activating the main relay, and almost nothing after. The rest of the current goes through a series of linear regulators. Except during a severe voltage dip when cranking, every mW burnt by the protection diode is a mW that doesn't need to be burnt in the 6V pre-regulator transistor.

It would be different if we were using something like the TLE9471, which has a buck regulator for the 3.3V or 5V supply. There the efficiency of using a MOSFET instead of a diode is worthwhile, and the chip includes a charge pump to enable using a lower-cost N-channel device.
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: microRusEfi

Post by stefanst »

One semi-mystery is solved. The screws holding the housing bottom to the top do seem to be M3.5 x 0.6. Because M3s just won't cut it and M4s are way too big.
I guess.
Post Reply