[azbill]

@azbill Eureka, you found it! Pack switching - You made me finally print out and read GM's patent application (see attachment below) for the two-pack 400/800 VDC pack switching process. The "dip" appears to happen when the charging process takes a short break so the EV controller can switch the 400 VDC AC compressor and IPE modules to the other pack. Since these 400 VDC devices can only be powered by one of the two 400 VDC packs when both packs are connected in series, the powering pack gets less charging amps and its SOC falls behind the non-powering pack, creating an imbalance. Periodically switching these loads back and forth from pack to pack helps to alleviate substantive imbalances. Relevant excerpt from the patent app text below (I inserted some translating data in bold):

0037] During charging at the second voltage level V2 [800 volts DC], there may be a need at times to power an electrical load aboard the mobile platform 20, e.g., to thermally condition to the battery packs 12A or 12B, a cabin of the mobile platform 20, etc.[power the 400 VDC AC compressor, IPE DC/DC converter, etc] Battery pack 12A provides such power when operating in mode (1 ). When operating in mode (2), this function is instead performed by battery pack 12B. The flexibility of using either battery pack 12A or 12B to energize the electrical load while simultaneously charging reduces the chances of a charge imbalance between the battery packs 12A and 12B. Such an imbalance might otherwise occur if one of the battery packs 12A or 12B were used for this purpose during charging to the exclusion of the other battery pack 12B or 12A.

FYI, when looking at the diagram below, the switches in-play for making the 400 VDC electric load change from pack 12A to pack 12B are Sets SA1/SA2 and SB1/SB2

View attachment 3381

You can also see resistors R1 and R2 with switches (PC1 and 2) for doing a final balancing of the packs before going back into parallel mode. That prevents a really high in-rush current.

 

[HVACman]

You can also see resistors R1 and R2 with switches (PC1 and 2) for doing a final balancing of the packs before going back into parallel mode. That prevents a really high in-rush current.

Exactly. Much like the little cell-level voltage-equalizing circuits except at the 100 kWh pack level The patent app also discusses using either physical or solid state contactors. Also, it makes several claims for applying the contactor-based pack-splitting technique to allow a native 800-VDC pack design to split to two sub-packs to charge with 400 VDC chargers. As compared to Porsche's 400/800 VDC DC-DC converter in 50 or 150 kW versions, Lucid's "Wunderbox" electronic DC-DC converter, or Hyundai's novel solution of using the drive unit inverter as a DC-DC converter.

Finally, the patent app includes the contactor logic to permit half-power EV operation even if one of the two packs fail.

Speaking of Porsche's DC-DC converter, this week is Charged EV's semi-annual EV Engineering Conference. Wed's 8 AM EDT session will be about the Taycan 150 kW DC-DC converter by Tier 1 supplier, Preh. It's free, anyone can register, no experience required Another session that performance EV enthusiasts might also enjoy: Explore Rimac's next battery platform

 

[PFRobinson]

For battery longevity/health it's best to keep it charged between 20/80% but Andy Oury said they designed Ultium to be forgiving and you can charge it however you like but its been best practice for full BEVs to set max charge to 80%.

You know more than GM engineers?

 

[Nebula1701]

You know more than GM engineers?

What? It's an established practice for BEV owners... Also GM has put out the 80% for recommend daily charging in their videos....

Here is one for the Lyriq....

 

[Littleshopofguitars]

Do they really know what’s best?

 

[azbill]

Do they really know what’s best?

I have an electrical engineering degree and can tell you leaving the battery at 100% is not good for longevity. Most manufacturers recommend 80 or 90%. Dendrites, which can grow in the cells are much more likely to grow with the cell voltage at the top, I.e. 100%.

 

[Dark-Fx]

You know more than GM engineers?

GM designed it so it wouldn't degrade so much by always charging to 100% that they would have to risk replacing the battery under warranty. It's still going to cause wear that could be somewhat avoided. If you're planning on keeping the truck for 10-15 years, you're better off playing it safe.

 

[HVACman]

Tom Moloughney just reported on this morning's IEVs podcast he has another Hummer EV on-loan and will be doing a bunch of charging tests this week.. VIN #00022....an early build, which is not a good sign that all the components will work to GM's specifications.

Someone needs to forcefully-remind GM brass they really need to keep press-vehicles in like-new shape, otherwise they risk QA-related press-drive problems and get "bad press", which defeats the purpose of having press vehicles. Exhibit A: He already has noted on a popular podcast the hood-closing mechanism isn't working properly

 

[HVACman]

I have an electrical engineering degree and can tell you leaving the battery at 100% is not good for longevity. Most manufacturers recommend 80 or 90%. Dendrites, which can grow in the cells are much more likely to grow with the cell voltage at the top, I.e. 100%.

That, plus fatigue-related mechanical damage to the anode graphite matrix due to repetitive swelling at the last 10% of charge. And the swelling is worse for Si-doped graphite, which is why Tesla's definitely shouldn't be frequently charged to 100%.

 

[azbill]

Tom Moloughney just reported on this morning's IEVs podcast he has another Hummer EV on-loan and will be doing a bunch of charging tests this week.. VIN #00022....an early build, which is not a good sign that all the components will work to GM's specifications.

Someone needs to forcefully-remind GM brass they really need to keep press-vehicles in like-new shape, otherwise they risk QA-related press-drive problems and get "bad press", which defeats the purpose of having press vehicles. Exhibit A: He already has noted on a popular podcast the hood-closing mechanism isn't working properly

Hopefully GM has told him about the pack switching, so that he will not be surprised when the charging suddenly drops and then comes back on. I would think most of the charging issues from early builds would be fixed with software updates.

 

[Dark-Fx]

Tom Moloughney just reported on this morning's IEVs podcast he has another Hummer EV on-loan and will be doing a bunch of charging tests this week.. VIN #00022....an early build, which is not a good sign that all the components will work to GM's specifications.

Someone needs to forcefully-remind GM brass they really need to keep press-vehicles in like-new shape, otherwise they risk QA-related press-drive problems and get "bad press", which defeats the purpose of having press vehicles. Exhibit A: He already has noted on a popular podcast the hood-closing mechanism isn't working properly

When I took my truck to Magna, the hood ended up in a state where it would refuse to close. I think someone pulled the emergency release and then tried to manually close it afterward. I wasn't present when they did it. Just stopped them from being able to drive it out of the plant for a bit while I was manhandling it to get it to close. The trick was to get the latch fully closed and after fully re-opened by using the button on the front. Seemed like it thought the latch was open when it actually wasn't completely.

 

[Nebula1701]

 

[azbill]

View attachment 3554

Looks like pack switching occurs more often when at high rates, i.e. above 300kw. That makes some sense because the imbalance will occur faster. I also suspect some of the charges were done on current limited chargers, MI and NJ in particular.

Is there any data on ambient temperatures for each charge?

 

[Nebula1701]

Looks like pack switching occurs more often when at high rates, i.e. above 300kw. That makes some sense because the imbalance will occur faster. I also suspect some of the charges were done on current limited chargers, MI and NJ in particular.

Is there any data on ambient temperatures for each charge?

He didn't mention any of that in his video comparing the three trucks in the video I posted in another thread.

Someone needs to poke him about it during the InsideEVs livestream tomorrow.

 

[HVACman]

View attachment 3554

GM needs to re-think their default 800 VDC charging strategy, at least until they figure out a scheme to reduce or eliminate these extreme pack-switching breaks.

I added up the PA 350 kW session "gained" kW vs SOC squares over the lower NJ 260 kW max speed curve and then subtracted all the "lost" kW vs SC squares to see which session was net more efficient at getting to 70% SOC. It was + 29 squares and -20 squares, so it probably was a little faster, but not by much. Probably about 40-45 minutes to 70%. The 260 kW rate only had one short balancing drop at 30% SOC. One reason for that is that the AC compressor probably didn't have to work near as hard, as 30% lower pack charging amps means half the pack resistance heat (I squared-R heat). This means a higher % kW going to charging and less going to the pack heat + AC compressor load. Less compressor load means fewer amps "borrowed" from the sub-pack supporting 400 VDC equipment, so that means a shorter, less severe balancing cycle.

Seems to me GM should reduce the maximum allowable charging speed, which would improve charging efficiency, reduce sub-pack imbalance issues, and give the pack a more consistent charging session. There is no evidence that grabbing EV-industry fast charging "bragging rights" with a 360 kW charging speed is buying owners faster net charging times for anything other than extremely short, low delta-SOC charging sessions that don't require pack-switching balancing breaks. On the other hand, this pack-switching scenario is creating a lot of confusion both among Hummer owners (like TFL) and among even the top charging experts in the world.

 

[Nebula1701]

GM needs to re-think their default 800 VDC charging strategy, at least until they figure out a scheme to reduce or eliminate these extreme pack-switching breaks.

I added up the PA 350 kW session "gained" kW vs SOC squares over the lower NJ 260 kW max speed curve and then subtracted all the "lost" kW vs SC squares to see which session was net more efficient at getting to 70% SOC. It was + 29 squares and -20 squares, so it probably was a little faster, but not by much. Probably about 40-45 minutes to 70%. The 260 kW rate only had one short balancing drop at 30% SOC. One reason for that is that the AC compressor probably didn't have to work near as hard, as 30% lower pack charging amps means half the pack resistance heat (I squared-R heat). This means a higher % kW going to charging and less going to the pack heat + AC compressor load. Less compressor load means fewer amps "borrowed" from the sub-pack supporting 400 VDC equipment, so that means a shorter, less severe balancing cycle.

Seems to me GM should reduce the maximum allowable charging speed, which would improve charging efficiency, reduce sub-pack imbalance issues, and give the pack a more consistent charging session. There is no evidence that grabbing EV-industry fast charging "bragging rights" with a 360 kW charging speed is buying owners faster net charging times for anything other than extremely short, low delta-SOC charging sessions that don't require pack-switching balancing breaks. On the other hand, this pack-switching scenario is creating a lot of confusion both among Hummer owners (like TFL) and among even the top charging experts in the world.

View attachment 3555

Speaking of reducing the maximum allowable charging speed... the Silverado WT in both pack versions is capped at 300 kW.

GM Online Order / Reference Guide

www.gmfleetorderguide.com

 

[azbill]

He didn't mention any of that in his video comparing the three trucks in the video I posted in another thread.

Also both EVGO and EA have gone to "power sharing" on their chargers, so that can affect test results as well, if other vehicles are charging at the same time. EA guarantees 150kw when power shared. The EVGO power shared chargers I have seen have two 175kw power supplies, so they likely guarantee 175kw.

The SUV Hummer with smaller battery is also capped at 300kw.

 

[HVACman]

Also both EVGO and EA have gone to "power sharing" on their chargers, so that can affect test results as well, if other vehicles are charging at the same time. EA guarantees 150kw when power shared. The EVGO power shared chargers I have seen have two 175kw power supplies, so they likely guarantee 175kw.

The SUV Hummer with smaller battery is also capped at 300kw.

Tom is aware of the power sharing scenario for the new EA chargers and very likely would have verified his sessions were not in a shared situation. One thing I was wondering after my last post was if the EA power electronics back in the EA cabinets that convert incoming 480 VAC to 800 VDC might have a thermal issue. The Hummer is about the only EV that would press them to their full 350 kW power output in real world charging conditions. Two of the three "breaks" didn't go down to zero kW. I would expect a temporary full shutdown would probably be required for actual pack "switching" (It did go to zero at the 15% SOC switch break). Instead, the charging power just throttled down to about 90 kW, which might indicate either an EA or Hummer thermal issue. As well, the 260 kW session dip didn't zero out and might have been a thermal issue. I wish we had volts/amps data. I'm also wondering if at the 260 kW session break at 30% SOC, if the pack just switched to 400 VDC charging mode so both sub-packs could remain balanced to charge session completion.

It is also very odd about the dip/bumps for the two 350 kW sessions at 55-65% SOC up to about 260 kW. Both dips went to the exact same 84 kW, then stepped up to exactly 180 kW, then up to 250-270 kW in the same pattern. I guess we could say the Hummer's BMS is in that awkward adolescent stage, subject to random mood swings. We can only hope it soon grows out of it

Again, it is time for GM to step up and explain to the Hummer world why the Hummer pack appears to have fast-charging ADD.

 

[azbill]

I noticed the flat spots of 85 and 180kw as well, but the pack switch seem to take it all the way to zero, the orange line at 15% and the gray line at 17%. I think the pack switching and the flat reductions are two different things.

It is well known that the EA ABB chargers can sometimes fail and output only ~40kw for 400V (125A limit at 360V nominal), so the same current limit would be ~80kw for an 800V system. This can happen when the temperature sensor in the CCS connector is failed or indicating an overheat situation.

 

[Nebula1701]