Perhaps not much help, but I recall that I had surge problems on my 813 p-p amplifier. The power transformer was a C-core 612-0-612 rated at 440mA. I cured the constant fuse blowing with a simple thermal anti-surge device in the mains supply. I suspect that the problem might have been worse had I been using toroids.
7N7
7N7
Oh dear! This was about 12 years ago. It was simply a little circular ceramic thing I bought from Farnell most likely. Search the catalogue of your component supplier for Surge Protection and I daresay that these days a superior product is available.
Those clever people around here I suspect, could probably offer some solid-state slow start-up circuit which would be better!
High-current toroids I understand, can cause trouble with surge, so I have always chosen slow-blow fuses for these, but with the scale of your amplifier perhaps something more is required.
7N7
Last edited:
Old high power tube amplifiers using transmitting tubes and multiple kV plate supply usually had a spark gap connected across the output tubes plates (OPT primary) to protect the OPT if anything goes wrong (unloaded secondary,oscillations,...), then the arcing will appear outside (across the spark gap) and not inside your (expensive) output transformer. A big bleeder resistor is placed in series with the spark gap. (a 10k/100W was used in the G.E.C 450W amplifier). I strongly recommend using such a protection in your amp to prevent any further transformer damages during the de-bugging of your project. Spark gaps are still available, search for "voltage surge protection" , or "spark gap protectors". Don't expect doing this with series connected zener diodes or Varistors,they will fail immediately. A spark gap is "auto-resetting" and (allmost) fail-proof. Of course this will not protect against arcing between windings and core,etc...
Another probable (and very likely) cause is inadequate insulation ratings of your transformer for your application. You should contact the manufacturer of the transformer and ask for the complete insulation test data's especially if the transformer has gone through a HIPOT TEST. (see below)
Problem is that only power transformers are tested for insulation ratings and these specs are usually not available for signal transformers.
Good luck for your great project...
The hipot test is a nondestructive test that determines the adequacy of electrical insulation for the normally occurring over voltage transient. This is a high-voltage test that is applied to all devices for a specific time in order to ensure that the insulation is not marginal.
Another probable (and very likely) cause is inadequate insulation ratings of your transformer for your application. You should contact the manufacturer of the transformer and ask for the complete insulation test data's especially if the transformer has gone through a HIPOT TEST. (see below)
Problem is that only power transformers are tested for insulation ratings and these specs are usually not available for signal transformers.
Good luck for your great project...
The hipot test is a nondestructive test that determines the adequacy of electrical insulation for the normally occurring over voltage transient. This is a high-voltage test that is applied to all devices for a specific time in order to ensure that the insulation is not marginal.
Last edited:
It might,but striking voltage will be unpredictable and hygrometric dependant in open air. Radio oldtimers made their spark gaps with a pair of razor blades mounted on a insulating (ebonite) board but much experimentation was needed.
Today these are called surge-arresters and have stable repeatable characteristics, available with striking voltages ranging anywhere from less than 100V to over 5 kV.
EPCOS has a wide range of these spark gap arresters:
http://www.epcos.com/web/generator/...pannungsableiter_und_Schaltfunkenstrecken.pdf
Last edited:
truth be told I would like a method of "seeing" the spark. For example.....during operation of the amp and this starts to oscillate on its own.....it would give anyone a clue as to something bein wrong.
gas surge arresters are commonly used against lightning strikes....certainly a very good, pratical choice.
gas surge arresters are commonly used against lightning strikes....certainly a very good, pratical choice.
They do this on modulation transformers. Put a spark gap across the primary and or the secondary.
across the secondary would not protect the primary winding and given the ratio of the transformer I doubt I would be able to precisely calibrate the distance between electrodes.
I checked the spark gaps on modulation transformers and it is exactly what I would need. I ordered some cheap (1 euro) test tubes (x5) made in borosilicate glass to house the spark gaps and/or suppressors. I will buy this epcos model
B88069X4251S102 EPCOS | Mouser
I just wished I had thought of this earlier....oh well it is a learning experience afterall. Hopefully it will help oher builders not make the same (expensive mistakes).
I checked the spark gaps on modulation transformers and it is exactly what I would need. I ordered some cheap (1 euro) test tubes (x5) made in borosilicate glass to house the spark gaps and/or suppressors. I will buy this epcos model
B88069X4251S102 EPCOS | Mouser
I just wished I had thought of this earlier....oh well it is a learning experience afterall. Hopefully it will help oher builders not make the same (expensive mistakes).
Another update:
I checked the connections and it would appear one of the ccs leads to the anode plates of the 6bl7 was a cold solder. It eventually snapped off the socket pin. Yet another reason for instability.
Also this (or the accident) destroyed the triode section of the 6bl7 which now reads weird voltages (like 87v anode plate voltage instead of 400+v)
I swapped the 6bl7 around and now I get normal voltages.
The bias setting network for the gu81m must however be replaced as the lowest bias is only -250v.
the mosfet bias setting resistor will definetely be replace by a 5600ohm resistor (instead of 15k).
I checked the connections and it would appear one of the ccs leads to the anode plates of the 6bl7 was a cold solder. It eventually snapped off the socket pin. Yet another reason for instability.
Also this (or the accident) destroyed the triode section of the 6bl7 which now reads weird voltages (like 87v anode plate voltage instead of 400+v)
I swapped the 6bl7 around and now I get normal voltages.
The bias setting network for the gu81m must however be replaced as the lowest bias is only -250v.
the mosfet bias setting resistor will definetely be replace by a 5600ohm resistor (instead of 15k).
Damage from the opt
Inviato dal mio GT-I9001 con Tapatalk 2
An externally hosted image should be here but it was not working when we last tested it.
Inviato dal mio GT-I9001 con Tapatalk 2
the damage appears quite deep inside the transformer which would point to something more than a random istantaneous event but then again I have no experience with this kind of power levels and output tubes.
The transformer can be rewound and before anything else happens I will install the GDT suppressors.
The transformer can be rewound and before anything else happens I will install the GDT suppressors.
Can you determine if the arcing appeared between primary to secondary windings , primary to core , or between adjacent windings ? Where is the blue wire connected ? Is the secondary (speaker output) floating or referenced to ground in your circuit ?
It's very unlikely your transformer failed due to over-current or overheating in such a short time, thus it can only be an insulation issue. As I said in one of my previous messages it would be good to contact the manufacturer of the transformer and ask them for complete specifications , maximum voltage ratings and high-voltage insulation tests (if any). This problem should NOT happen again and you MUST find the cause before going any further. Don't feel discouraged, this is not an ordinary project and you will be faced with many other difficulties, but it can be done. Problem is that very few members here have real experience with the building of such high-voltage high-power amps.
Keep us informed about your progress.
It's very unlikely your transformer failed due to over-current or overheating in such a short time, thus it can only be an insulation issue. As I said in one of my previous messages it would be good to contact the manufacturer of the transformer and ask them for complete specifications , maximum voltage ratings and high-voltage insulation tests (if any). This problem should NOT happen again and you MUST find the cause before going any further. Don't feel discouraged, this is not an ordinary project and you will be faced with many other difficulties, but it can be done. Problem is that very few members here have real experience with the building of such high-voltage high-power amps.
Keep us informed about your progress.
Your OPT will need to withstand peaks of AT LEAST 3kV with those specifications, and this is under normal operation, and no resonances. This is not easy to do on a toroid OPT. It is vital that your windong company knows this, not just the number of tuns etc (and I suspect you did use sectioning/sandwiching of layers).
The insulation used on toroids can normally hold 4-5kV, depending on what type is used, but it's not difficult to increase this using multiple layers. The leadout wire can be a problem, and should be additionally insulated with silicone tubing or similar means to prevent them arcing to other layers as they go from the inner layers of wire to the outside of the transformer.
And, MOST important - it must NOT be wound using a typical 'round th ring' winding technique used on power trafos. Rather, the winding must be wound by placing wire in a 'X turns forward on the ring, Y turns back, repeat' fashion, where X is more than Y. The windings themselves must be sectioned and spaced out on the core to prevent imbalances and too high voltage differentials, and also divided into insulated layers wound like I described.
When doing calculations, it would be well adwised to use a good 50-100% security factor. If I was doing it, 5kV sustained would be the minimum.
Finally, since you are running in pentode mode, keep in mind the plate is high-impedance and will not dampen internal trafo resonances as well as oscillations of the amp. Some measurement at low power and judicious application of Zobel networks can mean the difference in life and death of your transformer in these conditions.
thank you all for your input. I am not discouraged! On the contrary!
I heard sound from the amp and that is more than I need to keep going.
The transformer company is specialised in toroids with industrial applications of all types. They do have a winding method for OPT however where given instructions to wind the toroid by lem8r using another technique. Also they were instructed to insulate te transformer to 3kv at least.
Obviously more insulation will be installed.
I think that what happened was that a hige imbalance was created following a tube shutting off and the other conducting. The insulation was most likely punctured and the "hotspot" burned to the surface.
5kv insulation is a good approach. Thanks!
I heard sound from the amp and that is more than I need to keep going.
The transformer company is specialised in toroids with industrial applications of all types. They do have a winding method for OPT however where given instructions to wind the toroid by lem8r using another technique. Also they were instructed to insulate te transformer to 3kv at least.
Obviously more insulation will be installed.
I think that what happened was that a hige imbalance was created following a tube shutting off and the other conducting. The insulation was most likely punctured and the "hotspot" burned to the surface.
5kv insulation is a good approach. Thanks!
yes Tony, you are right. I read about the phenomenon in the last few days.
The only real solution is to dessicate the transformer and "pot" it.
I can do it with relative ease (plastic cover, liquid rubber or epoxy) but the problem is that if there is another issue with the opt rewinding it will be impossible and thus another mistake will ost much more.
I will install 4kv GDTs in clear glass tube envelopes so replacement and inspection should be easy in the future.
With insulation rated at 5kv the opt should be well protected.
The Core is large and the transformer heavy. What would be an inexpensive way of dessicating and potting the OPT?
Thanks
The only real solution is to dessicate the transformer and "pot" it.
I can do it with relative ease (plastic cover, liquid rubber or epoxy) but the problem is that if there is another issue with the opt rewinding it will be impossible and thus another mistake will ost much more.
I will install 4kv GDTs in clear glass tube envelopes so replacement and inspection should be easy in the future.
With insulation rated at 5kv the opt should be well protected.
The Core is large and the transformer heavy. What would be an inexpensive way of dessicating and potting the OPT?
Thanks
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.