Byd powerwall qnbbm unbalance wave
- Thread starter Mtour
- Start date
OffGridInTheCity
Moderator
- Joined
- Dec 15, 2018
- Messages
- 2,710
Nice set of batteries. Envious (in a good way) 
Is this one big 24vbattery bank?The 8s BattGo picture implies they are all parallel'edbut just curious as to how you have them hooked up if you care to share.
>I have noticed a wave pattern on my voltage checker. Anyone have any ideas
Maybe you could explain a bit more. I'm not getting what you're asking. Do you mean the screen of the BattGo is wavy? or voltages are going up/down?
Is this one big 24vbattery bank?The 8s BattGo picture implies they are all parallel'edbut just curious as to how you have them hooked up if you care to share.
>I have noticed a wave pattern on my voltage checker. Anyone have any ideas
Maybe you could explain a bit more. I'm not getting what you're asking. Do you mean the screen of the BattGo is wavy? or voltages are going up/down?
Thanks for the reply..
The powerwall is 2s12p 48v..the packs are highly used byd 8s 5.6kw packs, only getting half the capacity of new. I have interconnected the cells with 16g and fused each interconnect with a 10a fuse.
I have been researching the batrium bms system, really nice setup. From what I can see I would need a wm4 and 192 longmons, or could I get by with the basic 16s setup hooked up to the interconnect leads.
Thks
The powerwall is 2s12p 48v..the packs are highly used byd 8s 5.6kw packs, only getting half the capacity of new. I have interconnected the cells with 16g and fused each interconnect with a 10a fuse.
I have been researching the batrium bms system, really nice setup. From what I can see I would need a wm4 and 192 longmons, or could I get by with the basic 16s setup hooked up to the interconnect leads.
Thks
OffGridInTheCity
Moderator
- Joined
- Dec 15, 2018
- Messages
- 2,710
Haven't answered your original question and I'm still not sure how to help. Batrium is a great choice but you won't need192 longmons - you'd go broke I use Batrium - and you need 1 longmon for each unit in series. You have 24units - and if you hook the up as 2s12p (e.g. 2s units = 16 cellsin series for 48v) and 12p (12 units = 12 cellsin parallel) then you would need 16 longmons - one for each group of cells in parallel. A Watchmon4 withlongmons is capable of balancing on the order of500ah cellsand I believe that 12 units (cells) in parallel is still below that threshold. Just by example -If you hooked make 2 separate batteries of 2s6p - then you'd need 32longmons as you'd have groups of 6 cells in parallel.
Maybe if you take a look at @DavidPoz youtubes it will help. He recentlyinstalled 8 of your exact modules and did several youtubes. Here's an example where he talks about Serial and Parallel'ing his setup -
If you look (timewise) right before and after this youtube you'll others detailing his whole hookup.
I use Batrium - and you need 1 longmon for each unit in series. You have 24units - and if you hook the up as 2s12p (e.g. 2s units = 16 cellsin series for 48v) and 12p (12 units = 12 cellsin parallel) then you would need 16 longmons - one for each group of cells in parallel. A Watchmon4 withlongmons is capable of balancing on the order of500ah cellsand I believe that 12 units (cells) in parallel is still below that threshold. Just by example -If you hooked make 2 separate batteries of 2s6p - then you'd need 32longmons as you'd have groups of 6 cells in parallel. Maybe if you take a look at @DavidPoz youtubes it will help. He recentlyinstalled 8 of your exact modules and did several youtubes. Here's an example where he talks about Serial and Parallel'ing his setup -
Thanks for the info on the longmons, I think the issue I am having is voltage fluctuations on the bus bars, not so much the interconnects at higher amperage. I measured a voltage drop of 70mv across the 1/2 refridgeration soft copper bus at 60amps. I am feeding the bus + & - from opposite ends of the bus..
not2bme
Member
- Joined
- Oct 16, 2017
- Messages
- 517
Mtour said:
You want to take your voltage at rest to be accurate. Trying to take measurements while the battery is being charged or under load will skew your voltages if your bus bars aren't up to the amperage. I noticed that with mine.
But on the other hand if indeed your voltages are off then it's possible that those byd batteries are shot. The ones I've seen of davidpoz seems like some are swelling to the point that it probably has a high resistance and just a drain on the whole system. If your cells are particularly bad, having a parallel cable will not help much. A bms will not help either, not even with a batrium, you end up having to constantly discharge the packs just to maintain balance. The only way is to dremel off each cell and take out any bad cells and rearrange each packs then soldering them together. It's going to be a chore but the only way to solve this issue.
I have 26 of the byd pack, 208 cells.. One cell out of the 208 is down to 66ah capacity, the rest are between 110-145ah capacity. I was getting 1200ah in a 16s12p configuration, with the interconnects I am now seeing 1350ah usable 3.36v-2.9v. I ran them without a bms or balancer for the first three month without issue. I like these byd's, they have had a hard life, but I believe they are a quality battery that in my application could last a long time, max charging and discharging at .09C..
not2bme
Member
- Joined
- Oct 16, 2017
- Messages
- 517
Your cells are already 60% of it's rated output and if not properly matched will always cause an imbalance over time. Now that being said, 51mV isn't that bad. What's more important is how fast it goes out of balance. If you said that if you don't use any balancers, would get worse by 10mV every day then your packs are badly out of balance? 51mV isn't bad and it also depends on the SoC. My batteries go from anywhere from 5mV to 25mV in a single day depending on the charge level of my batteries.
About the wave pattern, is it possible you are looking at thevoltage monitor when you have the qnbbm plugged in? If so you could be seeing the qnbbm at work transferring from the lowest celll to the highest. Usually seen as a blip on any voltage monitor.
About the qnbbm, it actually works best if over 100-300mV to achieve 1A transfer. The greater the difference the better the transfer. At 50mV it may only transfer at 100mA. It's possible you can try other active balancers such as this https://www.aliexpress.com/item/4000252216391.html work differently by storing the energy into a supercap and thus give the 1-2A transfer no matter the difference in voltage (I have this board currently and actually want to test the real actual rate of transfer).
About getting the best balance for the pack, you need to find the sweet spot of your pack. The capacity of your pack is determined by the worst cell of your pack. So try not to use your pack to the max. In my example I'm going to use 4.2V as max and 3.5v as min since I've never had LFE batteries. 4.2V is the max so I would never use it at the max. In fact you get better life expectancy if you only reach about 90%. So I would keep my batteries at 4.1V instead of 4.2V. Then I would top balance at 4.1V. Then I would drain my pack until you see the packs start to deviate by 50mV or more. This number may be different for LFE since your working range and voltage curve is less linear. But what you want to try and avoid is having one cell charged too high, as that just generates heat since it's already charged, and that causes this loss and thus energy being lost and creating the imbalance. You notice this when it's charged towards the high end, which is nice when you try to cap your top end at 90% or less so there's room for any imbalance.
About the wave pattern, is it possible you are looking at thevoltage monitor when you have the qnbbm plugged in? If so you could be seeing the qnbbm at work transferring from the lowest celll to the highest. Usually seen as a blip on any voltage monitor.
About the qnbbm, it actually works best if over 100-300mV to achieve 1A transfer. The greater the difference the better the transfer. At 50mV it may only transfer at 100mA. It's possible you can try other active balancers such as this https://www.aliexpress.com/item/4000252216391.html work differently by storing the energy into a supercap and thus give the 1-2A transfer no matter the difference in voltage (I have this board currently and actually want to test the real actual rate of transfer).
About getting the best balance for the pack, you need to find the sweet spot of your pack. The capacity of your pack is determined by the worst cell of your pack. So try not to use your pack to the max. In my example I'm going to use 4.2V as max and 3.5v as min since I've never had LFE batteries. 4.2V is the max so I would never use it at the max. In fact you get better life expectancy if you only reach about 90%. So I would keep my batteries at 4.1V instead of 4.2V. Then I would top balance at 4.1V. Then I would drain my pack until you see the packs start to deviate by 50mV or more. This number may be different for LFE since your working range and voltage curve is less linear. But what you want to try and avoid is having one cell charged too high, as that just generates heat since it's already charged, and that causes this loss and thus energy being lost and creating the imbalance. You notice this when it's charged towards the high end, which is nice when you try to cap your top end at 90% or less so there's room for any imbalance.
not2bme said:
Au contraire, for LiFePO4 a 51mV imbalance between cells is really bad, possibly over a 40% capacity difference by below. This is because the LiFePO4 voltage profile is much flatter than for regular Li-ion, so a small change in voltage corresponds to a much larger change in capacity (which makes it much more difficult to design good fuel gauges for them). The data is from here and here. Divide voltage by 4 in the graphs to get per-cell values.
I was too quick to blame the cause of the wave pattern on the qnbbm's, they were added the same day the i finished the battery cell interconnects. Didn't cross my mind that the interconnects could cause the wave pattern in the balancing..
After doing a complete discharge (1350ah out) down to 3v the batteries stayed within 20-50mv until the end, then drifted to 70mv, no wave pattern at all. The wave pattern started to form right away on recharging the battery bank on 200amps of solar. I do not think the little amount of balancing the qnbbm's provide could sway the balance with 200amps going into them, 16amps each . After the batteries are finished absorption the wave pattern goes away.. most balancingI have measured from the qnbbm's is 550ma..
After doing a complete discharge (1350ah out) down to 3v the batteries stayed within 20-50mv until the end, then drifted to 70mv, no wave pattern at all. The wave pattern started to form right away on recharging the battery bank on 200amps of solar. I do not think the little amount of balancing the qnbbm's provide could sway the balance with 200amps going into them, 16amps each . After the batteries are finished absorption the wave pattern goes away.. most balancingI have measured from the qnbbm's is 550ma..