SmartSolar 100/20 drops to <10W intermittently for no obvious reasons

Question

question

semlohnhoj asked Jan 22, '23

SmartSolar 100/20 drops to <10W intermittently for no obvious reasons

I've got 5 x SmartSolar 100/20's (Firmware v1.61) with 4x170W panels conected to each. These are charging a 48V Lifepo4 battery bank (2 lots of 4x12.8V). There is also a SmartShunt (Firmware v4.12) in series.This is connected to a Smart RS Solar inverter (Firmware v1.11). The whole system is managed by a CerboGX (3.00~15) to which is connected to all the units via ve.bus and ve.can. All five of the MPPTS and the SmartShunt are also in a Bluetooth ve Smart Network.

Each group of 4 panels has different shading issue however the problem I'm having doesn't correspond to any particular shading or times of day.

The issue I have is that randomly, one or more of the SmartSolar MPPTs will intermittently drop out. Input PV voltage is fine (in fact, goes up a little), input PV current drops to near zero and output current falls to near zero. It will stay there for a random period of time and then corrects itself.

While this is happening, the other MPPT's will continue to work correctly (although sometimes, more than one MPPT will start to do this at the same point but correct itself at different points!).

None of these issues corresponds with temperatures, battery output loads. battery state of charge or anything else I can see. The MPPTs remain in Bulk mode at all times. It does seem to quite often happen when the battery voltage is around 54V (but not always).

At one point, when this was happening, I tripped the breaker between the MPPT and the solar panels and then turned it back on and the MPPT started working again.

I'm open to suggestions really because I'm stumped. My initial thoughts were it could be one of the following:

1) The fact they are in a ve.smart network as well as being managed by the CerboGX
2) The MPPT firmware has a bug

Some screenshots showing a few of the dropouts I have plenty other data if it helps :)

MPPT SmartSolar fault

1674415400169.png (115.8 KiB)

1674415537054.png (144.9 KiB)

1674415799354.png (69.9 KiB)

5

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

seb71 commented · Jan 22 at 07:45 PM

The whole system is managed by a CerboGX (3.00~15) to which is connected to all the units via ve.bus and ve.can. All five of the MPPTS and the SmartShunt are also in a Bluetooth ve Smart Network.

In a system managed by a GX device, VE.Smart Networking should not be used.

Is DVCC enabled in Cerbo GX?

Which device is set as the Battery Monitor?

Are you sure that increasing the loads has no effect?

1 ·

semlohnhoj seb71 commented · Jan 22 at 08:08 PM

Hi, thanks for the reply.

I'll give disabling the ve smart networking a go.

DVCC is disabled.

The Shunt is set as the battery monitor.

The AC output load doesn't seem to correlate with the issue.

0 ·

seb71 semlohnhoj commented · Jan 22 at 08:16 PM

Enable DVCC.

The AC output load doesn't seem to correlate with the issue.

What I meant is to try and see if increasing the load does bring the idling MPPTs back into production.

0 ·

semlohnhoj seb71 commented · Jan 22 at 08:42 PM

Ok. Up until the point they stop, the batteries are absorbing the current the MPPTs are supplying so there's still demand but I assume you mean extra load from the inverter.

0 ·

seb71 semlohnhoj commented · Jan 22 at 08:48 PM

Yes. When you see that drop in power from an MPPT, increase the load in the inverter (with a big load) and observe the results.

The battery is limited in how much power/current it can absorb. This is why the test with a load on the inverter is necessary.

1 ·

Modifications Space

Node Red Space

Answer 1 · 2023-02-09T22:54:05Z

semlohnhoj answered · Feb 09 at 10:54 PM

This issue still happens periodically. Interestingly, it triggered when my electric heater switched on today (in orange in the image below) at 12:18. The heater has however clicked on and off dozens of times during the day before and after with no effect. It dropped out 2 of my SmartSolar MPPTs one of which recovered after a few minutes and the one in the screenshot (MPPT5) took 30 minutes.

Again, no obvious pattern to the fault and no obvious way of getting logs from the SmartSolar to see why it decided to do it?

Can someone from Victron add anything to this question?

MPPT5 does this quite often, MPPT3 and one of the others (I forget which one) do it less often, and the other 2 almost never do it. Is this faulty units or a software bug maybe?

If I can't get a solution to this I may have to engineer a more direct remedy, i.e. inserting a rely in line with the panels to disconnect them for 30 seconds to kick the things back into life which seems like over kill on my part for an issue that should at least be identifiable.

1675982722189.png (114.5 KiB)

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Answer 2 · 2023-02-16T20:17:47Z

semlohnhoj answered · Feb 16 at 08:17 PM

Well, as an update to my issue with the MPPT dropouts, I haven't managed to nail down why it is doing it and I've disabled the Bluretooth network on them and tried applying a load to the inverter when one is doing this with no change to the MPPT.

What I have done is written a NodeRed flow to monitor all 5 of them and if one of them drops out when the other are working then it switched it off for 60 seconds and then turns it back on again. Problem solved, or more correclty, problem masked :)

I'd still like to have a proper fix for this but at least I have a stop gap. I can't understand why nobody else is having the issue which makes me wonder if I have an unusual setup having 5 MPPTs in a system or not having a BMS?

Would anyone recommend having DVCC turned on?

1

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

JohnC ♦ commented · Feb 17 at 05:33 AM

Hi @Semlohnhoj

Yes, use DVCC. It shouldn't hurt to have both SVS and SCS ticked, with the shunt as the source of both of those. Make sure that the mppts all have identical settings, and that the Tail (mppt one) is set low so it won't interfere.

Please don't have anything networked via the Smart bt. You don't need 2 networks, they could interact or throw quirks.

Bear in mind that the mppts will track every 10 minutes if they're pushing to reach a target V. This can take like 10 seconds or more as they swing the panel V several times. Often missed by the VRM sampling times, but can be shown by Victron Connect Trends. So don't get excited if you see odd spikes on the VRM graphs. Shouldn't happen though if V has reached target.

Flick your NodeRed thingy while testing.

1 ·

Answer 3 · 2023-06-14T13:54:49Z

semlohnhoj answered · Jun 14 at 01:54 PM

UPDATE:

I still get this issue and I've got DVCC on, latest firmware everywhere. Been through goodness knows how many revisions on the MPPTs, Shunts and Cerbo. I've even got a Lynx shunt now and that didn't alter anything.

I do have something to add though. The 2 MPPTs that do this most often wher both running at 0 today after a while this morning and flat refused to reset even using my NodeRed script to the point I though the things had fried. I tried the following:

1) Resetting to factory defaults!
2) Turning the charging on/off via bluetooth
3) Flipping the DC isolators
4) Rewiring MPPT 2 & 3 so they had each others panels. The fault still stayed with the MPPT.

I then unplugged the VE.direct cable from MPPT1 while I was monitoring it via bluetooth (no change - 0W). But, when I plugged the VE.Direct cable back in about 20 seconds later it burst back into life!

So, to sanity check this, I did the same with MPPT2 and the same happened!

Now because I use quite a few VE.Direct devices (5xMPPTs, 4xShunts) I have 2x 3 way USB expanders in 2 of the USB ports on the Cerbo so my suspicions pointed there, unfortunately MPPT1 and MPPT2 ironically are the only VE.Direct devices that connect directly to the Cerbo's builtin VE.Direct ports.

I tried unplugging the USB adapters and plugging them back in again and after that MPPT3,4 and 5 all started dong similar things to 1 & 2 to the point where I had to reboot the Cerbo to get them working.

I'll change the USB adapters in the first instance for powered ones which won't be easy but I'm getting a little disillusioned with this to be honest. If there was some sort of debug logging it would help, whether it was me or Victron that could see.

1

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

JohnC ♦ commented · Jun 14 at 02:33 PM

@Semlohnhoj

Victron have always recommended quality, powered usb hub(s) when doing this. I'd expect that to help greatly.

1 ·

question