|
Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members. |
|
Thread Tools |
24th Feb 2017, 6:42 pm | #1 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Rectifier Flashover and open circuit anode resistors
This morning I thought I'd fire up the EL34 monoblocs to do a few checks prior to doing a PSU change (new transformer as part of a separate PSU).
On switch on the GZ34 on one of them pretty much instantly had a "blue fireworks" display inside the envelope. Its a venerable Mullard, probably the thick end of 60 years old but has given sterling service up until that point. ON checking the amp it seems the 100 ohm 7watt WW resistors i used on each anode had gone open circuit, couldn't see any other visual signs of damage or over heating and the mains and output transformers measured fine with no open or short circuits anywhere. Nor could i measure any shorts to earth. So could a valve that failed like this on switch on take out the anode resistors? I measured the resistances on the valve pins with my DMM and there were no anode to anode or cathode to anode shorts measured cold. But I think the valve may be a bit "gassy" now as there seemed to be a definite blue glow (ionised gas?) afterwards. I'm going to check the other mono too as I have had anode resistors fail open circuit in the past which I put down to insufficient power rating.All voltages check out and the amps draw the correct current at about 70 to 75 mA per EL34. Any tips or ideas as to what happened? I feel that I may have used inferior wirewound resistors, these were the standard white cement "coffin" jobs that i bought from Maplins as i wanted them fairly sharpish and i wonder if a better type is available. I had thought 7 watts was more than adequate, but am wondering now. If the blowing of the resistors has saved the expensive output transformer from permanent damage then its been a small price to pay. Andy. |
24th Feb 2017, 7:18 pm | #2 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: Rectifier Flashover and open circuit anode resistors
Just a thought:
The reservoir/smoothing capacitors are rated at 500V and the volts can rise to around 480 until the amp warms up and they settle down at about 450 to 460V. I have had a bit of overvoltage issues in the past due to high mains etc (documented elsewhere on this forum) and wonder if the capacitors are failing? They are the right sort, proper PSU jobs but I have a feeling that I may have put them under a bit of stress. If so, I think i can re cap the amp but wire pairs in series so that they aren't working so close to their max operating voltage. I do know about using resistors across the capacitors when doing this. Which should also allow them to "Bleed" to earth when switched off. I always fit a "bleed" resistor across PSU capacitors usually only the one at the first cap but it allows all of them to lose stored charge on powering down. |
24th Feb 2017, 9:23 pm | #3 |
Heptode
Join Date: Apr 2003
Location: Cornwall, UK.
Posts: 989
|
Re: Rectifier Flashover and open circuit anode resistors
Andy, what's the value of the reservoir capacitor?
SimonT.
__________________
The honesty of imperfection.......... |
24th Feb 2017, 10:12 pm | #4 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: Rectifier Flashover and open circuit anode resistors
I should have pre empted that question!
50uF off the cathode, well within the 60uF maximum for a GZ34. You can safely assume I know about maximum capacitance across valve rectifiers. I am starting to think that the PSU electrolytics are underspecced on voltage rating and/or not robust enough for the task given albeit i bought them as being entirely suitable as PSU caps. Andy. |
24th Feb 2017, 11:45 pm | #5 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,901
|
Re: Rectifier Flashover and open circuit anode resistors
But the reservoir capacitor going short wouldn't have taken out the anode resistors?
Did something turn the EL34s hard on, and that current produced the pyrotechnics in the GZ34? David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
25th Feb 2017, 12:16 am | #6 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: Rectifier Flashover and open circuit anode resistors
I see where you are coming from David.
No I can't see that being the issue. The fault occurred too quickly, the EL34s wouldn't have got warmed up, the flashover happened pretty much as soon as the amp was switched on. I am suspecting that a venerable GZ34 finally decided to give up the ghost although the heater still works so it wasnt a total loss of vacuum. I can't see or measure anything that indicates any circuit fault that would turn the valves on that hard and I'd expect damage downstream of the PSU in that case. The open circuit resistors are a bit baffling too as there's absolutely no sign of damage or heat and to fuse the wire inside i would have expected to see some sort of thermal or mechanical damage, maybe cracking or disintegration of the cement casing? I haven't a great opinion of these "coffin" jobs as I've seen more than one give up even when used well inside it's parameters. All transformer resistances check out, I can't measure any resistance between windings that may suggest internal shorts, this on both output and mains transformers. I'm in the process of cutting my losses on the mains transformers anyway, the reason i fired the amps up today was to do a quick set of checks prior to modding them to be used with an offboard PSU via umbilicals. In the process I'll now replace the original PSU caps just as a precaution. The new PSU won't shove more than 460V out by my calculations and will be silicon rectified which I personally prefer, cooler and less wasteful of electricity! I just would like to diagnose the failure today in case there is a fault I have unwittingly induced at some point. But I can't see where if I have. The amps have always measured bang on design specs once the correct HT voltage was obtained, even when running a bit high there were no obvious signs of distress. The EL34 were always running inside their limits. Bit of a mystery so far but I'll crack it. Andy |
25th Feb 2017, 12:40 am | #7 |
Banned
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
|
Re: Rectifier Flashover and open circuit anode resistors
In my experience thats how GZ34/5AR4 "always" fail in this fashion. By that I mean I have had a few go bang, and always at switchon. I have changed very few that didnt carry the load anymore purely from old age. Generally they are ultra reliable.
Its good to see that toroid going into service Regards Joe |
25th Feb 2017, 1:12 am | #8 |
Dekatron
Join Date: Apr 2012
Location: Worksop, Nottinghamshire, UK.
Posts: 5,554
|
Re: Rectifier Flashover and open circuit anode resistors
I had one go gassy. It was in a radio that had been left switched off for a few weeks.
It glowed bright green well before the heater glowed. I just replaced it and dumped the old one as it was in the 1970s and new ones were easy to get. |
25th Feb 2017, 8:27 am | #9 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,901
|
Re: Rectifier Flashover and open circuit anode resistors
So it sounds like two or three independent faults then. The GZ34 did it's own thing, and somehow the anode resistors had gone open circuit while the amplifier was in storage. Maybe some sort of corrosion in some really nasty parts?
I'm not sure I like the idea of separate power supplies with umbilicals. The choice of connector and cables will be interesting to get very low drop on the heater connection and safety for the high voltage. On the plus side it will keep the stray flux of the mains transformer further from the amplifier itself, and some people tend to believe that anything which is more unusual, or makes the job harder must sound better. But I suspect that the constraints it imposes on the design of the grounding undoes any real benefit. You still have some stray flux from the heater wiring, so that needs careful twisting and routing. I'd opt for a bit larger chassis for the amplifier with a nice C-core banded transformer in a screening can. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
25th Feb 2017, 10:47 am | #10 |
Octode
Join Date: Jan 2003
Location: Guildford, Surrey, UK.
Posts: 1,960
|
Re: Rectifier Flashover and open circuit anode resistors
If the GZ34 is gassy, is it not possible for an anode to anode short to be formed by the arc and then the rectifier anode ballast resistors to be burnt out by the 900V or so that would be across them? An HT fuse in the MT secondary centre tap to earth would not stop this current.
Equally, if an anode to cathode short is caused by the rectifier arc, would not that apply a lot of AC across the 50uF reservoir capacitor and cause excess current through the ballast resistors? Ron |
25th Feb 2017, 12:22 pm | #11 |
Dekatron
Join Date: Jun 2005
Location: Warnham, West Sussex. 10 miles south of DORKING.
Posts: 9,147
|
Re: Rectifier Flashover and open circuit anode resistors
I would agree about normal failure here. This is exactly how most rectifiers pack up. I would break open the old GZ34 to salvage the base. Make up a plug in silicon replacement using 2 x 25 ohm resistors, 2 x BY127 rectifiers or similar and 2 x 4700pf caps and use this as a temporary replacement for the GZ34. This will confirm it was just normal failure. Burning out the surge limiters does protect the mains transformer and main amplifier components from serious breakdown. J.
|
25th Feb 2017, 1:39 pm | #12 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,901
|
Re: Rectifier Flashover and open circuit anode resistors
Duh!I can be damned stupid at times....
I thought the anode resistors were parasitic stoppers for the EL34s. The writing of the original post made me think the context had switched from the power supply to the amplifier itself before the anode resistors were mentioned. So rectifier anode resistors destroyed by an anode-to anode arc in a gassy rectifier is perfectly understandable. They may have saved your HT transformer. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
25th Feb 2017, 2:12 pm | #13 |
Dekatron
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,398
|
Re: Rectifier Flashover and open circuit anode resistors
Sounds as if you were lucky to have some relatively fragile limiter resistors feeding the rectifier! "7W wirewound" sounds as if they should have passed quite a bit of current for a while but they obviously gave in quite quickly- I reckon if you'd used some of those green vitreous enamel wirewounds, they would just have simmered quietly until either the rectifier melted or the transformer thrashed itself into a barbecued smoking mess. When I do an overhaul on a piece of valve kit, I try if possible to work in a small,flameproof limiter resistor in series with each rectifier anode on something like a ceramic tagstrip. The flameproof bit of the description is worth looking out for- some older carbon film types with that attractive shiny lacquering can make a picturesque spit of flame as they burn out...
|
25th Feb 2017, 7:08 pm | #14 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: Rectifier Flashover and open circuit anode resistors
Thanks for the replies guys.
So it seems the 60 year old Mullard decided to go to the great PSU in the sky by what I can glean here. One shouldnt feel to upset, its a good innings for a valve that had already seen sterling service in a leak stereo20 before i swopped it into the homebrews. ( I have a deep mistrust of Chinese rectifier toobs at voltages above 350 and currents above 150mA, possibly wrongly). Yeah I also think there's something else going on as the other mono has an O/C anode resistor and a couple of things spring to mind. 1) at the thick end of 500V on the anodes maybe 50uF is a bit too much for the GZ34, I will re read the Mullard data book on this as I suspect at higher voltages the max capacitance is lowered? 2) Those white cement coffin wirewounds from maplins (bought cos I needed em quick they aint cheap from there) are not really of any sort of quality and i wonder if they come with a weak bit of wire as standard? To David (radio wrangler), the external PSU is devised due to Joe bogs generosity of providing me with one of his custom wound toroids on an extra large core. It might prove yet that I will incorporate it into a stereo chassis but I'll need a Fork lift to move the thing i reckon! The heater drop is covered by the provision of 9V secondary so that will be rectified and regulated inside each mono as I can use the space where the mains transformer used to be, and make a nice thick ally plate to cover the gap and act as heat sink fir the regukator chip(s). theres a 5 amp version of the 317 and i think the LD108* series too. the signal valves are already served by their own dc regulated supply board so its a simple matter of plumbing in the 9Vac there. I'm sure Joe will be happy to furnish any wisdom further to this, he's promised to keep me on the safe and steady road. I found some military high voltage connectors on the interweb so that side of things is covered. the umbilicals will be kept very short, half a metre is more than enough for my needs. Yeah the grounding is going to need careful thought, I can't feel comfortable unless I know I have adequate safety earths on the kit and at the same time earth loops could be tricky. The worst thing that can happen here is that it won't work and I will just go back to the drawing board and have learned something along the way. No real metalwork and soldering before mid to late march as the house is still a building site and i can't even get in the shed and garage to sort it out! Onwards and upwards. Andy |
25th Feb 2017, 7:09 pm | #15 | |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: Rectifier Flashover and open circuit anode resistors
Quote:
|
|
25th Feb 2017, 7:14 pm | #16 | |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: Rectifier Flashover and open circuit anode resistors
Quote:
Personally these days I wouldn't go back down the road of thermionic rectification unless i had a stash of rectifier valves and a transformer or winding for its heater. Which for me now means onlythe half dozen Ez81 and a few EZ80 I have lying around, mostly NOS. |
|
25th Feb 2017, 7:17 pm | #17 | |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: Rectifier Flashover and open circuit anode resistors
Quote:
Andy |
|
25th Feb 2017, 7:23 pm | #18 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,901
|
Re: Rectifier Flashover and open circuit anode resistors
For those sorts of currents, although there are three terminal regulators rated to 5A, I'd be careful in choosing them, or I'd go for a design with a discrete power transistor. The Linear Technology LT1083/1084/1085 are the ones I most trust.
David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
25th Feb 2017, 7:28 pm | #19 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,901
|
Re: Rectifier Flashover and open circuit anode resistors
Another thought....
If you use silicon rectifiers, look for high voltage 'ultra-fast' diodes. It may sound ****** stupid at 50Hz, but the slow ones store charge, conduct for a short time in reverse, and then turn off with a snap! which snap creates a lot of radio interference. If you don't know what's in that chinese rectifier module, I wouldn't use one. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
25th Feb 2017, 9:59 pm | #20 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: Rectifier Flashover and open circuit anode resistors
Your words are wise.
Yes it was the Linear technology jobs I had in mind, quite a bit more expensive but even on paper they have a superior performance. At my level the extra cost isn't an issue. Again agreed, I wouldn't use anything lower in performance than something like a UF5408 and even then I use high voltage bypass capacitors to reduce the spike at switch off. I did my reading years ago in Morgan Jones' books and he explained how the spikes could reach right into RF. I did exactly this with my KT88 kit amp, I didn't go with the expensive boutique Schottky rectifiers but stayed with the standard UF5408 bypassed with 10nF 1000V polys and that made a slight but measurable difference. And UF**** are not a massive amount more expensive than the cooking 1N**** variants. As to those modules, the thing that makes me think is that the maker has encapsulated the lot in resin, maybe for a good reason but it prevents one from seeing what's in there. For the fiver i guess it's no hardship to solder a pair of UF rects with bypass caps direct to the valve base and the limiting resistors can be incorporated into the existing anode resistors. Again thanks for the input gents. Now when I bung these back on the bench is there a reliable way to test for surge current through the rectifier on switch on? It's easy enough to measure the steady state current by simply taking voltage readings across the anode resistors. I did wonder about sticking a 1 Ohm resistor after the cathode just to use as a measuring point and then calculate currents using Ohms law? |