Boiler PCBs

[diditmissen]

More a point of interest than a query really.

I have recently had to fix my combi boiler - a Potterton Puma 80 - for a fault which has been very intermittent for the past 5 years but which has recently been occurring much more frequently and has got to the point when I can't rely on the boiler to light.

The problem was that the boiler would fire on demand but would sometimes stop after a time ranging from a few seconds to a few minutes. It would then fire up again, often remaining lit but sometimes repeating the same light-fail-light-fail sequence. Problem occurred whether demand was for hot water or central heating. Obvious causes such as water/system pressure switches, pilot thermocouple etc were all checked and OK.

Luckily however, during a particularly bad episode, I happened to notice the sound of a relay "chattering" with an accomanying flicker of the demand LED on the PCB (cover was off at the time). Wiggling the wires to one of the multi-connectors was found to make the problem go away (or return if the boiler was firing!).

Removing the PCB and inspecting the tracks proved to be something of a revelation. This particular problem proved to be caused by the fact that the connector pins are soldered directly through the PCB without any further mechanical support, so pushing home a connector puts pressure directly on the solder pad and consequently the PCB tracks themselves. In this instance the whole of the copper track had broken free of the board and had fractured. The sporadic nature of the fault was probably thermal in origin and linked to the expansion of the materials as the boiler fired.

In this case it was a simple job to create a wire bridge across the break. Aside from the obvious weakness of this construction, further examination of the PCB revealed 14 - yes FOURTEEN - solder joints which I would have been ashamed to have made myself. Some were incomplete and allowed movement of the component lead. Some looked like possible dry joints and had a very "crystalline" look. I resoldered all of these. Ten minutes work to solve a problem which had been bugging me for five years.

What I'm really trying to say is that, if you were faced with a similar issue, then it would be well worth having a good look at the PCB before consigning it to the bin. By putting right what the manufacturer couldn't be @r$ed doing in the first place might save you a couple of hundred quid.

 

[Paul Barker]

Yes it is a valid point, and true for many pcb's from that period.

The same could be said of televisions from that same era.

Today they are flow soldered and used silicon components, so if it's the pcb it's in the bin. But in my experience the modern flow soldered surface mount component pcb's are more reliable.

When I get a fault in a boiler which leeds to the pcb I like you check all leeds etc first as it isn't always the pcb anyway it might the molded plug onto the gas valve or any other leed. It very often is the soldered joints surrounding relays as the movement and perhaps the heat liberate the joint. Then sometimes the relay itself has so many contacts in it's life and then like the mouses heartbeat it has reached it's designated number.

The other day I found myself servicing a boiler with a satic maximum burner pressure completely non adjustable at the modureg. On close inspection I found just 5v dc at the modureg, so I checked the value of the temp pot and the turmister both of which were correct. So it had to be a pcb based problem. It transpired that the leed plugged into the pcb for the modureg was not on the modureg socket, it was on a jumper which should never have been used. When I put it where it ought to have been the boiler modulated again. It does make you wander whether the previous 4 engineers actually checked burner pressure. In the event I was at the job 2hrs until it was all sorted and the customer obviously thought I was a compolete numpty as I seemed to be scratching my head a lot and geting nowhere for quite some time. As indeed you would in this scenario.

Similarly I once found a brand new Vokera connected up all wrong. So clearly it hadn't been bench tested either.

Basically our job isa very difficult one and we all deserve medals. Except for the guys who just put it back together as found and walk out quickly.