Pylon / Lux oddity ….

[Lincs Robert]

Hi,

Ok, so as a recap, my config WAS 5 x Pylon US2000 with a Lux ACS3600 - all duplicated so therefore able to charge/discharge @ 7200watts as arranged in master/slave config. Each Lux unit could “see” 250Ah of batteries as was confirmed on the LCD screen.

In order to fund my Chinese stacks, still in transit, I then sold 4 of the Pylons. This would reduce my capacity in the short term accepted. So now, I have 3 x Pylons with each of the 2 Lux units and all’s working - albeit with an issue.

The reduction in overall capacity coincided with the start of the cold snap, and the location of the kit is currently sitting around the 3 to 4 degree C mark.

The issue that has arisen is that my charge/discharge rate per Lux unit seems to have dropped to 2400, or about 4800W total. The Lux units show 150Ah of battery capacity each - so why the reduction in charge/discharge rate, which was 3600 per unit? Is the issue caused by the reduction in temperature, or is there some other problem?

My new cells are still some weeks away, so I’d like to get to the bottom of the charge/discharge issue.

So, any ideas much appreciated.

Cheers- Rob

[Stinsy]

When Pylontech batteries get cold they reduce the charge rate to 5A per battery so 3x would make 15A at 48V = 720W.

Usually once they’ve been charging for 20-30mins they warm up a bit and the charge rate increases back to 25A per battery.

[Tinbum]

[Lincs Robert]

Thanks, so on the basis that they are already charging faster then 720watts it obviously isn’t that……….
They are both doing exactly the same - so I have the same problem, but twice. The Lux units are displaying 150Ah each, so that’s correct, for the 3 batteries. Also, the LED indicators on all 3 are showing the same charge levels ……

[Oldgreybeard]

When Pylontech batteries get cold they reduce the charge rate to 5A per battery so 3x would make 15A at 48V = 720W.

Usually once they’ve been charging for 20-30mins they warm up a bit and the charge rate increases back to 25A per battery.

Yep, found this out the hard way the day before yesterday! A not very clearly documented feature, and as as I'd turned the max charge down to 2.5kW, to stay within the safe grid limit, it meant that the pack didn't reach the target SoC during the time window I'd set. Caused a bit of head scratching until I had a look at the history data in HA, and spotted the low initial charge rate.

The reduced charge rate seems to kick in at very roughly 2 deg C from what I could tell, but hard to be sure as I don't think the packs give out very accurate internal temperature readings (mine always seem to be lower than their case temperature, which doesn't seem right to me)

[Tinbum]

The reduced charge rate seems to kick in at very roughly 2 deg C from what I could tell, but hard to be sure as I don't think the packs give out very accurate internal temperature readings (mine always seem to be lower than their case temperature, which doesn't seem right to me)

See my post above for the temperate cut offs and the settings..

I find the battery temp reading goes up dramatically when the cells are balancing and is always slightly above the cell temps. I put it down to th location of the sensor.

[Oldgreybeard]

The reduced charge rate seems to kick in at very roughly 2 deg C from what I could tell, but hard to be sure as I don't think the packs give out very accurate internal temperature readings (mine always seem to be lower than their case temperature, which doesn't seem right to me)

See my post above for the temperate cut offs and the settings..

I find the battery temp reading goes up dramatically when the cells are balancing and is always slightly above the cell temps. I put it down to th location of the sensor.

Very sorry, Tinbum, wasn't wearing my reading specs so missed the table! Just been out and checked the battery shed (bloody cold out, -3 deg C already outside) and the battery cases seem to be sat at about 4 deg C. They are reporting an internal temperature of 2 deg C, but have just started charging at 2.5kW, split across 6 US3000C packs.

As an experiment, as as there is not going to be any appreciable PV tomorrow, I've set the to charge to 100% tonight, as much to get the whole pack warm as anything else. I think the cold soak over the past few days has exceeded the warning effect of having the inverter sat on the wall above them.

Might look at fitting the small tubular heater I'd planned to fit in there a couple of years ago, when I first installed them. I didn't bother with it, as the batteries didn't seem to get that cold last winter. This weeks' been a bit extreme for down here, though. First time in five years I've needed to get the salt out and spread it on the drive.

[Lincs Robert]

SORTED

I finally applied myself to this and got it sorted. Hercule Poirot would have been proud of me.

When I had a single stack of 8 Pylon US2000s and a single Lux unit, it was installed “professionally” - but there was no earth cabling between the batteries, inverters & earth.

When the second Lux was installed, the stack of 8 was split in half and became two stacks of 4. So far so good. I then had a “tidy up”, and earthed everything. Batteries and inverters back to the main earth on the consumer unit - which is fairly close by.

Then I added a new US2000C to each stack, again, earthing everything and all was good.

The fun started when I sold 4 of the US2000s - to fund the Chinese stacks. At the same time I relocated one of the stacks, now of 3, to allow space for the forthcoming Chinese stacks. Due to cable routings, I never got round to earthing the stacks of 3. After all, it had worked well enough before without an earth - hadn’t it?

So each stack of 3 was now a single US2000C and a pair of US2000s. All worked, except for the fact that each would only charge/discharge at 2400Watts. This triggered my original post. The Lux units were reporting seeing 150Ah of batteries & I was at a loss as to what was wrong. The wrong was consistent as each stack with its own Lux unit behaved exactly the same.

I had a ponder and wondered if the earthing may be the issue? It was. The short version is pretty much that two wrongs were making a right, removing one of them then caused the problem as described.

Pylon changed the pinout on the RJ45 from the US2000 to US2000C. They moved the gnd connection! Whereas, with my original, unearthed, config, the earth wasn’t needed as the gnd was via the data cable. Adding the US2000C then saw the earthing go via the hard earth - which was all good until I moved one set of batteries to make space!

I really was (honest) going to add the proper earth at some point, but never realised that it had been providing data grounding continuity. The earthing is now in place and all’s working as it should. BUT, the safety earth is masking a deficiency in signal grounding that I really need to get resolved.

I need to take a look at resolving the data cables and get it done properly so that I’m not relying on the protective earth connection for signal grounding. I’m surprised that it worked at all - although the Pylon manual for the US2000C talks about unreliable BMS operation - which I guess I discovered!

I fully accept it’s my own fault, but I really really don’t understand why Pylon would repin the signal ground in the way they did.

Happy xmas all …..

Rob

[Oldgreybeard]

Good news it's fixed, and even better that we now have all learned from your experience, so we'll understand if we get something like this happen in future.

[Lincs Robert]

Hello again,

This useful video explains all - the cable bit is towards the end, about 17 mins in.



He’s even gone to the trouble of giving the pinouts on how to resolve the problem and also makes a suitable adapter which he has listed on eBay.

https://www.ebay.co.uk/itm/334657045876 ... media=COPY

So, this “solution” is now buried within my tale of woe - do we have the ability to make a “sticky” , as on other forums?

Cheers