Page 8 of 13
Re: microRusEfi
Posted: Thu Dec 19, 2019 10:15 pm
by unfgr
Hardware-wise, both MCUs ara pin compatible,
what is your reasoning? unsuported functionality in M7 series?
Re: microRusEfi
Posted: Thu Dec 19, 2019 10:32 pm
by AndreyB
unfgr wrote: ↑Thu Dec 19, 2019 10:15 pm
Hardware-wise, both MCUs ara pin compatible
There is an opinion that 100 pin versions of F4 and F7 are NOT pin compatible. The reason to change the footprint from F7 100 pin to incompatible F4 100 pin is price reduction.
Re: microRusEfi
Posted: Thu Dec 19, 2019 10:43 pm
by unfgr
You are right, i stand corrected!
they have indeed different pinout in this package (a big screw-up from ST if i may add since they advertise full pin compatibility from their M4 to M7 series)
Oh well, no rest for the wicked... new PCB incoming...
Re: microRusEfi
Posted: Thu Dec 19, 2019 11:13 pm
by AndreyB
0.4.5 has bugs and there is a hope of 0.4.6 in coming days. more on Slack chat room
Re: microRusEfi
Posted: Fri Dec 20, 2019 1:34 am
by DonaldBecker
russian wrote: ↑Thu Dec 19, 2019 10:12 pm
unfgr wrote: ↑Thu Dec 19, 2019 10:09 pm
Will the 0.4.5 (2L) PCB work with the F7 MCU?
I believe that no, 0.4.5-2L is F4 only.
Correct. The 0.4 design is F4 only. A board cannot be made compatible with both F4 and F7.
The primary reason for switching to the F4 was that JLCPCB assembly regularly stocks the F4. The F7 is available as an 'extended' part, with changing availability. Today the 100 pin F7 part in stock is the F745, and they only have 18 of them. The only F767 type they have in stock is the 144 pin part, and they list only 4 of them (the minimum board order is 5 parts).
The STM32 family is pinout compatible (*).
(*) Here is minor detail behind that claim that makes it entirely false. The 100 pin package F7 made a major incompatible change to the STM32 pinout. One side of the chip has every signal shifted over by one pin. Major, stupid incompatibility. ST doesn't say exactly why they did that, beyond some nonsense statement about USB power. It appears that someone screwed up the bond-out pattern, and they didn't want admit it or throw away chips. So they broke compatibility for their customers.
Re: microRusEfi
Posted: Wed Jan 01, 2020 6:33 pm
by AndreyB
Code: Select all
Changes for 0.4.7
Corrected mixed-up Vref/Vbatt/AV4/AV5 inputs
Put USB Vin on a regular diode clamp, not the +V clamping rail.
Changed USB diode array +V clamping to the internal 5V rail.
Selected a different micro USB connector, a through-hole type that is more common and easier to solder. (The previous type had through-hole mounting ears so this doesn’t change the bottom clearance.)
Rotated the USB connector to follow wide-side-down orientation expectations.
Reduced brightness of USB power green LED (R48 currently 2.2K previous 1.0 Kohm, 2.5mA)
Moved debug connector away from USB connector (previously it barely fit)
Moved the Reset button adjacent to Program button for easier entry to DFU mode.
Changed the push button type to one with extending SMD leads for easier inspection.
Added termination (diode and 1206-size pull-up) to the LIN bus circuit.
Added a 0-ohm jumper for 5V standby power to Wake-On-CAN, remove to disable.
Added a 0-ohm jumper R21 to connect DFB-11 to the AV9 input, remove to disable
Removed thermal breaks on the bottom layer under TLE thermal pad. Solder mask remains.
Modified layout to improve bottom layer ground current from TLE8888 to mounting holes.
Changed TLE8888 ignition outputs from test point to 1x4 header hole size
Added a dual diode to allow STM32 RTC standby power simultaneously from coin cell header and from 12V constant (through main relay coil)
Changed pin assignment for communication header to allow I2C and CAN1 outputs.
Grouped test points on the edge of the top layer, renumbered and added 3.3V.
Changed freewheel diodes part and footprint to SMA (original was 500mA)
Added Tag-connect programming pads
New gerbers available
https://github.com/rusefi/hw_microRusEfi/blob/master/gerbers/rusEFI-Micro_R0.4.7.zip
iBom for this version
https://rusefi.com/docs/ibom/microRusEfi_R0.4.7.html
Re: microRusEfi
Posted: Fri Jan 03, 2020 12:49 pm
by AndreyB
Polarity validation image for 0.4.7
Re: microRusEfi
Posted: Sun Jan 05, 2020 3:18 am
by texasgdp754
I know there are the 4 ignition high side outputs on the 48 pin molex header for driving coils. Can the 1 x 04 header from tle8888 drive 4 additional coils with internal igbt. ?.
I see the ign in circuits are grounded in tle8888 and not hooked to stm32 like the inj in circuits . looking at data sheet. Still unsure.
Apologize if I'm missing the obvious here
Re: microRusEfi
Posted: Sun Jan 05, 2020 6:12 am
by DonaldBecker
The TLE8888 allows the four ignition trigger outputs to be controlled by either the four dedicated inputs or by using the SPI bus interface.
MRE grounds the four inputs, but they can still be controlled over the SPI channel. It's just a matter of software to do that.
The same is true of the 5 low-side outputs on the 1x5 header -- they can be controlled over SPI.
Re: microRusEfi
Posted: Tue Feb 04, 2020 8:10 pm
by Dron_Gus
Can some one with MRE run this
https://github.com/dron0gus/rusefi/tree/tle8888-dev code and check if 'pins' command shows valid diagnostic for TLE8888 pins? Or just copy-paste its output here.
Re: microRusEfi
Posted: Wed Feb 05, 2020 7:08 pm
by mck1117
I assume it's been tried on a tle8888 dev board already?
Re: microRusEfi
Posted: Thu Feb 06, 2020 12:55 pm
by Dron_Gus
mck1117 wrote: ↑Wed Feb 05, 2020 7:08 pm
I assume it's been tried on a tle8888 dev board already?
Nope. I do not have any HW with TLE8888 attached. Actually I have tle8888 and breadboard. But I better wait for real MRE that doing this wiring nightmare.
Re: microRusEfi
Posted: Sat Feb 08, 2020 6:20 pm
by AndreyB
Re: microRusEfi
Posted: Mon Feb 17, 2020 3:13 pm
by AndreyB
Finally available on eBay as a 75% assembled kit!
From $60 for kit up to $199 for final assembly in USA
ebay.com/itm/333517397424
Update: new eBay link
https://www.ebay.com/itm/333532119947
Kit instructions -
https://github.com/rusefi/rusefi/wiki/Hardware_microRusEfi_kit_instructions
Re: microRusEfi
Posted: Tue Feb 18, 2020 8:51 am
by Dron_Gus
- 97244d5s-1920.jpg (335.86 KiB) Viewed 37255 times
Re: microRusEfi
Posted: Sat Feb 22, 2020 4:02 am
by AndreyB
All six kits sold, still one last assembled microRusEfi 0.4.7 available!
Buyers have the option to stay anonymous or I can add them to
https://rusefi.com/forum/viewtopic.php?p=32462#p32462 list if they would allow.
Re: microRusEfi
Posted: Sat Feb 22, 2020 9:44 pm
by AndreyB
Re: microRusEfi
Posted: Sat Feb 29, 2020 1:45 am
by AndreyB
0.4.7 board, no load, 17C ambient. Q1 is still a clear winner @ 53C (36C above ambient)
Re: microRusEfi
Posted: Sat Feb 29, 2020 7:51 pm
by dexter_lab
it looks like some of the diodes not the Q1?
Re: microRusEfi
Posted: Sat Feb 29, 2020 8:22 pm
by AndreyB
dexter_lab wrote: ↑Sat Feb 29, 2020 7:51 pm
it looks like some of the diodes not the Q1?
Please note how thermal image is a bit shifted in relation to optical image.
Re: microRusEfi
Posted: Mon Mar 02, 2020 11:56 pm
by AndreyB
0.4.8 is being fabricated now:
1) more mounting points for microRusEfi as Plug&Play module project
2) more vias around Q1
3) two more components were moved from back to top, also a few more components are now pre-assempled by JLC
There is hope to resume sales of assembled units around March 25th.
Re: microRusEfi
Posted: Wed Mar 04, 2020 1:12 am
by Simon@FutureProof
So after discussions about cases and how we are tied to one specific Chinese case at the moment I took a few hours while I was off work to look into a viable alternative.
I did a bit of hunting and found a viable case - Hammond 1455L 1202BK
This case is 123mm x 103mm x 30.5mm and is the smallest stock case I could find. It comes in Silver, Black, Red and Blue.
I then worked out a way to mount the existing MRE plug to the stock case in a way that could be 3d printed and not look rubbish.
Adapter is designed from the CAD on the Molex website so is a very tight fit over the plug, Printing requires elephant foot compensation or trimming to make sure the first layer flash does not prevent fitting. You also need a fairly dimensionally accurate 3d printer.
Print time is about 1 hour using Hilbert curve top layers.
Yes I do still need to workout a way to mount the MRE board nicely. I was considering a trimmed piece of aluminum to sit into the slots of the case and bolt to the MRE board.
Note - This case is a very tight fit on the plug and a little has to be trimmed on it to fit nicely. This bit is shown on the last picture, it can be done with side cutters and cleaned up with a knife, the plastic pins on the bottom that go through the board need just the first few mm clipped off as you can see in the pictures.
Re: microRusEfi
Posted: Wed Mar 04, 2020 1:14 am
by Simon@FutureProof
3d printing STL is attached.
Re: microRusEfi
Posted: Thu Mar 05, 2020 2:42 am
by AndreyB
Kudos to
@abricos for pointing us at USB3131-30-0230-A - a cheaper AND hopefully easier to mount vertical USB connector. 614105150721 is x2.5 the price and 614105150721 is supposed to be used with 1mm PCBs.
Only weirdness is that mouser does not stock it. And digikey does not stock TLE8888.
Re: microRusEfi
Posted: Thu Mar 05, 2020 6:22 am
by AndreyB
Looks like the case we enjoy here has started as Bosch F01R00D072 - ours looks a bit different but mostly the same. And then for Dongfeng Sokon DFSK it became a bit less Bosch and a bit stamped instead of molded.
Re: microRusEfi
Posted: Fri Mar 06, 2020 2:42 am
by Simon@FutureProof
Re: microRusEfi
Posted: Fri Mar 06, 2020 6:34 pm
by AndreyB
36638-0002 would not work on current MRE I guess
While the pin alignment piece of plastic could be removed, the way the terminals are stamped from flat stock we cannot bend them into smaller row spacing.
Re: microRusEfi
Posted: Sat Mar 07, 2020 6:48 pm
by AndreyB
DonaldBecker wrote: ↑Fri Dec 20, 2019 1:34 am
Correct. The 0.4 design is F4 only. A board cannot be made compatible with both F4 and F7.
The 100 pin package F7 made a major incompatible change
Actually it could be that there are three or maybe four different stm32 LQFP-100 pinouts.
https://www.st.com/content/ccc/resource/technical/document/application_note/73/76/21/47/ef/d0/4c/16/DM00164538.pdf/files/DM00164538.pdf/jcr:content/translations/en.DM00164538.pdf
2 layer MRE 0.4.x uses "old device F4" pinout for 168MHz devices
the is also a different "new device F4" pinout for 180MHz devices which is NOT compatible with any MRE at this point.
And then we have 4 layer MRE 0.1-0.3 which uses F7 pinout which is a third different pinout.
Open question if there are differences within F7 family as well
Relevant documents:
https://www.st.com/en/microcontrollers-microprocessors/stm32f4-series.html
https://www.st.com/content/ccc/resource/technical/document/application_note/73/76/21/47/ef/d0/4c/16/DM00164538.pdf/files/DM00164538.pdf/jcr:content/translations/en.DM00164538.pdf Migration of microcontroller applications from
STM32F42xxx/F43xxx devices to STM32F7 Series devices AN4660
Update: maybe above info about F4 is not true? Maybe it's only 469 which is different?
https://rusefi.com/forum/viewtopic.php?p=24291#p24291
Re: microRusEfi
Posted: Sat Mar 07, 2020 8:13 pm
by AndreyB
Re: microRusEfi
Posted: Sun Mar 08, 2020 12:25 am
by Simon@FutureProof
We had a request for a blanking plug/cable gland to use on an MRE instead of the Molex plug so I had a quick go at it this evening.
It was not easy to work out a way to have the geometry of the Molex injection molded plug and still be able to print the part with good quality and printable holes but I think we got pretty close.
STL to follow.
