Hello from Puerto Rico

[spags]

Hi everybody,

I live in PR and own a 1994 EUDM miata. I'm interested in getting my car running on the Frankenso board (from the looks of it hardware wise it should just drop in).

Objectives: My ultimate goal is to get a fully programmable ECU as good as or better than stock. However I've read faqs and realize this is still under heavy development and expect (hope?) to get my hands dirty writing some software.

Rationale (Why i'm here): I really like tinkering with cars and computers. Also the ECU is one of the last bastions of the closed source doctrine. I believe that basic things like compilers, kernels, web servers, and industrial controllers (like ECU's) should be open for study and improvement. I also like the idea of using an RTOS to do this instead of a whole lot of special spaghetti code (MS i'm looking at you)

About the car: The car is a 1994 EUDM miata with a 1.8L BPZE engine. Other than a cold air intake, everything in the engine is stock. It is my daily driver although not my only car (by far my favorite though). All of the maintenance is done by me and is up to date. In the future some NA mods will take place (i.e exhintake, NB head). It will never be FI because i consider it the devils work.

About Me: I'm a programmer for a largish software company and have some experience in hardware programming. FPGA's back in college and arduino's more recently. As for programming skills I'm pretty handy from C and assembly all the way up to python and powershell (igh). I've already built my own ECU before (a DIYPNP MS). I was running it as my daily successfully until it was stolen (the ECU not the car).

Questions (all of them for the BPZE engine):
- Where do I get a Frankenso board? From what I can tell rev. 0.2 is the latest and most complete. I only see 0.1 on tindie
- Are there enough inputs to drive sequential injection and sequential spark (at some point in the future, with harness changes)
- Can A/C be made to work? It doesn't have to work yet but the hardware must be there.
- Is the Frankenso board PNP as far as wiring harness is concerned?
- Is there a linux toolchain?
- What kind of resolution are you getting on the outputs?

I'm still browsing through the docs and such so please excuse any uninformed questions.

Thanks.

[AndreyB]

Good morning, that's an impressive intro I hope you would stay around

- Where do I get a Frankenso board? From what I can tell rev. 0.2 is the latest and most complete. I only see 0.1 on tindie

I do not have 0.2 manufactured yet - if suddenly someone donates $200 I will order a batch of rev 0.2 right away. If this suddenly does not happen I would wait till i sell all the 0.1 boards. it's kind of a lame policy but I am not sure how to resolve it - I need to limit the amount of $ I am wasting on all this.

- Are there enough inputs to drive sequential injection and sequential spark (at some point in the future, with harness changes)

Yes for sure

- Can A/C be made to work? It doesn't have to work yet but the hardware must be there.

No idea. I hope so, but until this is done it's not for sure

- Is the Frankenso board PNP as far as wiring harness is concerned?

Yes, Frankenso is perfect for a 1994 Miata - same connector and the default wiring is for 1994 miata. For other cars like a 1996 miata you need to run more cross-wires on the board.

- Is there a linux toolchain?

the build server is on linux. I personally have not used the IDE on linux but I believe it has been done.

- What kind of resolution are you getting on the outputs?

http://rusefi.com/forum/viewtopic.php?f=2&t=2

[kb1gtt]

Good to hear from you and welcome along.

Some links that may be of interest if you haven't already come across them.
http://rusefi.com/wiki/index.php?title=Vehicle:Mazda_Miata_1994
http://rusefi.com/wiki/index.php?title=Manual:Hardware_Frankenso_board
http://rusefi.com/wiki/index.php?title=Manual:Start-a-project/en

Feel free to update the wiki also I didn't follow some of you acronyms like FI and NB head. Feel free to add them to this page
http://rusefi.com/wiki/index.php?title=Engine:acronym-phrases

About A/C, I see the schematic simple shows the AC is some how connected to pin 1J. We do not currently have anything connected to that pin, but we do have this WXX jumper thing that allows re-mapping of the ECU pin out. If it's a low side drive like an injector you can connect that 1J pin with a jumper via W10 wire to an unused injector channel. Long term we expect a little add on PCB's that goes across those WXX connectors which would remove those jumper wires. However we haven't gotten that far yet, so for now, it's either 0R resistors, or jumper wires if you need to change the pin out.

Also we have been making some progress on the IAC valve. A fellow in Texas has been really helpful in getting this ironed out.

[AndreyB]

I didn't follow some of you acronyms like FI and NB head.

the evil of acronyms Forced Induction? and Miata NB is the 2nd gen Miata - they've improved the head design. We've swapped an NB head onto a FWD Mazda BP engine in the yellow race car and it was a pretty cool improvement.

[spags]

Hi guys, thaks for the warm welcome.

kb1gtt, thanks for those links. The 1st was was especially helpfull.

russian, from the errata for the rev 0.1 frankenso i see that it's mostly silkscreen errors and pin inversions but should be made to run. If the 0.2 board is not set in stone I would have one feature request. A prototyping area would be nice to have in future versions. The DIYPNP MS has one and it's used for all sorts of freaky.

If i'm not missing anything rev 0.1 doesn't have any problems that can be fixed with some solder and wire.

Would you be willing to ship a unit to PR? Tindie wants $45 to send a box to PR. I think it's unreasonable.

I guess my first task is building a spreadsheet that ties the pinout in the ecu harness to the pins and functions in the frankenso board. I want to focus in A/C since it's a pre-requisite for me down here (The heat is unbearable). Next is setting up a toolchain and simulations.

[AndreyB]

Not set in stone and a prototyping area is a pretty cool feature to have eventually, but we are kind of out of space so unless we change it dramatically I doubt it would happen. Jared @ is the hardware boss anyway

https://answers.yahoo.com/question/index?qid=20110214151055AAjqJBM says that PR is USA for the purposes of USPS Flat Rate, weird why would Tindie not accept it this way. Anyway I've just added a separate shipping option for PR check if it makes more sense now? By the way I am out of the country till the weekend.

The Miata 1994 wiki page has the schematics and if you register on the wiki and tell me your password and promise not to do crazy shit I can give you Write permission so that you would finish the spreadsheet @ http://rusefi.com/wiki/index.php?title=Vehicle:Mazda_Miata_1994

[kb1gtt]

I've added a prototype area to the known issues list. Which is prompt me to consider it before the next spin. Thanks for the suggestion.

[spags]

I've added a prototype area to the known issues list. Which is prompt me to consider it before the next spin. Thanks for the suggestion.

Awesome, I think it's going to open the door for some good stuff.

As for the spreadsheet, I've done my own. I based it off the table in the miata wiki page. I've added some missing wire descriptions as well as a simple way of accounting for inputs, outputs as well as what uses a PWM or analog channel. It's still incomplete but you comments, corrections and additions are welcome.

I also had a look at the A/C schematics. 4 (2 inputs, 2 outputs)wires go into the ECU but i think you only need 2 (one input and one output) for basic functionality. Plus the software to enable the A/C and compensate with the ICV. Hardware wise it's just sensing switches and driving relays.

Russian, thanks for fixing that on tindie (we don't even know if we are part of the US ). Do you recommend i get the parts kit and put it together myself or get a finished board. I don't mind soldering but there seems to be a lot of SMD components on the board.

Here's the spreadsheet. https://docs.google.com/spreadsheets/d/1EO1QuAx5dOACP6xwKwAqhJgdPtN8wQgik_vFP-fsBc8/pubhtml please let me know if you would like access to make corrections. Feel free to copy it or post the info.

Thanks.

[tomiata]

Welcome to the club

re: linux tool chain
http://rusefi.com/forum/viewtopic.php?f=5&t=690&p=11185#p11185

[kb1gtt]

About AC, Is there an AC schematic some where? I can tell allot from a process / flow schematic. I see these pins of interest.

1J = AC relay. This is probably held low to turn on compressor. Can we confirm how this is activated?
1L = coolant fan relay. This likely turns the fan full power, and likely is pulled low when the AC relay is held low. Can we confirm how this is activated? This probably had very little to do with the AC, and is probably for turning the fans on to cool the engine off during stopped traffic.
1Q = refrig press switch. This is likely an over pressure indicator and probably turns the compressor off when a signal is seen on this pin. Can we confirm if this is activated when 12V or 0V?
1S = heater control unit. No idea what this means, or how it's activated.
2S = condenser fan relay. This likely turns the fan on like 1/4 power, and likely is pulled low when the AC relay is held low. Can we confirm how this is activated?

[AndreyB]

Do you recommend i get the parts kit and put it together myself or get a finished board. I don't mind soldering but there seems to be a lot of SMD components on the board.

I have two arguments for getting a finished board:
a) so far there are only 3 (three) persons in the works who have assembled one of these boards - myself included. Quite a few got distracted or horrified along the way
b) you will get the board faster and being more $ invested you are more motivated to not put this project on hold

On the other hand, if you get soldering paste, air gun and a small tip soldering iron, the assembly is pretty straight forward - just the question of time.

[spags]

Thanks tomiata. Will be building my toolchain during the week.

kb1gtt - This is where I found the schematics http://www.miataforumz.com/1990-200-Miata-Wiring/94sys.pdf
The AC schematic is the first one. Lists 4 wires going into the ECU. The pressure sensor and temp sensor as anded to the a/c switch AFIK as i can tell. That makes the system self regulating.

russian - You have an order for a built frankenso kit . Have to go hunt for a case this weekend.

[AndreyB]

russian - You have an order for a built frankenso kit . Have to go hunt for a case this weekend.

Sweet! I'll set it up for a 1994 Miata to the best of my knowledge

[kb1gtt]

Great, that tells me quite a bit.

Condensing fan and AC clutch are low side drive, so normal injector driver is fine. The pressure sensors are an input that let's you know it's OK to activate the AC clutch, and the fan blower motor has feedback which allows you to guess at how many BTU's are being pulled from the cooling system. AKA, if the fan is going slow, the AC clutch is duty cycled a certain amount, such that you avoid running into those pressure switches. If the FAN is full on, then you can let the AC clutch stay on 100% of the time, as the system will keep proper pressure levels. Or it should, then if something is mechanically wrong, the pressure switches are the fall back.

Hmmm, wheel speed sensors.....

[spags]

Low side drive? Still catching up on the hardware. Going to the glossary later to figure that out.

Are there enough outputs to run the A/C relay and the fan relay? The fan uses a relay (one speed) so as long as the output can drive the relay it should be fine.

The way I see it a typical cycle runs like this:

1 -The user pushes the AC button (input 1). That sends the signal into the ecu, baring that system pressure and temperature are within range. The ecu's opens the ICV and ramps up to 1200 RPM (IIRC)
2 - Once the engine is ready to take the load, ECU enables the AC relay. Compressor clutch engages and the system starts cooling. Condenser fan starts up to avoid system heating.
3(a) - The temp sensor reaches a tripping point and it tells the ecu to stop using ac. Reverse of 1 & 2 happens. Except fan stays on.
3(b) - The user juices the engine to above 4k RPM. ECU disengages the compressor to unleash all that insane horsepower (130hp). Reverse of 1 & 2 happens. Except fan stays on. It might disengage to protect the compressor clutch from damage. Not sure why but it does it.
4 - The user feels that the weather is great and drops the soft top. Disables the A/C: ECU kills the fan and the a/c relay.

Step 3 is the only reason i can think of having two separate outputs, one for the evaporator fan and one for the rest of the system. When the system is working the fluid gets hot and needs to be cooled even (especially?) when the compressor is resting.

None of this answers what that heater signal is though.

Hmmm, wheel speed sensors.....

Somewhat, that cable feeds from the speedo on the cluster which feeds from the tailshaft on the transmission. The cruise control a totally separate subsystem (thanks mazda!) Along with the accelerometers it could be used as some sort of skid detection. Retard timing if tranny speed and actual acceleration don't line up. Not sure if it would work.

[kb1gtt]

For power circuits, you have a loop that goes like this +12V to load to switch to -12V. That would be called low side drive as the switch is closer to the - side of the battery. However if you have +12V to switch to load to -12V, that's called high side drive. Low side drive is a bit easier and cheaper to do, as MCU's typically run at 0-5V which is closer to -12V. High side drive with MOSFET's requires additional circuitry. So many things are low side drive. However high side drive can remove the 12V, so it is some times used if you want to force the circuit to go inactive. AKA, a fuel pump in an accident can have twisted metal short the long wires that feed the fuel pump any were to GND and keep the pump going. While high side can de-energize the long wires and it would not matter if they get shorted to GND. The pump would be off either way.

You sound correct or at least mostly correct in your explanation of the A/C circuit. The only thing I see is that with heater fan at say 25% on, if the A/C clutch is 100% and you are at full RPM, you will likely create unsafe pressures, which will then cause the pressure switches to trip. So to prevent this it's likely that the AC clutch is disengaged before the pressure switches trip. I might guess that if you put the fan on 100%, then full rev the car, you won't have the AC clutch disengage and you'll limit that 130HP to 128HP, or what ever the AC consumes.

[AndreyB]

Sweet! I'll set it up for a 1994 Miata to the best of my knowledge

The board should be ready around Thursday, but I still need to order some high-wattage resistors to ship with the board - that's about the idle solenoid issue as in http://rusefi.com/forum/viewtopic.php?f=3&t=666&start=10

[jfive]

When I shipped a Flat rate box to PR for an ebay sale i had to pay an additional 5.00 over the already great price. Shipped a heavy demon carb for 16.95! I'd double check that its the same as US 50 states price, but I'm sure its not 45.00. UPS is crazy high compared to USPS.

[spags]

My board is here!!!

Thanks russian. Everything looks good, still have to program it and run tests. But its here!!

Will have some questions as I work through programming and testing. Stay tuned!

Edit: It's already programmed. Just wanted to add that the boards come pre-flashed.

[AndreyB]

as I work through programming and testing. Stay tuned!

Good

board testing mode

there are some testing commands once the board is inside the car

By the way, your http://rusefi.com/wiki/index.php?title=Manual:Engine_Type is
set_engine_type 24

[spags]

Hi guys, I've been a little distracted, death in the family and people flying in from all over. I've managed to sqeeze some time to play around with the board. Still not mounted on the car (hope to do that this weekend).
I hooked up the board to the provided USB TTL dongle. I got the following after connecting to it and typing 't'

Code: Select all

[ludwig@horton java_console_binary]$ picocom -b 38400 /dev/ttyUSB0
picocom v1.7

port is        : /dev/ttyUSB0
flowcontrol    : none
baudrate is    : 38400
parity is      : none
databits are   : 8
escape is      : C-a
local echo is  : no
noinit is      : no
noreset is     : no
nolock is      : no
send_cmd is    : sz -vv
receive_cmd is : rz -vv
imap is        : 
omap is        : 
emap is        : crcrlf,delbs,

Terminal ready
4989 ts_p_al

I also got the java console working and tested it, seems to work fine (compiled from repository). I found that russian has already set up the engine type for a miata.

I can't seem to get TunerStudio to work though. Here's what I have
v2.6.04
Running on linux I get two serial devices, the USBTTL dongle (/dev/ttyUSB0) and the micro usb port (/dev/ttyACM0)
I'm using the *.ini from the repository although I haven't reflashed the card yet. That might be an issue. Don't want to reflash until after i install in the car.
Using the /dev/ttyACM0 device (or the ttyUSB0 device for that matter). It fails to test. And gives me the following errors in a dialog If i go ahead and try to use it as is.

Screenshot of TS error.

rusefi_screenshot.png (41.8 KiB) Viewed 18644 times

I'm probably missing something. Will keep working on it during the weekend.

[kb1gtt]

I seem to recall there is a jumper to change which port you get TS vs terminal / dev console. You may need to set or remove that jumper. I'm sure russian can comment about the jumper.

[puff]

http://rusefi.com/wiki/index.php?title=Manual:Software:User/en
Jumpers

Communication mode
Grounding PB1 flips the default communication port mapping
Configuration reset
Grounding PD6 resets saved configuration to default state - see http://rusefi.com/forum/viewtopic.php?f=5&t=373&p=9571&hilit=PD6#p9571
Board Testing mode
Grounding PB0 enabled board testing mode. See http://rusefi.com/wiki/index.php?title=Manual:Hardware_Test_Mode

[AndreyB]

On the attached screenshot we can see rusEfi dev console protocol messages in Tuner Studio error dialog. You are connecting to the wrong port.

Flipping the port mapping would help in terms of making sure that the dongle is wired up properly. If in doubt please attach a picture of how the USB thingy is plugged into the board.

Sorry about loss

Feel free to PM me your phone number if you want to chat on the weekend re: trying it in the car for the first time. I hope we would not fry anything, but unfortunately I cannot guarantee that.

[spags]

Ok. Info pm'd. Will keep trying to get it to work tonight.

[spags]

Hey guys,

Time has been at a premium this week. Wish I could dedicate more time to rusefi but I have my family visiting for the week. I managed to get
some time to plug the ECU into the car. Didn't get around to starting it but verified that critical systems are working. Here are my findings until now.

What i got to work:
- IAC control works. Did not fiddle too much with it but I could hear the PWM working on the valve. Changing the pwm duty cycle changed the valve.
- TPS sensor works and reads ok (I think it needs re-calibration; will only go up to .87)
- IAT reported sane values, same for MAF.
- CAM sensor works. I get trigger signals as well as sane RPM readings when cranking (didn't have enough sense to check the tach though)
- Spark works, on both banks (got juiced while testing it).
- Fuel pump turns on at the right time.
- Nothing broke! ....yet.

What still needs testing:
- Injector control. Any way of testing this short of pulling out an
injector?
- Fan control?
- Alternator control (see wierdness below)

What was wierd:
- Got a low voltage light on the dashboard, not sure if it's supposed to be like that. Will track that light down on the schematics to the ECU when I get a chance.

Hopefully will get a chance to plug it in and test before the week is out. Next time I will be going for an engine start.

I've attached logs of the runs. In case they are of interest.

[AndreyB]

What still needs testing:
- Injector control. Any way of testing this short of pulling out an
injector?
- Fan control?
- Alternator control (see wierdness below)

Believe it or not injectors are pretty loud - you would hear them if you engage them while the engine is not spinning. http://rusefi.com/wiki/index.php?title=Manual:Software:dev_console_commands#I.2FO_testing has the commands which would allow you to engage all the injectors one by one.

It's a shame but I did not provision a fan relay wire in your board I am a child of cold weather apparently. Good news there are enough spare transistors on the boards, I will provide you with the info on where to put the jumper wire hopefully tomorrow.

More or less the same story with alternator control Sorry it's all so raw. This would be the 2nd thing to add.

[spags]

Russian,

Thanks, will probably check the injectors tomorrow night or saturday morning. Will SMS you after everything checks out to see if you're available so we can try for a start.

Re: Cooling fan, it's going to be a must for driving it around. Here we can still see 32 degC during the day . No rush though, not of much use if I havent even got the car running yet
Re: The alternator and battery light. Not a big issue for me. The alternator on the NA's is self regulating, the NB is not self regulating and it needs to be done at the ECU. I'm hoping its simply an issue of a line that has to be set hi or something like that. For the future and NB miatae; there is a regulator circuit posted on the MS forums, should be easily adaptable to rusefi.

[AndreyB]

Cooling fan logic is already there - it's a question of adding a wire and a property for the pin assignment, nothing more.

Alternator: here is 1994 schematics:


It kind of seems like the ECU does play a role there, seems like it is pulling the field wire down?