question

nicholas.nouwens avatar image
nicholas.nouwens asked

RS 450/100 - Up in smoke

Hi there,

Installed a RS450/100 3 months ago. The unit charges 2x14.3kwH lithium batteries which feed a Multi II 48/5000. GX to control hardware/ESS. Sizing of the PV arrays was done using Victron's PV sheet/calculator, and all checked out.

  • First Tracker runs with 8 x 605W Canadian CS7L-605MS (4840W)
  • Second tracker with 6 x CS6W-545MS (3270W)

This past weekend the unit went up in smoke following (appears to be) some of the internal component failure. Fortunately no other damage or fire spread.

The maximum PV input voltage recorded on the unit to date is around 335-340V so well within the 450V nominal voltage,

I know that charge controllers normally fail due to poor terminations on the DC or PV terminals, but this is not the case here. Unit is in a well ventilated area, and it is winter at the moment, so the unit could not have been operating very hot.

I've actioned a warranty claim with my reseller, but I remain concerned as to why this could have happened in the first place.

Anything that I'm missing?

Thanks,

Nick

img-1942.jpgimg-1938.jpg




MPPT Controllers
img-1942.jpg (700.6 KiB)
img-1938.jpg (815.1 KiB)
5 comments
2 |3000

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

allen-dengler avatar image allen-dengler commented ·

img-20240422-wa0023.jpgIn the past 2 weeks I have had 2 RS450 /100 Mppts go up in smoke, so bad that in both cases a fire extinguisher had to be used. One installation is just on 2 years old and the other is just over a year old. The first one unfortunately was offline so I can't see the fault code, the second ones fault code is 201. Both systems have 8x 545watt JA Solar panels per string so the voltages (VOC) is under 400V. One has been returned and I did the RMA for the second one today. The one is HQ2138 the other is HQ2139. img-20240417-132604.jpg

1 Like 1 ·
nickdb avatar image nickdb ♦♦ allen-dengler commented ·

There were issues with early models, RMA these, all those issues are resolved in new units. Both are from a similar manufacturer period.

1 Like 1 ·
diederik avatar image diederik nickdb ♦♦ commented ·
I have the same device.

How can I tell if it is an early model?


0 Likes 0 ·
janieronen avatar image janieronen commented ·
Were there package damage as top of case looks bent? Sure hard to see if not opened top part of case before it burned.
0 Likes 0 ·
nicholas.nouwens avatar image nicholas.nouwens janieronen commented ·
@JaniEronen I recall that it is bent on both sides after removing the cover. I don’t think this was damaged before installation. Will share photo when I strip the unit out.
1 Like 1 ·
5 Answers
nickdb avatar image
nickdb answered ·

Can you share the serial number?

1 comment
2 |3000

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

nicholas.nouwens avatar image nicholas.nouwens commented ·

@nickdb No Problem - HQ22496NVTC

0 Likes 0 ·
nicholas.nouwens avatar image
nicholas.nouwens answered ·

So, 3-4 weeks on, the unit has been returned to the local distributor. Unit has been replaced by Victron under warranty.

The shortcomings on the RS450-100A are known, affecting units with serial numbers HQ2216 up to and including HQ2236. These units would typically fail by displaying a "E200" error, reporting a DC Voltage failure. There is a replacement scheme currently in place for this, with a default warranty extension offered by Victron on units already in the field.

However, with Serial number HQ2249, the smoky RS450 in question is outside the "problem batch" and seems to have displayed a unique mode of failure.

There hasn't been any feedback from Victron Support on my unit, nor any inquiry into the specifics of the installation.

I appreciate that the unit has been replaced under warranty, but some further insight as to what might have caused the unit to catastrophically fail (i.e an inherent fire risk) would be both helpful and appreciated.


3 comments
2 |3000

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

Hi @nicholas.nouwens

I saw when you first posted, and forwarded it on internally. Apologies for the failure, and the hassles, and for not responding directly to you then, an internal communication misstep over who would respond.


You're on the right track returning the unit via the RMA process.


Given the way it failed, an RS engineer will get notified, and depending on the details may request that the unit is shipped back to R&D in The Netherlands for further investigation.


You don't need to do anything related to this, and it's rare that there is any extra info required from you. The machine can usually do most of the talking.


As for insights as to why, it can vary and we could only speculate at this stage. Unlike with the Error 200 within that batch range, there is no specific guidance or pattern of failure for what happened here. I have seen catastrophic failure from environmental factors like vermin ingress and copper swarf from bridging two contacts on the PCB, or perhaps there is something else internally. Until an engineer can investigate, and detect a pattern it's only speculation.


With that said, I am confident enough in the design that I run MPPT RS 450/100 and 450/200 in my own home, but also cautious enough of all high power electronics that I have a smoke detectors, fire extinguishers, and fire blankets installed near by.

0 Likes 0 ·
nicholas.nouwens avatar image nicholas.nouwens Guy Stewart (Victron Community Manager) ♦♦ commented ·

Hello Guy,


Thank you for taking the time to respond.


I've been advised that the faulty unit has subsequently been returned to the Netherlands for inspection.


As what you may expect, I have some concern that there may be an inherent fault in my installation which may have caused the failure - hence the push for feedback.


I agree that any premature conclusion on the reason for the fault would be speculative at this point in time.


I await further feedback from Victron's support office.



0 Likes 0 ·
nickdb avatar image nickdb ♦♦ nicholas.nouwens commented ·
Hi, it is unlikely that an external fault could have caused this, the system will generally trigger alarms if that is the case. I would not worry too much, but it never hurts to check everything.


As an aside, like Guy, I installed smart smoke detectors in the garage and other areas so if any magic smoke escapes I am notified on the phone.

0 Likes 0 ·
James Pitstock avatar image
James Pitstock answered ·
  • 1000008561.jpgFor record my RS 450/100 went up in smoke after 2 years of solid operation. The unit was still powered up when I got home with error #200 internal error. The fault occurred on sun rise. A lot of black residue covered the surrounding surface from the fault. My unit serial # is HQ2144.
  • 1000008558.jpg
  • 1000008556.jpg

1000008556.jpg (3.1 MiB)
1000008561.jpg (2.7 MiB)
1000008558.jpg (3.2 MiB)
2 comments
2 |3000

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

klim8skeptic avatar image klim8skeptic ♦ commented ·

@James Pitstock Is there any recommendations regarding the use of crimps with the PV input terminal block?

ss-rs.png


1 Like 1 ·
ss-rs.png (1.9 MiB)
James Pitstock avatar image James Pitstock klim8skeptic ♦ commented ·
Hi, nothing recommended from the installation manual. The installation picture did show the use of yellow ferrules. It is very common to use a ferrule with fine stranded small gauge wires. A good square ferrule has a good contact surface with the terminal. The terminal on the MPPT would also clamp bare stripped PV spec wire well too if carefully inserted. Either way both methods done correctly will provide a low resistant connection.
0 Likes 0 ·
Alex Pescaru avatar image
Alex Pescaru answered ·

Hi All,

It seems that all units are damaged on the MPPTs side.

And all had arcing damage from the DC high voltages, because all have strings towards the upper end of the voltage limit.

Fact that I've observed even on my Multi RS with similar MPPT topology, is that the soldering process is done by a machine, without a real human to additionally resolder the high voltage, high frequency components and inductors.

An automatic wave soldering leaves very little solder on the PCB. During time and temperature variations, that very little solder could crack, especially because we are talking here about switching inductors that generate microscopic high frequency vibrations.

Such a crack or any imperfect contact will lead to microscopic arcing that further melt the solder and soon became catastrophic failures as seen.

So time to think about things.......

Like a coincidence, just talking with @Alexandra about HF things which are 2-3 years old since development, like these units....

Maybe @Guy Stewart (Victron Community Manager) can please tell us about the conclusions for that failed unit last year that returned to R&D?...

Thank you,

Alex

3 comments
2 |3000

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

Alexandra avatar image Alexandra ♦ commented ·

The production process and some other items were changed since the early ones. I think you will understand oif not all the details are shared. It is interesting that is was some units not all from that time.

But as you can see, to give good warranty back up means being able to support the customers you already have.

0 Likes 0 ·
Alex Pescaru avatar image Alex Pescaru Alexandra ♦ commented ·

Sure I understand that not all details are shared, the important thing is to learn from mistakes and to correct them. And to go to school and learn how to look for details... :-)))

Problem with warranty in my country is that the law says max 2 years. Try to say to a seller after 2 years to replace the unit because Victron says 5 years and he will laugh you in the face.

Maybe some production processes have changed, but for sure the soldering one has not. My unit from mid 2023 is having soldering deficiencies.

Take a look at the picture below and see for yourself. Click on the picture name, below the picture, for a larger size. For sure some will say that for those FETs the soldering is enough and maybe they are right for those transistors, even if up to 25A are passing through the points highlighted, but rest assured, for an induction coil passing the same current and vibrating at 30kHz or more is NOT.

Also take a look at the smear of silicone on the left capacitor. That silicon is to fix the capacitor to the board, but look how it stays only on top of the capacitor with no contact to the PCB. The manufacturer is putting that silicon without care only to justify the quantity of material used? Where is the quality control??? How this PCB has passed? Because for sure we are paying premium prices and some of this price should find itself in a better quality control, right? Otherwise we are paying premium prices for a "Chinese like manufactured" product....

pic.jpg

0 Likes 0 ·
pic.jpg (326.2 KiB)
Alexandra avatar image Alexandra ♦ Alex Pescaru commented ·

So a seller issue not a victron issue.

For the record then the seller is a box mover not likely to be authorised. (And if they are authorised need some discipline)

I am sure you can contact victron direct and work the warranty out through other channels such as the support page.

0 Likes 0 ·
James Pitstock avatar image
James Pitstock answered ·

Update

Completed a warranty support request on Saturday and was contacted by the distributer Monday morning. A new unit will arrive in 2 days and I am required to pack the old unit in the box of the new unit. Return of the old unit is organised by the distributor. Victron has a very efficient distribution system and I'm very happy with the fast turn around and minimal impact to my system.

2 |3000

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