[mudder]

I would measure the voltage with everything connected, and the key off. If you don't measure the actual pack voltage at the MCMe connector, then disconnect the MCMe connector from the MCM and then measure the MCMe connector voltage. You should get the full pack voltage when the key is off.

 

[LowCarbon]

I installed the current hack board set at 20% today. I logged LiBCM on my test drive. I got this:

Key:ON
Charger: OFF
METSCI buffer sync..
-1,0,A, 178,147,V, 3.724,3.721,V, 3000,mAh, -0.2,kW, 23,C
-1,0,A, 178,147,V, 3.724,3.721,V, 3000,mAh, -0.2,kW, 23,C
-1,0,A, 178,147,V, 3.724,3.721,V, 3000,mAh, -0.2,kW, 23,C
-1,0,A, 178,147,V, 3.724,3.721,V, 3000,mAh, -0.2,kW, 23,C

which repeated continuously. Is this normal, or did I mess something up? The car drove fine and seem zippier. No IMA or CEL lights.

I also installed the monster ferrite on the HV cables while I had the cover off. Is this supposed to improve the reliability of the 20x4 display? It's been behaving itself since the ferrite installation.

 

[mudder]

Yes, that looks like normal display data LiBCM sends while driving. You can change which data is displayed with the $DISP= command (for help, see the help screen which is presented whenever you plug in LiBCM and open the Serial Monitor.

Yes, the ferrite greatly reduces noise coming from the IGBT chopper (inside the PDU). The ferrite prevents this noise from reaching the junction board, hence it is no longer present in the batteries, the LiBCM PCB, and the IMA wire harness. 4x20 noise shouldn't occur with the ferrite installed... although I can't 100% verify that because I don't get display issues in my car except under forced conditions.

 

[BLS]

I would measure the voltage with everything connected, and the key off. If you don't measure the actual pack voltage at the MCMe connector, then disconnect the MCMe connector from the MCM and then measure the MCMe connector voltage. You should get the full pack voltage when the key is off.

Ran the checks, 171 on mcm and 180 from pack so the cheap oscilloscope might be a bit dodgy or I'm just plain measuring it wrong. Ill play around with it tomorrow and Monday to hopefully have a good reading if not ill grab one of amazon.

 

[mudder]

The voltages should be the same at the MCMe and pack. It's likely your DMM has low input impedance, which is causing the 9 volt difference. Given that there is 20 kOhm ESR between the MCMe and pack, I propose your DMM has 380 kOhm input impedance. What is the DMM model number so I can verify this?

If the 9 volt difference is as I've proposed, then it sounds like the LiBCM PCB is working properly. Given that the known-working MCM I sent you also isn't working, I propose the issue is low resistance between the HVDC bus and chassis ground. To properly measure that resistance, we're going to need to apply a large voltage between HVDC- and chassis ground... do you have any test equipment that can do that? If not, I can send you a current-limited power supply to test for leakage.

I can also send you a differential probe that will work with pretty much any oscilloscope.

...

Another proposal: it might be time to come visit me in Chattanooga... you're considerably closer than most insight owners, but it's still going to be an all day ordeal for you. But I have all the tools - and the knowledge - to troubleshoot whatever the issue you're seeing is. I'll do the testing at no charge, given my vested interest in figuring out why LiBCM isn't working in your car... thus far, I believe yours is the only car LiBCM hasn't worked in.

 

[Balto]

The voltages should be the same at the MCMe and pack. It's likely your DMM has low input impedance, which is causing the 9 volt difference. Given that there is 20 kOhm ESR between the MCMe and pack, I propose your DMM has 380 kOhm input impedance. What is the DMM model number so I can verify this?

If the 9 volt difference is as I've proposed, then it sounds like the LiBCM PCB is working properly. Given that the known-working MCM I sent you also isn't working, I propose the issue is low resistance between the HVDC bus and chassis ground. To properly measure that resistance, we're going to need to apply a large voltage between HVDC- and chassis ground... do you have any test equipment that can do that? If not, I can send you a current-limited power supply to test for leakage.

I can also send you a differential probe that will work with pretty much any oscilloscope.

...

Another proposal: it might be time to come visit me in Chattanooga... you're considerably closer than most insight owners, but it's still going to be an all day ordeal for you. But I have all the tools - and the knowledge - to troubleshoot whatever the issue you're seeing is. I'll do the testing at no charge, given my vested interest in figuring out why LiBCM isn't working in your car... thus far, I believe yours is the only car LiBCM hasn't worked in.

@BLS You should totally take him up in this offer. I wish I was close enough that visiting Mudder would be only an all day venture I would 100% do it.

 

[mudder]

Updates:
-I will be out of country JUN16-JUN22. My wife woke up this morning with the urge to visit the British Virgin Islands. This afternoon she booked airfare and rented a house. I'm not complaining.
-I still haven't received the latest LiBCM PCBs. Customs is really enjoying making me fill out completely unrelated worksheets... so far I've done "ball bearing worksheet" and "Aluminum extrusion worksheet" for US Customs. I've been quite clear to them that this package contains printed circuit boards (HSTUS 8534.00.00.40), but that doesn't seem to make sense to them. I've also submitted the invoice twice now. So far PCB manufacturer has been completely unhelpful, and I won't be using them anymore... even at work, where we ordered QTY20,000 PCBs from them last year.
-Assuming the PCBs arrive tomorrow (JUN13), I will only have time to build QTY1 and start testing the various hardware changes. Certainly I won't manufacture (or ship) new LiBCM units until I get back (starting JUN23).

 

[BLS]

Another proposal: it might be time to come visit me in Chattanooga... you're considerably closer than most insight owners, but it's still going to be an all day ordeal for you. But I have all the tools - and the knowledge - to troubleshoot whatever the issue you're seeing is. I'll do the testing at no charge, given my vested interest in figuring out why LiBCM isn't working in your car... thus far, I believe yours is the only car LiBCM hasn't worked in.

I'm using a basic Crenova MS8233D from amazon. My power supply only goes up to 30v which i highly doubt is enough. Thank you for the offer, I greatly appreciate you doing this for me. Ill send you a PM and we can discuss a good day to meet for when you get back from paradise . In the meantime ill go back through and doublecheck everything to make sure its not going to be something obvious and get the car prepped for the trip.

 

[mudder]

The MS8233D appears to have a 20 MOhm range, but I'm not sure it's selectable (it might only be auto-ranging). If you push the "func" button while in the resistance mode, does anything happen that would indicate the range is changing?

Apologies LiBCM isn't working in your car. I'm glad it's not your primary vehicle...
For the trip up, are you familiar with how to connect the DCDC cables to the HVDC output? This will allow the DCDC to charge the battery while you're driving. You don't need the Arduino for this to work... but without it you'll get a check engine light (the car is still drivable).

If you'd like, I can send you the HVDC power supply and differential probe tomorrow... so you can play around with it all while I'm on vacation. I'll write up some instructions so you can test it while I'm gone.

 

[BLS]

Yeah nothing happens so autoranging ig. If only I could use my fluke from work .

are you familiar with how to connect the DCDC cables to the HVDC output?

I do not. Is this the cut the white/grn wire to keep the 12v charged mod?

Ill take you up on that offer. Thank you so much. If somehow we solve this remotely Ill ship them back all together with your mcm since I have received the 40% mod from Bulldog.

 

[mudder]

RevD PCBs finally cleared customs!

 

[Balto]

RevD PCBs finally cleared customs!

As Dr. McCoy once said
'The bureaucratic mentality is the only constant in the universe'.

 

[mudder]

Update: PCB passed my automated tester! Deeper testing starts tomorrow.

...

Update: Finishing up my first round of bench testing. Everything works as designed. I'm going to throw it in the car tomorrow and verify all is good. Here's a video showing the RevD grid charger control circuitry's thermal performance:

 

[mudder]

FYI: I sent @BLS the following troubleshooting package today:
-replacement LiBCM PCB (to verify it's not the PCB). FYI: This wasn't previously an option, since I only had QTY2 RevC PCBs (and nothing else). Now that I have the RevD PCB, @BLS is getting RevC serial #1, which has been working in my car for months.
-Lambda GEN 300/2.5 power supply for testing HVDC leakage, including various test cables.
-a differential scope probe (to measure MCM'e' voltage).
-various test cabling/connectors, so we can connect this equipment to the IMA system.

If the replacement LiBCM PCB doesn't solve the issue, then @BLS and I are going to schedule a Zoom call to troubleshoot further. I have duplicates of all equipment listed above, so we can probably figure out the root cause. Note that I previously sent @BLS a known-good MCM, so if the issue persists, then it's almost certainly a pre-existing condition elsewhere in the IMA system.

If remote troubleshooting doesn't work, BLS and I have tentative plans to rendezvous in Chattanooga. That's a 6 hour drive each way, which ends up being a long day for @BLS... so hopefully we can figure it out remotely.

 

[sharpmj]

John, has anyone else had a problem of not restarting if left in autostop for a few minutes.

 

[mudder]

I don't believe others have experienced this issue. Maybe you're just the first to report this issue? Anyone else see this behavior?

How often does it occur? Are there specific triggers that cause it to occur?

If you didn't get a CEL - and the SoC was above 20% - then it's almost certainly an MCM issue, not LiBCM. Pretty much anything LiBCM could do wrong to prevent auto-restart would cause a CEL.

 

[tervic]

I haven't seen an auto stop restart problem.

 

[LowCarbon]

What I have noticed is that if I slow down with the car in neutral, autostart kicks in at about 10 mph. Then if I put it in 2nd rather than 1st, it doesn't start up. If I move the stick towards 1st, even if I don't actually shift into 1st, it starts up. At least that's the way I remember it. There's some kind of weirdness with autostop when shifting to 2nd rather than 1st. None of this is new with LiBCM. Sharpmj, could this be what is happening to you?

 

[sharpmj]

I don't believe others have experienced this issue. Maybe you're just the first to report this issue? Anyone else see this behavior?

How often does it occur? Are there specific triggers that cause it to occur?

If you didn't get a CEL - and the SoC was above 20% - then it's almost certainly an MCM issue, not LiBCM. Pretty much anything LiBCM could do wrong to prevent auto-restart would cause a CEL.

I think it's the second time it has happened, and only since last firmware update about 10 days ago. Here's what happened, pulled up to my driveway with 60% charge, parked in neutral with auto stop engaged. Out of car and walked over to get mail from mailbox, got back in car after maybe 2 minutes at most and put car in gear, no autostart. Went from 1st to 2nd and back a couple of times, still no autostart. Turned off ignition waited for screen to turn off, tried to restart, nothing happened. Did the ignition off, wait for screen off maybe 5 more time and then finally IMA start engaged and started.
Only concern for this situation would be maybe stopped for road construction or stuck in traffic, left car in autostop and when time to go wouldn't be able to start car.
Also no CEL.

 

[mudder]

Interesting behavior. Note that in a situation where you "need to start the car now," you can always engage the backup starter by holding the key in start for several seconds.

One thing you can do to see if LiBCM is intentionally preventing autostop:
1) plug LiBCM into a laptop and open the Arduino serial monitor
2) Type $DEBUG. LiBCM responds with QTY3 'YES/NO' data values:

-Has LiBCM limited assist since last cleared?: NO**
-Has LiBCM limited regen since last cleared?: NO
-loopPeriod exceeded since last cleared?: YES***

**If this response is 'YES', then it's possible LiBCM was telling the MCM the battery was dead while you were in autostop.
***For now, this value will almost always be 'YES'... due to how the 4x20 display is reset during each keyOFF event.

These values are persistent until cleared by the command $DEBUG=CLR.

3) If the first stored value is 'YES', type $DEBUG=CLR (which will set the value to 'NO').

4) The next time this failure-to-restart issue occurs, plug your laptop in and type $DEBUG. Is the first response 'YES' or 'NO'?
Note: you could check this value right now, but if you see 'YES' then the result is inconclusive (because if LiBCM has ever limited assist the answer will be 'YES'). However, if the answer is 'NO' right now, then LiBCM certainly isn't telling the MCM that the battery can't autostart the car right now.

...

UNRELATED:
NOTE: Starting right now, I am out of country for the next week. I will not have access to insightcentral, except through my cell phone... hence if I do respond to questions posed on ic.net, my answers will be brief.