Speak EV - Electric Car Forums banner

41 - 59 of 59 Posts

·
Registered
Joined
·
4,394 Posts
The PreCharge Contactor coil driver durations are about 200ms for the pulse and about 1000ms for the main on period.

I'm guessing this is some sort of logic check to make sure the Precharge Contactor is not welded shut?????
 

·
Registered
Joined
·
65 Posts
The behaviour that you have recorded ( a brief pulse on the Precharge contactor followed by a longer "On" period ) corresponds to what was seen by Civjdh in post #11, when he was using LEDs, rather than a 'scope, to indicate the status.
 

·
Registered
Joined
·
4,679 Posts
@freddym

Your scope has confirmed what the fault code suggests - negative and pre-charge contactors are being powered however the ECU is not happy with the voltage across the capacitor being correct within the specified amount of time so aborts instead of engaging the positive contactor.

Is it possible for you to put the scope on the HV out of the battery pack ? You could measure this on the 20 amp fuse inside the MCU inspection hatch, or at one of the cables for the heater or A/C compressor.

Be aware that the HV system is fully isolated from the ground of the car - you cannot connect your scope ground to both the 12v ground and HV negative at the same time as you will cause the HV system to fail the isolation test when it checks the insulation resistance between the HV system and normal electrical system. The entire HV system including battery, motor, inverter etc "floats" relative to the chassis with opto isolators / transformers etc being the only connection between the two systems. (Even the CMU's in the battery pack use opto-isolators for the can bus signals)

You could connect a battery powered scope (I assume yours is?) to the HV negative and MCU fuse (or just across the PTC heater plug) and use a simple opto-isolator circuit to connect another channel to the pre-charge contactor relay coil if you want to look at the relative timing of the contactor and HV voltage rise although I'm not sure the relative timing is that important - the HV voltage rise curve alone is probably all the info you need.

It would be very interesting to look at the HV voltage charge curve on a storage scope - how high does the voltage go, (relative to actual battery voltage) how long does it take to get there, and indeed does it rise at all! This would help confirm or rule out a number of possibilities.
 

·
Registered
Joined
·
4,394 Posts
Many thanks for all the very informed replies.

First of all to answer the queries on measurement of the HV voltage rise at the MCU.
Is it possible for you to put the scope on the HV out of the battery pack ? You could measure this on the 20 amp fuse inside the MCU inspection hatch, or at one of the cables for the heater or A/C compressor.

Be aware that the HV system is fully isolated from the ground of the car - you cannot connect your scope ground to both the 12v ground and HV negative at the same time as you will cause the HV system to fail the isolation test when it checks the insulation resistance between the HV system and normal electrical system. The entire HV system including battery, motor, inverter etc "floats" relative to the chassis with opto isolators / transformers etc being the only connection between the two systems. (Even the CMU's in the battery pack use opto-isolators for the can bus signals)

Yes the scope I'm using is portable so floating wrt to true earth. If I remove the lid of the MCU (later design of MCU so possibly doesn't have an inspection hatch) I can use the 20A fuse to measure precharge positive voltage rise (time and amplitude). Not sure where I will pick up the negative side of the traction pack within the MCU. Using a 10:1 probe will presumably enhance safety whilst making the measurements.
Since the scope only has a single earth reference to cover both channels, it would be inadvisable to make a simultaneous measurement of the 12v coil driver output????
To measure the traction pack voltage directly (as alternative to Canion, Diagbox or HOBDRIVE) I guess the Chademo port is the best access point (but would need a hack to energise). All suggestions most welcome.

Best get cleaning the MCU (mud, dirt etc)

thanks all.
 

·
Registered
Joined
·
4,679 Posts
Yes the scope I'm using is portable so floating wrt to true earth. If I remove the lid of the MCU (later design of MCU so possibly doesn't have an inspection hatch) I can use the 20A fuse to measure precharge positive voltage rise (time and amplitude). Not sure where I will pick up the negative side of the traction pack within the MCU. Using a 10:1 probe will presumably enhance safety whilst making the measurements.
From memory the Battery HV negative is also available within reach of a probe through the MCU inspection hatch as the hatch covers where the 5 HV cables come into the upper compartment of the MCU. I'm sure your later version will still have the hatch - without the hatch you wouldn't be able to disconnect the cables before removing the MCU.

Or if you measure at the PTC heater plug then you have +/- directly available on the plug. (But that would require the car up on ramps and removal of the plastic under covers so not as convenient!)
Since the scope only has a single earth reference to cover both channels, it would be inadvisable to make a simultaneous measurement of the 12v coil driver output????
Like I said, you can't do this as the HV and 12v/chassis electrical systems must remain isolated or it will at the very least trigger a fault condition that will shut down the HV system, worst case you zap your scope...

I'd just measure the high voltage charge up ramp - that should give the info needed.
To measure the traction pack voltage directly (as alternative to Canion, Diagbox or HOBDRIVE) I guess the Chademo port is the best access point (but would need a hack to energise). All suggestions most welcome.
I don't like the sound of that hack. I would say that Diagbox/Canion/Hobdrive would be sufficiently accurate to measure the battery side of the contactors, so as a first pass I would try that. They are able to report the traction battery voltage with the key turned on before attempting to go into ready mode.
Best get cleaning the MCU (mud, dirt etc)
It sounds like you think I'm suggesting you remove the entire MCU top cover - I'm not, just the small cable termination cover near the front edge of the top of the box - it's about 10cm x 5cm in size, held on by just a few small bolts with a re-usable gasket. I have photos somewhere but I'd have to dig them up. I'd probably pull the safety plug first in case you drop a bolt inside and then put the plug back in just before you're ready to take the measurement.

Just pop the rear motor bay cover off and find where the 5 orange cables enter the MCU (the box on the right looking into the boot) and remove the small cover near where the cables go in.
 

·
Registered
Joined
·
4,679 Posts
141094


The 20 amp fuse is in the top left. The two bolted terminal lugs on the left go to battery HV +/- so measure there. The three on the right go to the motor.

I wouldn't bother with more than a quick dust down of the top cover. As long as the flange where the o-ring is is clean you won't have any trouble sealing it up again.
 

·
Registered
Joined
·
4,394 Posts
View attachment 141094

The 20 amp fuse is in the top left. The two bolted terminal lugs on the left go to battery HV +/- so measure there. The three on the right go to the motor.

I wouldn't bother with more than a quick dust down of the top cover. As long as the flange where the o-ring is is clean you won't have any trouble sealing it up again.
IMG_20210221_162917.jpg
 

·
Registered
Joined
·
4,394 Posts
Sorry for the replies, my phone is misbehaving and licking up. The redesigned MCU has a larger inspection hatch which covers 100% if the upper layer. Negative and positive battery pack supplies and easy to connect to.
 

·
Registered
Joined
·
4,679 Posts
Sorry for the replies, my phone is misbehaving and licking up. The redesigned MCU has a larger inspection hatch which covers 100% if the upper layer. Negative and positive battery pack supplies and easy to connect to.
Interesting - I was aware the OBC got redesigned but this is the first I've heard of a redesigned MCU!

I wonder how that fits into the fault analysis when some P1A15 faults are purportedly in the MCU itself - does that mean old and new MCU's have common circuitry where the fault is occurring or does that perhaps point the finger in a different direction such as a faulty pre-charge resistor or contactor? (which are both inside the battery enclosure...)

Looking forward to seeing the voltage ramp - I don't think anyone else with this fault has taken an accurate voltage ramp snapshot on a scope yet?

It should be able to divide the fault tree between an actual problem with the voltage ramp and a problem with the sensing of the voltage ramp, which lead off in two entirely different troubleshooting directions...
 

·
Registered
Joined
·
4,394 Posts
Interesting - I was aware the OBC got redesigned but this is the first I've heard of a redesigned MCU!

I wonder how that fits into the fault analysis when some P1A15 faults are purportedly in the MCU itself - does that mean old and new MCU's have common circuitry where the fault is occurring or does that perhaps point the finger in a different direction such as a faulty pre-charge resistor or contactor? (which are both inside the battery enclosure...)

Looking forward to seeing the voltage ramp - I don't think anyone else with this fault has taken an accurate voltage ramp snapshot on a scope yet?

It should be able to divide the fault tree between an actual problem with the voltage ramp and a problem with the sensing of the voltage ramp, which lead off in two entirely different troubleshooting directions...
Mystery deepens.
Measured the same voltage with the scope as reported by HobDrive peak, about 220v. Scope of course is 22v because of 10to1 probe.

Then It got a bit complicated. After removing the 200A fuse to the OBC, I got a ready. Measured voltage on the scope within 2volts of HOBDRIVE.

So MCU measurements must be OK
 

Attachments

·
Registered
Joined
·
4,394 Posts
Mystery deepens.
Measured the same voltage with the scope as reported by HobDrive peak, about 220v. Scope of course is 22v because of 10to1 probe.

Then It got a bit complicated. After removing the 200A fuse to the OBC, I got a ready. Measured voltage on the scope within 2volts of HOBDRIVE.

So MCU measurements must be OK
20A fuse of course.
 

·
Registered
Joined
·
4,394 Posts
Next start was a fail, only 166v not 220v any more.

Time rise was only about 20ms, the charging of the "condenser" terminated. This was not timed out, looks more like a fault in the contactors, charging resistor, EVECU or the 12v supplies!
 

·
Registered
Joined
·
4,679 Posts
Mystery deepens.
Measured the same voltage with the scope as reported by HobDrive peak, about 220v. Scope of course is 22v because of 10to1 probe.

Then It got a bit complicated. After removing the 200A fuse to the OBC, I got a ready. Measured voltage on the scope within 2volts of HOBDRIVE.

So MCU measurements must be OK
If the voltage reported by Hobdrive agrees with the scope I think we can rule out a measurement error in the MCU.

Assuming you're looking at the right figure in Hobdrive - Hobdrive can report both the MCU reported voltage across the MCU capacitor and also the BMS reported voltage inside the battery pack which I think is just the sum of the CMU reported voltages from each cell bank. Both of course should read the same when the contactors are engaged.

One question though - 220 volts is way too low - for an 80 cell car that's only 2.75 volts per cell. That can't possibly be right??

Full charge is 4.1 volts per cell, 4.1 x 80 = 328 volts. Fully discharged is about 3.6 volts per cell, so 3.6 x 80 = 288 volts. 220 volts would be low enough to set fault codes.

Looking up P1A15 I note the following:


STEP 7. M.U.T.-III freeze frame (FFD) data

Check the freeze frame data.

Freeze frame data
  • Item No. 15 : EMCU: Condenser voltage

OK: 220 V or more
Bit of a coincidence that 220v is the threshold where a fault code will be set ?? Could it be right on the edge of triggering the fault?

Also, the picture you posted doesn't show the rise time, it looks like a near vertical line. Can you expand the time base a lot so that the rise time is stretched across most of the screen so I can see the shape of the curve?
 

·
Registered
Joined
·
4,394 Posts
If the voltage reported by Hobdrive agrees with the scope I think we can rule out a measurement error in the MCU.

Assuming you're looking at the right figure in Hobdrive - Hobdrive can report both the MCU reported voltage across the MCU capacitor and also the BMS reported voltage inside the battery pack which I think is just the sum of the CMU reported voltages from each cell bank. Both of course should read the same when the contactors are engaged.

One question though - 220 volts is way too low - for an 80 cell car that's only 2.75 volts per cell. That can't possibly be right??

Full charge is 4.1 volts per cell, 4.1 x 80 = 328 volts. Fully discharged is about 3.6 volts per cell, so 3.6 x 80 = 288 volts. 220 volts would be low enough to set fault codes.

Looking up P1A15 I note the following:



Bit of a coincidence that 220v is the threshold where a fault code will be set ?? Could it be right on the edge of triggering the fault?

Also, the picture you posted doesn't show the rise time, it looks like a near vertical line. Can you expand the time base a lot so that the rise time is stretched across most of the screen so I can see the shape of the curve?
Sorry, for my poor reporting. 220v is a failed READY, reported value from HOBDRIVE and scope, both are very close. Pack is about 50%SOC, about 314V.

With the 20A fuse removed, the reported Voltage rise and measured scope voltage rise was 166V. This might just be a random fluctuation.

The interesting aspect is that I'm getting a lot of " no response" when turning the key to start. Nothing, no clacks, HV warning light. Turning the ignition key to the off position and trying again seems to work on maybe 2nd, 3rd, even 4th try.

On the voltage rise, here is a P1A15 trace, about 22ms to hit 208V measur Failed READY voltage rise.jpg
Failed READY voltage rise.jpg
ed by scope at the MCU. Scope voltage is divided by 10.
 

·
Registered
Joined
·
4,394 Posts
From the German P1A15 thread, the combined capacitance in the MCU and OBC is 1000microF, pack voltage is 314v, precharge resitance is 24 Ohms

If these numbers are correct for a 2016/2017 Czero (with 80 cells and later design of MCU), the time constant should be 24msec.

Or after 24msecs one should be seeing 63% of 314V or 198v. The scope recorded 208v after about 22msec. (the difference is possibly down to different assumptions as to the capacitance of the system?).

The curious thing is why does the capacitor charging terminate after only 22msecs????
 

·
Registered
Joined
·
668 Posts
From the German P1A15 thread, the combined capacitance in the MCU and OBC is 1000microF, pack voltage is 314v, precharge resitance is 24 Ohms

If these numbers are correct for a 2016/2017 Czero (with 80 cells and later design of MCU), the time constant should be 24msec.

Or after 24msecs one should be seeing 63% of 314V or 198v. The scope recorded 208v after about 22msec. (the difference is possibly down to different assumptions as to the capacitance of the system?).

The curious thing is why does the capacitor charging terminate after only 22msecs????
I havent tested this but I don't think the MCU needs to charge to full pack voltage, just enough to prevent arcing on the main + contactor. It could charge the capacitor to half pack voltage then bring in the main contactor if the capacitor has charged sufficiently? Otherwise there might be a slightly-too-long delay when starting the car perhaps? As I say this isnt tested and is just a theory but maybe we shouldnt expect to see full pack voltage from precharge. Someone who wanted to test this could disconnect the main + contactor and see what voltage you get in the MCU. Would be a bit of a faff to test but would be interesting.

Cheers.
 

·
Registered
Joined
·
4,679 Posts
From the German P1A15 thread, the combined capacitance in the MCU and OBC is 1000microF, pack voltage is 314v, precharge resitance is 24 Ohms

If these numbers are correct for a 2016/2017 Czero (with 80 cells and later design of MCU), the time constant should be 24msec.

Or after 24msecs one should be seeing 63% of 314V or 198v. The scope recorded 208v after about 22msec. (the difference is possibly down to different assumptions as to the capacitance of the system?).

The curious thing is why does the capacitor charging terminate after only 22msecs????
Are we sure the MCU is reporting the correct voltage during the charge period?

Can you show a screenshot of Hobdrive to make sure you're looking at the right parameter? There are at least two different traction battery voltage readings in Hobdrive that are coming from different ECU's.
 
41 - 59 of 59 Posts
Top