UPS2 pump failure

[MrDuck]

A Grundfos UPS2 15-50/60 pump has just failed on me whilst I was recommissioning a small micro bore system after cleaning. It blew a 5 amp fuse on start-up and on inspection one of the electronics boards has failed short circuit causing a flashover between the live and neutral terminals internal to the pump. Definitely no way water has got in.

Grundfos are getting me a replacement under warranty but I am concerned at a catastrophic failure in an almost new pump (manufactured week 34 2013, only installed in May and not used until two weeks ago). Am I the only one with this problem? I only installed the UPS2 because the original British Gas supplied Grundfos pump was badly balanced and caused a lot of pipe vibration.

 

[Agile]

I have not heard of one like that.

Can you post a picture of the burnt parts?

Tony

 

[sooey]

They were manufactured with a design fault causing them to draw too much current on start up. There was a modification to fix this, sounds like yours was one of the faulty ones.

 

[MrDuck]

Agile - please find attached a photo showing the back of the terminal board and the inside of the housing.

I think that what happened is that initially the 5A plug-top fuse blew on switch-on at the start of recommissioning after draining down and refilling (I had the pump directly connected to a plug-top as I had been cleaning the system continuously for two days and needed to keep the pump on). Puzzled, after inspection I fitted a new fuse and switched on again.

This time there was a phut sound from the pump and it then measured open circuit. I suspect that the narrow circuit trace by the triangle of small holes (presumably an internal safety fuse) vaporised due to the pump fault (which maybe had got worse?) but the plasma it produced allowed a flashover between the L and N terminal contacts causing a lot of damage until the plug-top fuse blew. Not good and could have caused damage to my boiler had the pump been connected through it.

On testing the power supply board is a now dead short from the filter onwards (no obvious cause or other damage). Why the pump failed is a complete mystery - some sort of surge I suppose - but disappointing, and using a narrow circuit trace to interrupt 240v mains may not be satisfactory as this example shows. An enclosed fuse would be better.

Incidentally, the pump instructions give no guidance as to the correct protective fuse rating, which is clearly important.

View media item 81643
Hope this is useful. Do you know if spare parts available? It seems a pity to waste the rest of the pump.

 

[MrDuck]

Agile - Should have mentioned that I cleaned the pcb up quite a bit after removing it and before photographing. The original flash debris was over a much larger area.

 

[Agile]

The current surge is of little importance if powered from a plug.

But it was enough to weld the contacts on small relays as used on programmers.

They fitted a resistive lead as a start. I think about 40 ohms to limit the inrush current.

They take about 0.7 amps so a 5 amp fuse is about right, although a three amp mains fuse would also be fine.

Without seeing the components on the top side I cannot work out exactly what the function is. Your suggestion is possible correct.

Are you sure the filter is not shorted? Or is it only inductive by-fillar inductor?

Tony !

 

[MrDuck]

The short is on the pump side of the filter, that I do know, but I would have to remove components to check further, and as the board is so badly flashed I wouldn't want to repair it and use it again. I believe there has been some redesign of the pump and I hope that it has addressed the issue.

 

[Happyplumber]

That particular model had quite a few issues and Grunfos will change them out without quibble

They had nightmares when the pumps were first changed over to the new highly efficient design.

This seems to have been ironed out now ,we were unfortunately their testers!

 

[Phill Peck]

I've had a few of the updated ones fail.

I had a new one I installed on a new system that was unable to move water

I've not installed on since

 

[Boilerman2]

My Company no longer use Grundfos pumps largely due to early obsolescence of heads for Commercial pumps and due to many problems with the UPS2 range

 

[MrDuck]

None of this is good news, is it? I have just received an FOC version 3 pump from Grundfos today and hope it will be successful. It seems to have a built in delay between powering on and starting the motor. I presume this is to stop the current surge. Manufacture of v3 started end of June 2014, I am told.

Boilerman2 - I appreciate your advice but what manufacturers has your company found satisfactory? All the A-rated pumps are pretty new and electronics has to be more unreliable than a simple induction motor, if only because of the greater complexity. At least I get a 5 year warranty with the Grundfos. Also, what UPS2 problems have you found? Is it just the current surge or have their been other difficulties?

 

[Boilerman2]

MrDuck

I cannot say if it is good or bad, but we are currently fitting Wilo products based on customer service and parts for old pumps availlability

 

[MrDuck]

Agile (and others) - on investigation it was the bridge rectifier in the power supply that had failed. Have replaced this (pence) and cleaned up the mess from the flashover and all is now well. Also varnished over the area of flashover to prevent a recurrence. Grundfos let me have a replacement pump FOC anyway so now I have a spare! Not too happy though that it should have failed in the first place - I prefer a pump that's bulletproof.

Thanks to everyone for your comments.

 

[Agile]

A bridge rect used on mains is usually rated at a minimum of 800v and would rarely fail although some do.

Usually it would be caused by a spike in the mains ( lightening etc. ) but in your case it was after the filter which would usually remove the worst spikes.

Glad we know the cause now but I am still surprised at it flashing over unless the back emf on the motor somehow caused it.

Tony

 

[MrDuck]

I suspect overheating may just be a possibility. The spec sheet for the rectifier recommends large areas of pcb copper adjacent to each contact to take heat away, but the design I repaired does not have this. The pump failed when it was quite hot after pumping a mixture of hot water and air due to a partial airlock, so the rectifier in its enclosed environment may have got very hot as the pump "grabbed" at the mixed air and water flow. But I feel a pump should cope with this.

The part was 800v rated as you suggest and I agree it should not have got spiked as there is protection against this. I have replaced like for like using a Fairchild DF08S part.

The flashover was, I think, due to the pcb fuse (a very narrow bit of track near the main terminals) vaporising and the vaporised copper in turn allowing a flashover between the live and neutral contacts on the terminal block. Either way a big 'phut'. I have varnished the terminal block joints now to offer some protection against this happening again but a better design would have the pcb fuse track on the other side of the board, or better still have an enclosed fuse (there is a reason plug top fuses are so heavily made with sand filling to suppress vaporised material!).