diy solar

diy solar

Valence XP Super Thread

Have you checked the balance of internal cells? It may be, if the battery were kept instorage for too long that internal unbalance prevent the battery to deliver full power...
If you have screenshot of the Valence software page it could help.
Two batteries read outs where posted about a week ago.
 
I remember hearing about your batteries that weren't working right. I'm curious what their history is? Did you get them directly from a seller who got them from their original installation? Or did you get them from somebody In between that person and you who may have abused them?
Hi Tavis, I ordered them off eBay 2019 came form San Diego. Just got around to to hooking everything up in Dec. Th 2000W inverter would shut down on start up. Found the negative cable on the battery using one lug was not tight,(bolt to long...saw your post of this) added copper washers. Now under a 34% load ~60amps runs OK, but bump the load to ~130amps, batteries voltage drops quickly below 12V, inverter shuts down. I did try my 2 lead/acid deep cycle batteries they preformed worst with shut down at even lower amps. The microwave 1,200watt (~160amps) shuts down the inverter within mins. Cables and fuses getting hot, so I am going up to 2/0 cables and higher quality fuses, and order a hydraulic crimper. My hand crimper goes only to 1/0 and even though they seem tight and multiple meter shows no resistance, just not confident I have the best crimps. And the cable are only about 3.5' runs + and -. I ordered the load test should be in this week, will test both Valences to see how they preform. But back too your question; possible they came damaged, or its possible the inverter/charger might have over charged them?(Fast charge setting 14.6V/ 65amp max /floats at 13.7V) And its possible the 110V heat I have been running the in the RV could have shut off and the inverter/charger tried to charge them. My solar control has temp cut off, inverter does not, why I have been running heat in the RV. I am planning to build my own battery pack with 4x 280 3.2 cells with 250/300amp BMS controls and heating pads. Batteries take a month maybe longer from China. Now I wished I had gotten a 3000watt inverter, but the 2000watt does pull less a load on stand by. Just have to limit what is turned on when I turn on a higher load. Inverter can start the Gen. but I have to get the kit from Onan for auto start. ($300). This upgrade of the RV is getting expensive!
 
The above description sounds to me ALOT like bad cables. Put a shunt between the invertor and cables, and then compare with the shunt at the cable-to- battery ends, and see if you get a current spike at the invertor end. I think your cables are bleh. I'll suspect bad crimps from this angle. I have a 190amp load on my 12v and everything is cold.
 
Hi guys,

I have a pack of 22 batteries with the black enclosures (v1 I think), I got them hooked to the Valence BMS on 20th of January and took status right before doing so. Then I took another one on 24th and then on 6th of Feb.
The pack is rewired into 5 48V groups and 2 batteries not wired, just as part of the comms network (as the BMS expects 22).
I have also changed the mode of work to "Backup" i.e. the batteries are not supposed to be discharged all the time and then charged as they were before for a while (stupid I know).
What can be seen is that the cell spread has improved a lot, but the SoC of the batteries and the battery modules voltage varies significantly (with SoC drifting further apart).
The charge voltage (the hybrid inverter is charging the pack) is 55.2V per group i.e. 13.8 per module at present.
What would you recommend that I do to get them to balance, obviously the BMS itself won't suffice. Should I charge them to a particular voltage battery by battery? What charger should I use for this, any recommendations ?
Should I be worried by these numbers, maybe is the 1st and most important question :)

Thanks all !


20.1.202124.1.20216.2.2021
Batt. NGroupU, mVCell Spread, mVSoC, %Batt. NGroupU, mVCell Spread, mVSoC, %Batt. NGroupU, mVCell Spread, mVSoC, %
111393514497.25111386336100111384530100
22136926281.96221370232100221373132100
331376911280.78331372031100331369324100
44135454976.8644135443775.2944135633957.25
54136924199.61541369128100541374036100
63134231080.3963134913180.7863135394679.22
73136132780.3973135862178.8273135822154.9
8113396381.188113415482.358113415487.84
9313409581.579313427479.229313443450.2
104134571876.86104134622173.73104134612145.88
11513415876.47115134171072.9411513416923.53
125135594777.25125135652676.8612513594261.57
135135996877.25135135703376.86135135821764.71
14213397375.6914213404476.08142134431472.55
15-13343675.6915-13341772.9415-13336745.88
16113432777.65161134521878.82161134663580.78
172136317179.61172135881678.82172135681562.35
185135818878.04185135713281.18185135841299.61
19-13349579.6119-13346478.4319-13335575.69
20213403476.4720213409373.7320213405244.71
21413426377.2521413431575.2921413441949.41
22113398377.2522113410677.25221134382267.06
 
Hi guys,

I have a pack of 22 batteries with the black enclosures (v1 I think), I got them hooked to the Valence BMS on 20th of January and took status right before doing so. Then I took another one on 24th and then on 6th of Feb.
The pack is rewired into 5 48V groups and 2 batteries not wired, just as part of the comms network (as the BMS expects 22).
I have also changed the mode of work to "Backup" i.e. the batteries are not supposed to be discharged all the time and then charged as they were before for a while (stupid I know).
What can be seen is that the cell spread has improved a lot, but the SoC of the batteries and the battery modules voltage varies significantly (with SoC drifting further apart).
The charge voltage (the hybrid inverter is charging the pack) is 55.2V per group i.e. 13.8 per module at present.
What would you recommend that I do to get them to balance, obviously the BMS itself won't suffice. Should I charge them to a particular voltage battery by battery? What charger should I use for this, any recommendations ?
Should I be worried by these numbers, maybe is the 1st and most important question :)

Thanks all !


20.1.202124.1.20216.2.2021
Batt. NGroupU, mVCell Spread, mVSoC, %Batt. NGroupU, mVCell Spread, mVSoC, %Batt. NGroupU, mVCell Spread, mVSoC, %
111393514497.25111386336100111384530100
22136926281.96221370232100221373132100
331376911280.78331372031100331369324100
44135454976.8644135443775.2944135633957.25
54136924199.61541369128100541374036100
63134231080.3963134913180.7863135394679.22
73136132780.3973135862178.8273135822154.9
8113396381.188113415482.358113415487.84
9313409581.579313427479.229313443450.2
104134571876.86104134622173.73104134612145.88
11513415876.47115134171072.9411513416923.53
125135594777.25125135652676.8612513594261.57
135135996877.25135135703376.86135135821764.71
14213397375.6914213404476.08142134431472.55
15-13343675.6915-13341772.9415-13336745.88
16113432777.65161134521878.82161134663580.78
172136317179.61172135881678.82172135681562.35
185135818878.04185135713281.18185135841299.61
19-13349579.6119-13346478.4319-13335575.69
20213403476.4720213409373.7320213405244.71
21413426377.2521413431575.2921413441949.41
22113398377.2522113410677.25221134382267.06
The bleed off resistors are very very tiny. It could take weeks or months in some cases of holding them at 99% For them to come to balance. Any battery that suddenly jumps to 100% has finally reached balanced according to its internal computer. You may see a battery at 20% suddenly move to 100. If it's taking too long and you get impatient you can take the side of the battery cover off and inject 3.8 V into the low cell.

I haven't had any luck flashing different battery configurations into the older BMS with round communication cable. Since there is a reason valance wants to know the string Total number in each String I suppose it could do a better job keeping you balanced if it were programmed properly. Unfortunately I've only figured out the newer bms.

Your state of charge mismatch Is probably faults. The amp metres are not accurate and so the state of charge becomes inaccurate over time. Getting them fully balanced will pull The state of charge metres back together.

. For the easy but slow way of balancing adjust your float voltage so that the highest cell is 3.8 v. Occasionally check on it and you'll have to raise the float voltage a little to maintain the 3.8. Basically this maintains the maximum potential balance speed. When all state of charge metres read when 100 or have read 100 at any point in this stage you are finished.
 
Thank you Travis, and everyone reading this :)
I will raise the float voltage and observe.
 
"I have a pack of 22 batteries with the black enclosures (v1 I think)..."

Sounds like a Sainsburys/Smiths EVs-converted Ford Transit fiasco battery pack...? Did you get them from Martin in Sheffield?

I would think your U-BMS is not working due to it still being configured for 22 series modules and therefore not balancing them 'inter-module'-ly.

I have 4P4S pack of the same V1 modules but I'm still in the process of wiring them up. I am going to use one of these on each S string... http://www.deligreenpower.com/deligreen-products/130-en.html#ad-image-0

As a matter of interest, I did a couple of full charge/discharge cycles on one of the 100 or so modules I have and it came out at 137Ah (11.5V - 14.3V @ 15A charge/7A discharge). Amazing considering they all sat un-loved in a van in a yard for about 6 years before all 50 vans were sold at a salvage auction.
 
Has anyone written an Excel 'thing' to graphically display the standard Valence internal BMS data recordings - the interesting stuff at least?
 
Hi guys,

I have a pack of 22 batteries with the black enclosures (v1 I think), I got them hooked to the Valence BMS on 20th of January and took status right before doing so. Then I took another one on 24th and then on 6th of Feb.
The pack is rewired into 5 48V groups and 2 batteries not wired, just as part of the comms network (as the BMS expects 22).
I have also changed the mode of work to "Backup" i.e. the batteries are not supposed to be discharged all the time and then charged as they were before for a while (stupid I know).
What can be seen is that the cell spread has improved a lot, but the SoC of the batteries and the battery modules voltage varies significantly (with SoC drifting further apart).
The charge voltage (the hybrid inverter is charging the pack) is 55.2V per group i.e. 13.8 per module at present.
What would you recommend that I do to get them to balance, obviously the BMS itself won't suffice. Should I charge them to a particular voltage battery by battery? What charger should I use for this, any recommendations ?
Should I be worried by these numbers, maybe is the 1st and most important question :)

Thanks all !


20.1.202124.1.20216.2.2021
Batt. NGroupU, mVCell Spread, mVSoC, %Batt. NGroupU, mVCell Spread, mVSoC, %Batt. NGroupU, mVCell Spread, mVSoC, %
111393514497.25111386336100111384530100
22136926281.96221370232100221373132100
331376911280.78331372031100331369324100
44135454976.8644135443775.2944135633957.25
54136924199.61541369128100541374036100
63134231080.3963134913180.7863135394679.22
73136132780.3973135862178.8273135822154.9
8113396381.188113415482.358113415487.84
9313409581.579313427479.229313443450.2
104134571876.86104134622173.73104134612145.88
11513415876.47115134171072.9411513416923.53
125135594777.25125135652676.8612513594261.57
135135996877.25135135703376.86135135821764.71
14213397375.6914213404476.08142134431472.55
15-13343675.6915-13341772.9415-13336745.88
16113432777.65161134521878.82161134663580.78
172136317179.61172135881678.82172135681562.35
185135818878.04185135713281.18185135841299.61
19-13349579.6119-13346478.4319-13335575.69
20213403476.4720213409373.7320213405244.71
21413426377.2521413431575.2921413441949.41
22113398377.2522113410677.25221134382267.06
I had an interesting problem balancing module to module this week. My valence branded BMS is set up for nearly 40 batteries is series which is also how I have my battery's arranged. I have a small LED hooked to one of them. It Took a long time but its voltage Eventually dropped far enough for the BMS to take notice. For the last 6 months the bms as activated every bleed off resistor in the module balancing circuits except for the battery that was low because of the LED. So all this time none of those batteries were balancing in relation to each other. They were only balancing in relation to the one low battery. I installed 3000 more Watts of Solar and for the 1st time in 6 months I reached full charge and one of the batteries tried to go over voltage & the BMS shut it all down.

So that got me worried about your set up. You're module to module Balancing is not going to work properly when you have 2 batteries that are not going up in down in state of charge synchronously with the rest.

Now this reminds me of a special kind of battery balancer I found a while ago. What it did was took an extra battery that wasn't tied into your Ceres parallel circuit. It used that battery as a sort of Capacitor. It tries to keep all the other batteries balanced to it. So it ends up rising and falling Together with the other batteries. This may be the only way you could properly use the valance BMS To be in combination with this other balancer. Because I haven't met anyone who knows how to program the version one bms. I am not sure if it was just a prototype or if it was ever in mass production.

Honestly your probably better off using a Computer with the 485 adapter to keep the block's balanced In combination with some of the cheap Amazon balancers.

And also in combination with a Cheap bms that will cut the relay to the battery bank if the total battery banks voltage goes out of spec. https://www.ebay.com/itm/Wireless-0...att-Power-Combo-DC-Meter-Ammeter/352638402761.
 
Has anyone written an Excel 'thing' to graphically display the standard Valence internal BMS data recordings - the interesting stuff at least?
Yes but he is using grafana. I got his system and I love it. Im learning how to make graphs just how I like them presented. And I can check on my system with my phone when I am not at home.
 
Hi martinwinlow, yes I got them from Martin, they were posted in the DIY EV forum couple of years back :) A throwback to the times we used to shop with pallets from the UK...
Hi Travis, thanks for the suggestion, I will be getting these balancers, 5 for 200$ straight from China - not bad. About time to get my setup in order and maximize its use :)
If I keep the Valence BMS with 20 modules only will it still make the cells balance?
I still haven't had the time and motivation to setup the arduino BMS by Seb303, I just might after I will be using standalone balancers for each series anyway. This would give me better observability too.
 
I had an interesting problem balancing module to module this week. My valence branded BMS is set up for nearly 40 batteries is series which is also how I have my battery's arranged. I have a small LED hooked to one of them. It Took a long time but its voltage Eventually dropped far enough for the BMS to take notice. For the last 6 months the bms as activated every bleed off resistor in the module balancing circuits except for the battery that was low because of the LED. So all this time none of those batteries were balancing in relation to each other. They were only balancing in relation to the one low battery. I installed 3000 more Watts of Solar and for the 1st time in 6 months I reached full charge and one of the batteries tried to go over voltage & the BMS shut it all down.

So that got me worried about your set up. You're module to module Balancing is not going to work properly when you have 2 batteries that are not going up in down in state of charge synchronously with the rest.

Now this reminds me of a special kind of battery balancer I found a while ago. What it did was took an extra battery that wasn't tied into your Ceres parallel circuit. It used that battery as a sort of Capacitor. It tries to keep all the other batteries balanced to it. So it ends up rising and falling Together with the other batteries. This may be the only way you could properly use the valance BMS To be in combination with this other balancer. Because I haven't met anyone who knows how to program the version one bms. I am not sure if it was just a prototype or if it was ever in mass production.

Honestly your probably better off using a Computer with the 485 adapter to keep the block's balanced In combination with some of the cheap Amazon balancers.

And also in combination with a Cheap bms that will cut the relay to the battery bank if the total battery banks voltage goes out of spec. https://www.ebay.com/itm/Wireless-0...att-Power-Combo-DC-Meter-Ammeter/352638402761.

Yes but he is using grafana. I got his system and I love it. Im learning how to make graphs just how I like them presented. And I can check on my system with my phone when I am not at home.
Wish I knew what I was doing on the coding end of the spectrum to make my thunderstrucks putty communication actually useful. That alone is reason to go with what you guys got cooking. Cheaper then setting up a victron vrm to frankly which is what I was in the process of doing.
 
Wish I knew what I was doing on the coding end of the spectrum to make my thunderstrucks putty communication actually useful. That alone is reason to go with what you guys got cooking. Cheaper then setting up a victron vrm to frankly which is what I was in the process of doing.
Yes it will make the cells balance just fine, no problem there
 
The black circles on the side of the u27-12xp reveal screws. You can remove the side cover and you'll have access to the individual wires going to the cell blocks so you could actually hook up a traditional cell balancer like a Batrium or many other far cheaper than Batrium choices. But if you have more than a few batteries that would be a hell of a lot of wires and I'd recommend getting the actual BMS designed for these.

Photo courtesy of http://cogito44.free.fr/technologie.html#wa-anchor-jrc1z2om1dqjuxl91c
Hi Travis, thank you for this thread! I just bought 2 U27-12XP’s and am wondering if I can add a Valence RT series and use the more robust onboard BMS from it to connect to the XP’s? Has anyone tried this? The comm connections look to be the same.

Also, do you have any experience with the Thunderstruck VCI BMS. As it looks like valence LV BMS is almost impossible to find.
 
Hi Travis, thank you for this thread! I just bought 2 U27-12XP’s and am wondering if I can add a Valence RT series and use the more robust onboard BMS from it to connect to the XP’s? Has anyone tried this? The comm connections look to be the same.

Also, do you have any experience with the Thunderstruck VCI BMS. As it looks like valence LV BMS is almost impossible to find.
I doubt it will do anything other than confuse the rt. Maybe one benefit it could cause the XP To Flash every 5 seconds instead of 20 indicating that its on board balancing is activated. Of course this won't balance battery to battery on the XP side. And I would worry that the battery to battery balancing on the RT side could get confused and do some bad things. In order to find out you have to feel the bleed off resistors for heat indicating there active in boTh xp & RT and check the voltage of the cells And modules over time to make sure that things aren't behaving in a way detrimental to anything.

This brings up an interesting idea that maybe somebody could activate the cell balancing of a 12V XP battery bank by Utilising a board from a single RT .
 
Last edited:
Hi Travis, .
do you have any experience with the Thunderstruck VCI BMS. As it looks like valence LV BMS is almost impossible to find.
No I've never tried the thunderstruck and don't have any experience with it other than seeing people happy with it on 12 V systems. Don't forget I convert HV units to LV and Flash them properly for your battery bank configuration.
 
I doubt it will do anything other than confuse the rt. Maybe one benefit it could cause the XP To Flash every 5 seconds instead of 20 indicating that its on board balancing is activated. Of course this won't balance battery to battery on the XP side. And I would worry that the battery to battery balancing on the RT side could get confused and do some bad things. In order to find out you have to feel the bleed off resistors in the RT and check the voltage of the cells And modules over time to make sure that things Aren't behaving in a way detrimental to anything.

This brings up an interesting idea that maybe somebody could activate the cell balancing of a 12V XP battery bank by Utilising a board from a single RT .
I agree. It may be worth a test. Would you place it in the center of the 3 batteries or at one end?
 
No I've never tried the thunderstruck and don't have any experience with it other than seeing people happy with it on 12 V systems. Don't forget I convert HV units to LV and Flash them properly for your battery bank configuration.
Do you sell the LV versions? I’m definitely in the market.
 
Back
Top