Help! I Purchased "used" megasquirt i think its broken

[powermedic]

I Purchased a "used" megasquirt v3.57 online and now I am having issues. I will try and list what I have found so far.

This is the pre made 3.57 board but has a few modifacations. It has mapdaddy which was poorly installed and it looks like pin 2 which should be ground is "burned" through the board, I think its the PO that is to blame. the gound makes contact sometimes but i fear it may be shorting out between the 4 layers? there is no power to the map daddy, infact there is no power to a portion of the board (c2-c10 and r2-r11) I do not think this is normal.

when it is hooked up to the computer (but not the car) the engine map gauge fluctuates or is pinned at max 413.5 kpa. coolant temp is stuck at 82C and intake temp is 19C. if you put a jumper across either of these inputs and the sensor return ground they fluctuate slightly but do not change value. throttle is stuck at 116%. the voltage reads 29 volts all the time even tho i measured 11.5 from my power source.

I have contacted the PO but he has not gotten back to me.

any testing tips or advice would be appreciated.

thanks

[jsmcortina]

Sounds pehaps like there is a short on the analogue 5V supply on the board - that would give false readings on all analogue inputs. Check the schematics for the V357 board in the manual.

James

[powermedic]

I have traced the 5v all the way to the "L2" where it has power on the one side but not the other, there is ~144k of resistance through it and no voltage on the other side. is this correct? there are no solder joints anywhere near this.

thanks for you help so far.

I also do not have 5v to the map sensor, so I think we are on the trail.

[billr]

That 144K ohms through L2 sounds way too high. I would expect something under 100. Is that measuring right at the L2 leads? Post some pictures of the inductor and that area of the board; back side of the PCB, too, if you think there is any damage.

Edit: Yeah, the spec on that Digi-Key P/N is "48 mOhm Max" (.048 ohm). That is for the V3.0 through-hole board, but I expect the V3.57 SMD part to have a very similar DC resistance.

[powermedic]

There is no damage to the board. Nothing soldered in that area. I ordered a new one, will be here on Monday. I will post an update after it is in. Thanks

[DaveEFI]

There is no damage to the board. Nothing soldered in that area. I ordered a new one, will be here on Monday. I will post an update after it is in. Thanks

If you want to play in the meantime, simply solder a wire link across the existing one. The choke is only to help with any possible interference on the supply.

But do use a suitable fuse in the power supply to your MS - or a low current (max 1 amp) supply. It could be the reason the choke has failed is a short somewhere after it.

[powermedic]

My power source is 12v and 150ma

[DaveEFI]

My power source is 12v and 150ma

That's not enough to run an MS and Stim. Needs to be 500mA. Actual current draw shown here is 320 mA - but you need a bit more when loading code.

[powermedic]

msfail.jpg

So I replaced the L2 now I have ~5v to the sensors, however the map is always pinned at max voltage, and the tps is also pinned at max voltage. the coolant temp and the intake air temp do not change when u short them to ground. it is the same when plugged into my new power supply (12v at 600mA) or plugged into the car, no change...

is there something wrong with my wiring? I checked it twice, seems fine.
also I only saw a glimpse of cranking signal once.

here is a screenshot of what happens when its plugged in to the car or just power.

[powermedic]

its also telling me that battery voltage is 29v, but testing with a multi meter its 11.58v
I also plugged power directly to the ms without my harness, so I know that its internal.

[powermedic]

We removed the mapdaddy but left the header pins in. there is no continuity between any of the header pins. however there is still full signal on the map. we measured 0v to the reference pin while this was happening.

when testing pin 23 on the daughterboard we have 0v, assuming a full map signal it should have 5v?

also pin 21 and 22 have ~2v...

[DaveEFI]

Use your stim and a DVM and check that you get a varying voltage at the actual processor input from one of the sensors which isn't working. Make that say coolant, and pin 25 on the daughter card measured to ground. As you vary the CTS pot on the stim, you should see the voltage at pin 25 change too. I suspect you'll find it doesn't change.

L2 being open circuit is a pretty unusual fault - although I've not repaired many V3.57 boards. I'd wonder just why it failed, and check all of the Vref rail after it, since that is the common factor for all those sensors not working. I'd first check C20&22 haven't failed short circuit. It's possible you have a burnt out PCB track too, if L2 had failed.

[elaw]

Also check the voltage at processor pin 31.

The analog-to-digital converter in the processor is "ratiometric": it doesn't return absolute input voltage, it returns the ratio of the input to a reference voltage. So unrealistically-high readings can result from either the input being high or the reference being low. And pin 31 is the +5 reference connection.

[powermedic]

ok,

tested the change in voltage from the coolant temp sensor at pin 25, with my stim plugged in, it is 0.0v with max resistance, and goes to 0.8v with minimum resistance. exactly the same with the intake air temp on pin 24.
with just power and no stim plugged in (ie open) it get 0.8v to both
i get just under 5v at pin 31.

c20 has no power to it and c22 seems to have 5v on the one side, but not the other...

there are no burn marks, or other issues on this section of the board, but is it possible that something happened here? L2 is right beside C22...

[billr]

I think you need for somebody with more electronics or MS experience to work on this board. Where are you located? There may be an experienced member nearby who can take a look at it.

[powermedic]

Vancouver island, British Columbia, Canada

[DaveEFI]

ok,

tested the change in voltage from the coolant temp sensor at pin 25, with my stim plugged in, it is 0.0v with max resistance, and goes to 0.8v with minimum resistance. exactly the same with the intake air temp on pin 24.
with just power and no stim plugged in (ie open) it get 0.8v to both
i get just under 5v at pin 31.

c20 has no power to it and c22 seems to have 5v on the one side, but not the other...

there are no burn marks, or other issues on this section of the board, but is it possible that something happened here? L2 is right beside C22...

Refer to the schematic for the power supply. C20 and C22 are wired in parallel across the Vref 5v rail. After L2. They should both have 5v and ground. If not, the track has burnt out between them. Check both aren't short circuit. My guess is one is - which caused the L2 failure.

[elutionsdesign]

I'm in Bellingham, just across the boarder, and I do MS repair.

[powermedic]

Thanks, sent it to a local guy.

[powermedic]

For those who were interested, and as a thanks to those who helped me, I heard back from the guy I sent it to and he said
The 2 pedestals for the two coil drivers weren't insulated from the drivers, and there bases were thin metal pushing into 5v reference traces. That means there was >100 volts separated by a very thin film of "solder mask". This caused a major short on the board which is why the ADC (analog digital converter) fried, along with a few other components.

so its done, hoping to get my money back but who knows.


thanks again to all those who helped.