SUPRIMA 30 BLOWING PCB FUSE

[mmalone]

Guys... I dont have my multimeter or fuses handy tonight when called to a mates house , suprima 30 s plan system , no lights on..

checked pcb and fuse had blown, had 1 x 3amp quick blow in my box so retried but disconnected GV and Fan, turned the boiler control knob to 0 no controls calling for heat, switch on at spur and blew straight away...

Tomorrrow I will get to the bottom of this and get my meter on the case but me thinks the pump is the problem....

The pump is wired direct to the boiler and Im sure the pump cycles on an overrun with these. I replaced the pcb last year on this on and no signs of anyprobs and mate advised its never when to reset once...

just wondering if this is common in blowing the internal fuse or if im overlooking something...

anyone confirm this is likely to be a pump issue or not ??

 

[ollski]

Pump is always suspect number one with fuses blowing.

 

[pierrot]

Pump is always suspect number one with fuses blowing.

DITTO

 

[croydoncorgi]

I really, REALLY wish that boiler manufacturers and board suppliers would agree amongst themselves that it's a Bad Idea to put mains voltages on boiler PCBs. There are several infamous examples which feature here regularly (eg Puma, Suprima, ...). Most of the failures are either because the board gets too hot (due to amps going through it to run fan, pump, etc.), or the actual components (especially miniature relays) carrying mains go bang, or the header pins fall out of the board (Puma!!!).

Given that there are NO microcontrollers or microprocessors that run directly off the mains, it seems to me a sensible design decision to make PCBs a low-voltage area. When did you last hear of a Personal Computer PCB failing?

Of course, you still need to run at least the pump off mains, so ther would need to be one or more externally-mounted relays - but that's no big deal.

It would also be a lot safer to work on!

 

[mmalone]

Pump is always suspect number one with fuses blowing.

DITTO

sounds good to me too, normally see the spur fuse blow on a lot of boilers when this happens and not the actual board internal fuse but the pump is direct via the pcb on these just like a combi.... hopefully tomorrow its an easy fix and is the pump..
really gotta be as its the only other component which draws lots of power
and ive eliminated the fan and GV... actually forgot the pump was via pcb to start with hence the reason I didnt start here !! as it normally ios no1 suspect...

was hoping this was the case, cause replacing the pcb again aint cheap on these as we all know...

 

[ChrisR]

Not long ao I had a suprima high limit stat leaking current to earth and blowing fuses. Looked OK on a multimeter though.
Finding it was easy though I wouldn't have bet any money on it being that. Started with GV, fan and pump disconnected and it blew a fuse, so even I could work it out then.

CC there isn't any problem whatsoever putting mains voltage on PCB's. All power supplies use them, and think of tv's and monitors with thousands of volts. It's "mature technology". If Potty had spent just a couple of pounds more on their PCB's they'd have cut the failures hugely, and they know it perfectly well. The only reason that they didn't is that they chose not to, for what I assume to be business reasons.
Going double sided (copper) with plated-through holes is a major step, which most other manufacturers took years ago. (That's where the inside of the hole through the board is plated with copper connected to at least a ring plated on each side, so the solder appears on both sides of the board and surrounds the pin as it goes through the board.)

 

[mmalone]

Not long ao I had a suprima high limit stat leaking current to earth and blowing fuses. Looked OK on a multimeter though.
Finding it was easy though I wouldn't have bet any money on it being that. Started with GV, fan and pump disconnected and it blew a fuse, so even I could work it out then.

CC there isn't any problem whatsoever putting mains voltage on PCB's. All power supplies use them, and think of tv's and monitors with thousands of volts. It's "mature technology". If Potty had spent just a couple of pounds more on their PCB's they'd have cut the failures hugely, and they know it perfectly well. The only reason that they didn't is that they chose not to, for what I assume to be business reasons.
Going double sided (copper) with plated-through holes is a major step, which most other manufacturers took years ago. (That's where the inside of the hole through the board is plated with copper connected to at least a ring plated on each side, so the solder appears on both sides of the board and surrounds the pin as it goes through the board.)

chris, will keep that in mind mate if the pump aint the problem... had I had more fuses or my mulitmeter I would have found it easy enough as really didnt have much else to go for...

dont think its a leakage of current to earth I would expect the mains electricity breaker to trip which in this case aint happened from what ive seen.. will let you know if it is though and very handy to keep this in mind for future...

 

[Agile]

Whilst the pump is a likely failure point, there are supression components on the older PCB which can blow the fuse.

Just to make it more difficult this will sometimes only happen intermittently and only when the full mains voltage is applied so you cannot test for it.

I encounter this about once or twice a year.

Try disconnecting the pump first and if this does not cure it then expect the PCB.

If it really is not the PCB then consider the wiring harness or the o/h stat Chris mentioned.

Tony

 

[croydoncorgi]

I never said that mains on PCBs COULDN'T be done. But I still argue it SHOULDN'T. As boilers come to need more than just a bunch of resistor / capacitor networks and relays, and get memory, ASICs, processors, etc., putting mains in close proximity becomes harder to deal with anyway. Producing microcircuits immunised against mains and high voltage transients must be more expensive than moving mains away from the PCB altogether. OK - power supplies DO use a printed circuit but only 'in isolation' from the rest of the system, in most examples.

And before you remind me - yes of course something needs to be done with the HT spark generator for the boiler ignition. But it's already in effect isolated from the rest of the PCB for safety reasons.

Tony's comment about 'suppression components' giving problems is yet another reason for low voltage boards - mostly these don't need any suppression.

 

[ChrisR]

I'm going back a bit but where I worked we always found it cheaper to make one board with everything on it than separating things, space permitting, etc. The economics would be tilted if you could make a separate board very cheaply; if the quality standards are higher overall then one board wins. Much of the "quality" that's needed in higher power circuits is because the components weigh more and put physical stress on the boards.

Ideal went both ways in the M series boilers so there can't be much in it. (Only later versions have HT dept separate)

Again , saying " But I still argue it SHOULDN'T be done" is coming from the point of view of wanting reliable pcb's. That isn't potterton's aim, or they would do other things first to achieve it. They didn't want Suprima boilers which never need spares buying for them. Installers bought them not users, and installers don't care how long they last as much as how easy they are to lift! The economic climate seems to have changed, as we're seeing more 5 year guarantees, etc.

It may just be that they've got their engineering better organised and can make things with a more predictable life. MTBF figures have been available on every component the mfr buys, for decades. (Mean Time Between Failures). Otherwise a mfr trying to manage quality versus cost wouldn't touch an unknown component.

Ford have always been very good at it. At one time the design life of a car was 60,000 miles. So if a wheel bearing always lasted 70,000 they would make it cheaper. The shape of the failure vs age curve is obviously very important, you want them all failing at 60,001 miles. . Last I heard the figure was up to 100,000 miles. As an example, cam belts. Once they were all 40,000 miles life, now (eg Fiesta Zetec) is 100,000 miles.

Same is true of suppressing components. If you want them more reliable, make them bigger and they'll be happy with 1000 volts, or normal mains for lots longer. That's a fairly expensive component though, so a typical place to use board-life-limiting bits.
I still have some 12kV capacitors somewhere, wot I broke with too many volts. Half inch thick ceramic, 3inch diameter, split apart. (RF welding of plastics)

"..the HT spark generator for the boiler ignition. But it's already in effect isolated from the rest of the PCB for safety reasons. .."
Not really, flame detection is fairly intimately connected with the HT in most boilers, and not particularly unreliable.
ALso think Netaheat spark generators - 20 odd years and you don't see many of those failing.

The heat producing stuff, Amps, is of course much higher in low voltage circuits too. (I squared R)

 

[John.B]

Nice post Chris, I reckon the MTBR on a Suprima must be 18 months to 2 years then! I,ve heard of some Suprima,s PCB,s failing under 12 months. Talking of PCB,s failing, starting to get more call for Glow-worm Micron,s PCB,s at about 30 month,s.

 

[ChrisR]

It's probably more like half a dozen years. In an ideal world you'd love to get graphs of "number of failures" vs "time" like the second one here.

With that shape graph, in the "early" white bit of the coloured graph you still have something like 1%, which is a lot of suprimas, but that's in less than a quarter of the "mean" llifetime.

In realilty you get a skewed graph for loads of reasons, which lifts the figures in the early years, or overall, or late. eg these all lead to early failures:
=Bert dropped the board
=Software problems by bad design - the electrodes get condensation on in the winter (Suprima, Micron) so the board locks out unless it has several goes at lighting
=Software problems by cock up - quick switching upsets the software, which does something daft like turn the pump off so the boiler overheats (Micron)
=Installer error - rain through a misaligned flue gets to a thermistor/aps (Micron, Chaff Britony, all) or something critical and blows the board

Then there are the things which give a (nice predictable) lightbulb-like failure pattern where something starts burning out from day one, and they all tend to fail at a similar age. Like overheated diodes and resistors(Suprima), cables (Lynx).

If you superimpose all the fault patterns you can end up with a flattish curvy W shape. Lots of early failures, a bunch around an average, and a rise when everything's old.
I could go on - oops, too late .

 

[croydoncorgi]

Judging by user comments on this forum and elsewhere, users in general are not that naive any more (even if they once were). Based on usually-satisfactory PC and other domestic electronics experience, they get very hacked-off if things go bang.

Potterton have shot themselves in both feet as far as the Suprima is concerned. The name has such a reputation for 'bad electronics' that many people will never buy it, even though Potty has now moved to a Honeywell board which seems to be reliable - or at least 'average' rather than 'poor'. And presumably despite charging their captive market for ALL the new bits required, I doubt that overall Potty came out ahead financially. Too much loss of goodwill.

In the end, it comes down to Honeywell versus Pactrol versus ..... I don't think Honeywell would produce a special 'replace annually' PCB just for Potterton (even if others might have)!

 

[mmalone]

guys. just to let you all know IT WAS THE PUMP thi time..

cheers...

 

[kevplumb]

got off light there then