It would be interesting to better understand possible failure modes. I’ve only owned valve power amps for a decade or so (Prima Luna Prologue 2 and the Leak) and have yet to have any valve or other failure.
I assume given the huge number of surviving Leaks, Quads etc nothing irreparable tends to happen. Plus in my time as a volunteer on the Manchester SSEM replica I’ve never seen a valve failure do any damage beyond not working yet. With that machine the main point of failure tend to be old carbon resistors being way high or low and the circuit simply not working as a result. We had one resistor start smoking a while back, but replacement was all that was needed.
So, if a Leak Stereo 20 goes wrong how does it tend to go wrong? What should we look out for? I’m more interested in failure modes in a properly restored example, i.e. one where 60 year old resistors, caps etc aren’t the issue.
I’m assuming one channel will go noisy/off if an ECC83 fails, nothing more catastrophic than that, but I’m not clear about the possible failure modes of the EL84s and GZ34. What does one keep an eye out for?
I had a Mullard GZ34 fail on my Radford recently with these symptoms in the opposite order. A month or two of fuse blowing at two to three week intervals followed by a visible internal flash. After that it was only giving out 40-ish volts.Next most likely is GZ34 failure although Mullard ones seem very reliable. If using a modern GZ34 this is a likely suspect. Usually internal sparking followed by fuse blowing.
I had a Mullard GZ34 fail on my Radford recently with these symptoms in the opposite order. A month or two of fuse blowing at two to three week intervals followed by a visible internal flash. After that it was only giving out 40-ish volts.
You are correct Tony, it is almost completely original. Yes, it has been looked at and tested - and given a clean bill of health in all other respects! The problem (the bit I grasped) was gas in the rectifier as shown by a retreating getter. The more detailed explanation has been forgotten in the subsequent period, I apologize.The fuse-blowing would have raised deafening alarm bells for me. IIRC your amp is ‘part restored’ rather than being a full rebuild too, so I’d want to know if something else being borderline had actually taken the GZ34 out (they have a reputation of being exceptionally long-lasting). Hopefully it is the GZ34, but I’d still get it looked at and tested as at £100+ these days you don’t want to be blowing any more!
PS I’d be curious to know how a failing GZ34 could blow fuses weeks before giving up? I’d have thought it would just be a bit saggy and show voltage drop on the other side. How would it blow a fuse before the point of catastrophic failure?
John_73, you'll be fine with 1W carbon film. The actual dissipation here is around 500 to 600 milliwatts.
Thanks Jez. I was initially going to order the following for the cathode resistors:
https://uk.rs-online.com/web/p/through-hole-fixed-resistors/4851975/
But the recent suggestion that these also function/double in a safety role gave me pause for thought.
What voltage do these resistors each ‘see’? The carbon films offer 500V working voltage whereas the wirewounds linked above are only 120V.
Thanks. Anything to be gained by going for the Mills non-inductive wirewounds or is that (yet) another case of audiophile excess?
Does anybody know the real-world operating voltages for C3 and C5 please? From the circuit diagram it seems C3 would be 160V max, but nothing is there for C5. I have 0.1uF 400V sets and 200V sets of the K40Y-9 Russian paper in oil types, and would prefer to use the 200V ones if possible as they’re less bulky, but wasn’t sure if the voltage rating was a tad too low. Cheers.