New BMS for bq76PL455A based batteries

Topics concerning OEM and open source BMSes
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

Using a spare board is a great idea, it isolates it from any wiring loom capacitance too.

Other random thoughts:

The noise might be better or worse at one end of the pack. Is it possible to reconfigure the chain to try connecting at the other end?

As a temporary test you could try adding a reasonably sized HV cap (it would need to be a HF cap, polypropylene metal film or similar) between the module end of the battery pack and chassis gnd. This would reduce the noise voltage at that point and also introduce another lower impedance ground path in parallel with the daisy chain - both these would tend to reduce noise current in the daisy chain link. Note - I don't know enough about your wiring setup to really comment on how good or bad an idea this is!

It's worth checking all the grounds, particularly at the charger. I doubt there is a problem there, I think this is inherent in floating the HV system, but worth a look.

Is the charger L and N the right way round? Again doubt it's the problem but no harm checking.
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: New BMS for bq76PL455A based batteries

Post by catphish »

Pete9008 wrote: Tue Jul 05, 2022 9:24 am As a temporary test you could try adding a reasonably sized HV cap (it would need to be a HF cap, polypropylene metal film or similar) between the module end of the battery pack and chassis gnd.
I think this is definitely worth a try. Do you have a suggestion as to a reasonable size for this capacitor?

PS. Do you think there's any chance of an isolating transformer on the signal lines solving this? Would probably be the easiest solution to retrofit!
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

catphish wrote: Tue Jul 05, 2022 10:43 am I think this is definitely worth a try. Do you have a suggestion as to a reasonable size for this capacitor?
Just big compared (say 10-100x) the daisy chain impedance so 100nF should do it (but If a 1uF was available I'd use that).
catphish wrote: Tue Jul 05, 2022 10:43 am PS. Do you think there's any chance of an isolating transformer on the signal lines solving this? Would probably be the easiest solution to retrofit!
Possibly, not something I'd considered but it would be a nice solution. The trouble would be finding an available/suitable one as it needs to have a pretty good spec. An ethernet POE one might do the trick?
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: New BMS for bq76PL455A based batteries

Post by catphish »

Pete9008 wrote: Tue Jul 05, 2022 9:24 am The noise might be better or worse at one end of the pack. Is it possible to reconfigure the chain to try connecting at the other end?
This would rather difficult and messy even to test because the master connector is on the positive side of each battery, and the daisychain cables are very short accordingly.
PXL_20220701_164335847.jpg
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

catphish wrote: Tue Jul 05, 2022 11:09 am
Pete9008 wrote: Tue Jul 05, 2022 9:24 am The noise might be better or worse at one end of the pack. Is it possible to reconfigure the chain to try connecting at the other end?
This would rather difficult and messy even to test because the master connector is on the positive side of each battery, and the daisychain cables are very short accordingly.

PXL_20220701_164335847.jpg
Does that mean the daisy chain comes of the pack positive side?
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

You've got me wondering about an isolation transformer now, it would be a simple solution. The only thing worrying me is that I've never seen it done before on an RS485 type link - although that doesn't mean it won't work! Transformers also tend to be a bit of a pain, they rarely show anything like ideal behaviour.
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: New BMS for bq76PL455A based batteries

Post by catphish »

Pete9008 wrote: Tue Jul 05, 2022 11:11 am Does that mean the daisy chain comes of the pack positive side?
Yes. Each battery has 2 data connectors, the connector on the positive side connects to the next master, and and connector on the negative side connects to the next slave. They are not reversible, you could put the BMS at the negative end, but the wiring would look very odd.

I've ordered a couple of Ethernet transformers, and a 1uF film capacitor. Fingers crossed that one of them works!
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

catphish wrote: Tue Jul 05, 2022 11:40 am Yes. Each battery has 2 data connectors, the connector on the positive side connects to the next master, and and connector on the negative side connects to the next slave. They are not reversible, you could put the BMS at the negative end, but the wiring would look very odd.
That's a shame, if I had to guess which side would be noisier it would be the positive.
catphish wrote: Tue Jul 05, 2022 11:40 am I've ordered a couple of Ethernet transformers, and a 1uF film capacitor. Fingers crossed that one of them works!
Not keen on the cap as a fix, especially on the positive side, it's more a test to add confidence in the diagnosis. And while I do like the transformer idea I'd only give the it a 50% chance of working (magnetics have probably caused me more grief than all the other component types combined over the years). It's definitely worth trying though given the relative ease of implementation.
User avatar
joromy
Posts: 371
Joined: Fri Jun 28, 2019 12:56 pm
Has thanked: 1 time
Been thanked: 3 times

Re: New BMS for bq76PL455A based batteries

Post by joromy »

I hav struggled with the same noise problems, with isolating comm just with capasitors. It is not a good solution in noisy enviroment!
It's to high impedanse, and that is bad in noisy enviroment.
You want low impedance opto coupler, or a digital isolator.

This is used on the tesla battery boardit is Quad-Channel, but it's possible to get in other sizes.
https://no.mouser.com/c/semiconductors/ ... ies=Si864x
Thomas A. Edison “I have not failed. I've just found 10,000 ways that won't work"
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: New BMS for bq76PL455A based batteries

Post by catphish »

Another option that may work is just adding more common mode chokes / ferrites to the transmission lines.
If all else fails, I will employ a nasty software hack, charge at full current for some period, then pause the charging to allow a BMS reading, then resume.
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

I'm going off the capacitor idea, even just as a diagnostic test. It shouldn't damage anything but there is a small possibility that it could and I'd hate for it to damage either your charger or inverter!

Ferrites or chokes should help, it depends how much of an improvement you need.
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: New BMS for bq76PL455A based batteries

Post by catphish »

Mini update on this:
1) I tested with the BMS powered from my laptop USB. It works! So we're correct about the diagnosis.
2) I shorted B+ to chassis GND (don't try this at home kids). This didn't seem to help.

My transformers have now arrived so when I get an opportunity I'll see if the comms work though a transformer.
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

catphish wrote: Wed Jul 06, 2022 6:00 pm 1) I tested with the BMS powered from my laptop USB. It works! So we're correct about the diagnosis.
:D
catphish wrote: Wed Jul 06, 2022 6:00 pm 2) I shorted B+ to chassis GND (don't try this at home kids). This didn't seem to help.
Bit surprised by this, how did you make the connection?
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: New BMS for bq76PL455A based batteries

Post by catphish »

Pete9008 wrote: Thu Jul 07, 2022 8:53 am Bit surprised by this, how did you make the connection?
I did it with a pair of multimeter probes joined together, one to the charger chassis and one to the B+ terminal. Not 100% sure it was a good connection.

On a related note, when running from the laptop battery and working, unsurprisingly, connecting laptop GND to vehicle GND immediately breaks it.

Transformer testing to follow shortly.
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

OK, that makes sense. Even if the probes made a good connection there will have been enough inductance in the leads to mean it didn't provide an effective earth at HF. Not surprised that connecting the laptop earth broke it either, that's a nice confirmation too.

Much happier about the results now, when a test doesn't give the expected result it's often a sign that you don't fully understand the problem and if you don't fully understand the problem you may well have the wrong solution!

Which transformers have you gone for?
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: New BMS for bq76PL455A based batteries

Post by catphish »

Pete9008 wrote: Thu Jul 07, 2022 9:10 am Which transformers have you gone for?
This is the transformer I'm going to test. I chose is quite arbitrarily. I think the main question will be whether its capacitance is too high for the weak drive of the system, but it's a starting point and I believe is designed for the frequencies we're working with.

https://docs.rs-online.com/3ea2/0900766b81577209.pdf

They're rather small, soldering to test will be a challenge! If it works I'd definitely build a PCB for it.
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: New BMS for bq76PL455A based batteries

Post by catphish »

Mini update: the transformer works! I haven't looked at the waveforms, tested in the car with noise, or anything much, but touching the wires on each side of the transformer, I consistently receive data, encouraging!
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

All the ethernet transformers are pretty similar TBH. The capacitance should be OK for driving and the inductance is high enough too. For me it's whether the primary to secondary capacitance is low enough to do the job. When you wire it up be careful with your wiring to keep the two sides physically separate.

The above results are making me think that the noise is actually pretty high in frequency, say 50-100M+. Might be an idea to check with a scope as a better idea of the noise bandwidth would be useful in terms of working out the best way to deal with it.

Edit - just seen your last post - great news!
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: New BMS for bq76PL455A based batteries

Post by catphish »

I think I will now order a tiny board from JLC, just to house 3 transformers and 3 pairs of molex connectors. That will allow me to test this properly in the car, and to get a scope on there to see how it affects the waveforms. Now I know a transformer works, it probably makes sense to replace the whole capacitor-filter setup on the board, but I'm not in a hurry to buy a whole new batch of boards again!

Edit: Might design a board with just the transformer. This would be small enough to heatshrink inline with the cable for testing.
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

catphish wrote: Thu Jul 07, 2022 9:40 am I'm not in a hurry to buy a whole new batch of boards again!
This is why hardware is a pain compared to software and why I always try to bodge test circuits onto the first prototype wherever possible and only go to a respin when I've seen it work. Experience has taught me that there's often at least one more problem :(

The flip side is once you have something that has been bodged, and is installed and working, there is always the tendency not to pull it all apart again to do it right!
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: New BMS for bq76PL455A based batteries

Post by catphish »

Pete9008 wrote: Thu Jul 07, 2022 9:48 am The flip side is once you have something that has been bodged, and is installed and working, there is always the tendency not to pull it all apart again to do it right!
As long as the bodge works, the important thing is remembering to update the schematics, so that if you (or anyone else) wants to replicate later, the fix is there!
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

catphish wrote: Thu Jul 07, 2022 9:40 am Now I know a transformer works, it probably makes sense to replace the whole capacitor-filter setup on the board
Thinking about it the ethernet transformer includes a common mode choke so that might make the one already on your board redundant. You can't get rid of the series capacitors at the far end, and since they are smaller they have a larger effect on performance. One thing I wouldn't do is try and use the other channel in the transformer for a second BMS channel. It might work but on the other hand you might get crosstalk problems.

I'd still test a transformer on the car (even if it is not installed in the battery box) before committing to more boards. The primary to secondary capacitance could still let enough noise through to cause problems.
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: New BMS for bq76PL455A based batteries

Post by catphish »

catphish wrote: Thu Jul 07, 2022 9:40 am Edit: Might design a board with just the transformer. This would be small enough to heatshrink inline with the cable for testing.
Turns out SOIC16 breakout boards can be bought off the shelf, so will just use one of those and solder the wires, at least for testing.
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: New BMS for bq76PL455A based batteries

Post by Pete9008 »

catphish wrote: Thu Jul 07, 2022 9:51 am As long as the bodge works, the important thing is remembering to update the schematics, so that if you (or anyone else) wants to replicate later, the fix is there!
Splitting hairs but to me bodge implies that you don't expect it to survive long term (vibration, sealing, robustness etc) usually because it has just been done quickly to prove the principle. If it's good enough on all those counts then it's a fix that needs documenting and fixing on the next release.

Unfortunately documentation has never been my strong point!
Post Reply