Hello all,
I had a frightening experience a few days ago when I was testing my aikido circuit. I switched on my power supply and was waiting for the tubes to warm up and the voltage to rise. All of a sudden I saw some big blue sparks in my rectifier tube. I immediately turned it off. My brother said that he heard some scary popping and buzzing from the computer speakers.
I have a feeling that I might have too much capacitance. What also confuses me is that this is not the first time that I have powered up the circuit (although not for long periods).
Here is the configuration:
power transformer is 300-0-300 100mA
5Y3GT--470R--40uf--2490R--220uf--11700R--320uf--Aikido circuit
Does anybody have any idea what might have caused the sparking? Can the tube still be used?
I had a frightening experience a few days ago when I was testing my aikido circuit. I switched on my power supply and was waiting for the tubes to warm up and the voltage to rise. All of a sudden I saw some big blue sparks in my rectifier tube. I immediately turned it off. My brother said that he heard some scary popping and buzzing from the computer speakers.
I have a feeling that I might have too much capacitance. What also confuses me is that this is not the first time that I have powered up the circuit (although not for long periods).
Here is the configuration:
power transformer is 300-0-300 100mA
5Y3GT--470R--40uf--2490R--220uf--11700R--320uf--Aikido circuit
Does anybody have any idea what might have caused the sparking? Can the tube still be used?
Hi lawbadman,
Your filter caps are new, not NOS - right? Also, your 5Y3 is only rated for 20uF assuming a 50 ohm impedance in the AC supply (from RC-27).
The tube may still be okay, but lose the 40 uF input cap. 10~20 uF should be fine.
-Chris
Your filter caps are new, not NOS - right? Also, your 5Y3 is only rated for 20uF assuming a 50 ohm impedance in the AC supply (from RC-27).
The tube may still be okay, but lose the 40 uF input cap. 10~20 uF should be fine.
-Chris
I agree, that's a typical symptom of using too high as smoothing cap for the rectifier concerned. The 470 ohm might protect the rectifier from the 40uF smoothing cap to some extent but 16uF would be safer.
So if I replace the 40uf with a 10uf I will be safe?
I have had this configuration (excluding the 11.7K and the 320uf) for a long time now with different circuits and I have never had a problem. Is this the tube telling me it has had enough?
I have had this configuration (excluding the 11.7K and the 320uf) for a long time now with different circuits and I have never had a problem. Is this the tube telling me it has had enough?
Hi lawbadman,
-Chris
😀 I'd say so. Replace the first cap with a 10uF. Your tube may like you again. Certainly other 5Y3's won't hide in the darkness when you come looking.
-Chris
This is the classic symptom of too much capacitance. I have one of my TubelabSE amplifiers that uses a 100uF input cap on a 5AR4. Yes this is too much. It was the only cap that I could find when I built this one, and I am too busy to change it. About once a month there is fireworks in the 5AR4 when I turn it on. The tube is an old RCA that I took out of another amp. I look at this as a controlled experiment to see how long it takes it to blow up. It has been in there for about almost two years, and hasn't fried yet. DON"T do this with an expensive tube, you know that it will blow up the first time you turn it on. This amp also has an undersized power transformer that probably helps the rectifier survive.
From RC-30, the max. hot-switching transient current for the 5Y3 is 2.5 amps per plate. The 470R resistor limits hot-switching current to less than one amp in this circuit, even if the capacitor is discharged as it would be at power up. As such, there shouldn't be any problem running a first capacitor of 40uF. As an aside, RC-14 specifically talks about using a 40uF first capacitor with a total effective plate-supply impedence of a meager 10R per plate, suggesting such a configuration is ok. I've been running a 5Y3 in my stereo SE 6V6 amp with 150R + 100uF for years with no problems. I suppose I should admit that amp has a somewhat anemic power transformer as well, but I don't think it matters all that much in this case... 😉
Of course, this *particular* tube seems to be having problems with the capacitance of the circuit, but I don't believe that the circuit is at fault. Dropping the value of the first capacitor will probably drop your B+ voltage a bit, which you may or may not care about. If you want to continue using *that* 5Y3 without otherwise altering the circuit, you could add a switch in series with the 470R (which you would leave off until the circuit has equalized) and another, much larger, resistor (say 10K) from the 5Y3 to the 40uF cap. My personal opinion, however, is that you should just get another 5Y3, and leave the circuit alone.
Of course, this *particular* tube seems to be having problems with the capacitance of the circuit, but I don't believe that the circuit is at fault. Dropping the value of the first capacitor will probably drop your B+ voltage a bit, which you may or may not care about. If you want to continue using *that* 5Y3 without otherwise altering the circuit, you could add a switch in series with the 470R (which you would leave off until the circuit has equalized) and another, much larger, resistor (say 10K) from the 5Y3 to the 40uF cap. My personal opinion, however, is that you should just get another 5Y3, and leave the circuit alone.
OK, here is an update on my power supply.
Today I finished building my 12b4 pre amp so I needed to test it. I replaced the 40uF with a 22uF cap, and took off a few other stuff which left:
5Y3GT----470ohm------22uF------3490ohm------220uf-----3300----82uF
I connected everything together and fear fully flipped the switch for a smoke test. Well everthing started out great, but then something happened which was quite unexpected. When the voltage (at the 82uF cap) got to about 190volts @ 35mA the transformer started buzzing, I turned it off as soon as I heard it. Good news is that there was no smoke.
So the question now is what caused the buzzing in the trafo?
I measured the secondarys and they measured 83 and 75 ohm, so I dont think that it burnt out. The weird thing about the trafo is that it is always hot (but not too hot that I can't touch it) even at low loads. The trafo is a Hammond 272BX by the way.
This power supply is getting to me now I figure I will just buy a new trafo and move to a nice SS power supply.
Any thoughts as to what might have happened?
Any suggestions are welcome.
Thanks all
Lawrence
Today I finished building my 12b4 pre amp so I needed to test it. I replaced the 40uF with a 22uF cap, and took off a few other stuff which left:
5Y3GT----470ohm------22uF------3490ohm------220uf-----3300----82uF
I connected everything together and fear fully flipped the switch for a smoke test. Well everthing started out great, but then something happened which was quite unexpected. When the voltage (at the 82uF cap) got to about 190volts @ 35mA the transformer started buzzing, I turned it off as soon as I heard it. Good news is that there was no smoke.
So the question now is what caused the buzzing in the trafo?
I measured the secondarys and they measured 83 and 75 ohm, so I dont think that it burnt out. The weird thing about the trafo is that it is always hot (but not too hot that I can't touch it) even at low loads. The trafo is a Hammond 272BX by the way.
This power supply is getting to me now I figure I will just buy a new trafo and move to a nice SS power supply.
Any thoughts as to what might have happened?
Any suggestions are welcome.
Thanks all
Lawrence
There's buzzing and there's buzzing. With a new filter setup, the current waveform will be different. Maybe the peaks are causing the buzz but it may still be ok. (it isn't a hard thumping buzz is it?)
As far as the earlier diagnosis, a rectifier that is worn will arc. Lowering the capacitance may help because the rectifier sees easier going, but such a rectifier is still worn. It just has more trouble giving out the current.
Could you look at the ps output on an oscilloscope?
As far as the earlier diagnosis, a rectifier that is worn will arc. Lowering the capacitance may help because the rectifier sees easier going, but such a rectifier is still worn. It just has more trouble giving out the current.
Could you look at the ps output on an oscilloscope?
Hi Lawrence,
You have lowered the peak current drawn, so that's easier on the transformer too. But your secondary resistance on the HT winding should be more equal. If you did measure this carefully with a meter up to the task, I'd say you have a shorted secondary.
Pull the rectifier and measure the AC voltage of the unloaded secondary winding. Be very cautious (measure from the socket pin, top side maybe). The voltages should be the same wrt ground. If not, it's new transformer time.
-Chris
You have lowered the peak current drawn, so that's easier on the transformer too. But your secondary resistance on the HT winding should be more equal. If you did measure this carefully with a meter up to the task, I'd say you have a shorted secondary.
Pull the rectifier and measure the AC voltage of the unloaded secondary winding. Be very cautious (measure from the socket pin, top side maybe). The voltages should be the same wrt ground. If not, it's new transformer time.
-Chris
Another thing to consider should the transformer prove not to have a secondary winding short is the type of 5Y3 you have. I have 5Y3 that have indirectly heated cathodes as well as ones that are filamentary types, if you have one of the indirectly heated types it is quite possible that the tube has a short/partial short between the cathode and plate of one of the sections. (This is a common failure in chinese made 5AR4 which is where I first encountered it, incidentally genie mullards can fail the same way.) Filamentary types usually are not seriously damaged by an occasional arc, although it may damage the filament locally.
Don't ditch the rectifier tube just because you are having a problem with this particular project. I use them in almost everything I design and have no problems at all.
One thing that helps is it put an UF4007 or similar 1Kpiv diode in series with each plate, all the advantages of thermionic rectification with some additional margin for piv issues and also seems to help arc over during warm up. For very high voltage applications near the max rating of the specific rectifier tube this can make a very big difference in reliability.
Don't ditch the rectifier tube just because you are having a problem with this particular project. I use them in almost everything I design and have no problems at all.
One thing that helps is it put an UF4007 or similar 1Kpiv diode in series with each plate, all the advantages of thermionic rectification with some additional margin for piv issues and also seems to help arc over during warm up. For very high voltage applications near the max rating of the specific rectifier tube this can make a very big difference in reliability.
Hello guys,
Thanks for all the help.
The power supply now works, I bought a new 5Y3 tube and tried it out and the power supply now works just fine, no sparks and no humming.
Thanks again
Lawrence
Thanks for all the help.
The power supply now works, I bought a new 5Y3 tube and tried it out and the power supply now works just fine, no sparks and no humming.
Thanks again
Lawrence
My contribution is thus quite late, but I would like to comment on the minimum rectifier plate impedance matter.
Let me firstly say that I totally respect the experience of members who had a similar experience as Lawbadman. But we have a problem, as some observed. The minimum plate (serie) impedance includes the power transformer winding resistance, which for the kind of supply using a 5Y3, I would guess would already have been in excess of 50 ohm. (If Lawbadman has the patience and don't mind to humour me, perhaps he could measure his transformer's secondary d.c. resistance and let me know.) It is also an impedance, somewhat higher than the d.c. resistance. Then, if I understood correctly, the 470 ohm would also have safeguarded the circuit completely, for almost any capacitance.
I would carefully say that I have in the past associated internal rectifier sparking more with physical problems in the rectifier than transformer voltages as low as 300V. Also, my experience with too high a capacitance (too low a serie impedance) mostly reflected in the rectifier overheating.
Lastly I must unfortunately concur with Kevinr regarding the quality of present-day rectifiers - to be pessimistic, also more general, more's the pity. I was shocked to find indirectly heated 5Y3s; the characteristics are totally different!
I notice other experienced members on this thread, who would perhaps consider commenting on the above.
With thanks.
Let me firstly say that I totally respect the experience of members who had a similar experience as Lawbadman. But we have a problem, as some observed. The minimum plate (serie) impedance includes the power transformer winding resistance, which for the kind of supply using a 5Y3, I would guess would already have been in excess of 50 ohm. (If Lawbadman has the patience and don't mind to humour me, perhaps he could measure his transformer's secondary d.c. resistance and let me know.) It is also an impedance, somewhat higher than the d.c. resistance. Then, if I understood correctly, the 470 ohm would also have safeguarded the circuit completely, for almost any capacitance.
I would carefully say that I have in the past associated internal rectifier sparking more with physical problems in the rectifier than transformer voltages as low as 300V. Also, my experience with too high a capacitance (too low a serie impedance) mostly reflected in the rectifier overheating.
Lastly I must unfortunately concur with Kevinr regarding the quality of present-day rectifiers - to be pessimistic, also more general, more's the pity. I was shocked to find indirectly heated 5Y3s; the characteristics are totally different!
I notice other experienced members on this thread, who would perhaps consider commenting on the above.
With thanks.
I have been able to (in desperation) reduce/prevent the tendency of a worn rectifier to arc, by increasing series resistance. Not enough info to draw a conclusion though.
5U4G started to spark!
Hi,
I have just joined diyAudio on the strength of reading your posting regarding the rectifier valve problem.
I have Quad Forty amps and they have been dogged with problems virtually from new. I won't bore you with the endless list of faults but one I would like some opinion on is this rectifier sparking. After reading what other members have suggested as being possible faults I never considered or thought possible. It’s made very interesting reading.
The rectifier valves in the Quads are 5U4G's and a good while ago I had to replace one due to sparking. This particular fault was due to a fractured connection inside the valve and replacing it put it right. However, there have been two successive cases of sparking but of sort described by one of the guys on the forum. When I first switch the amps on, just as the heater starts to produce enough heat to start the thermionic reaction a shower of miniature sparks (quite nice display) from the top of the metal structure. It doesn’t last more than a few seconds then it's gone and all works absolutely fine.
As I said before, there have been many faults with the amps and I have changed leaky and short capacitors BUT only for ones of exactly the same value, maybe slightly higher voltage but no difference in the capacitance itself.
Phew! Sorry it's been a long-winded post but I am a bit puzzled by this. If anyone can help out or at least point me in the right direction I would be more than grateful.
Many thanks,
Jules.
Hi,
I have just joined diyAudio on the strength of reading your posting regarding the rectifier valve problem.
I have Quad Forty amps and they have been dogged with problems virtually from new. I won't bore you with the endless list of faults but one I would like some opinion on is this rectifier sparking. After reading what other members have suggested as being possible faults I never considered or thought possible. It’s made very interesting reading.
The rectifier valves in the Quads are 5U4G's and a good while ago I had to replace one due to sparking. This particular fault was due to a fractured connection inside the valve and replacing it put it right. However, there have been two successive cases of sparking but of sort described by one of the guys on the forum. When I first switch the amps on, just as the heater starts to produce enough heat to start the thermionic reaction a shower of miniature sparks (quite nice display) from the top of the metal structure. It doesn’t last more than a few seconds then it's gone and all works absolutely fine.
As I said before, there have been many faults with the amps and I have changed leaky and short capacitors BUT only for ones of exactly the same value, maybe slightly higher voltage but no difference in the capacitance itself.
Phew! Sorry it's been a long-winded post but I am a bit puzzled by this. If anyone can help out or at least point me in the right direction I would be more than grateful.
Many thanks,
Jules.
Quad 40 sparking
As the Quad 40 is not as famous as some other amps, please post first resistor size, first capacitor size, second resistor size. If you can read the expected voltage of the b+ winding of the transformer off a print or something it would be cool. As a first suggestion, you could look at the production date for the first capacitor, usually a four digit number with the first two being the last two digits of the year made, and the second two digits being the number of the week. If the production date was more than 10 years ago, there may be an age problem of the capacitor. Then read the high voltage FAQ, discharge your cap through a resistor to ground, (heat shrink around the resistor) measure cap voltage to make sure it is zero or less than 1/2v, and measure the resistance of the first resistor to make sure it as advertised.
As the Quad 40 is not as famous as some other amps, please post first resistor size, first capacitor size, second resistor size. If you can read the expected voltage of the b+ winding of the transformer off a print or something it would be cool. As a first suggestion, you could look at the production date for the first capacitor, usually a four digit number with the first two being the last two digits of the year made, and the second two digits being the number of the week. If the production date was more than 10 years ago, there may be an age problem of the capacitor. Then read the high voltage FAQ, discharge your cap through a resistor to ground, (heat shrink around the resistor) measure cap voltage to make sure it is zero or less than 1/2v, and measure the resistance of the first resistor to make sure it as advertised.
- Status
- Not open for further replies.