Hello,
I just plugged in my newly built double mono preamp and after a few minutes one of the trafo's started smoking... (B+ measured as expected however)
I unwired it and tested it without a load; everything fine (all the voltages measured fine); seems it had been the plastic covering the windings that was smoking and not the internal isolation. (lucky!!!)
Rewired it again and after a few minutes it began to smoke again...
Now: my problem is that I really don't understand why the windings get so hot (which is the real illness, smoking plastic is just the symptom). Both trafo's are identical besides the plastic covering the windings, that probably explains why one trafo smokes and the other doesn't (they both seem to get quite warm instantaniously). I used one of the trafo's before in a el84se pentode integrated with silicon diodes that drawed a lot more current + the filament winding isn't being stressed either. A el84se draws around 100mA; this preamp (12b4) draws 30mA = 15mA for each trafo (= 15% of the el84 current draw but the trafo gets much hotter)!!!
I'll give a short description of the setup as I'm not an electrical expert and maybe the problem is clear to someone reading:
HT winding is 125-0-125V; each anode of the rectifier tube is connected to one of the 125V taps (because I need 250V), two diodes are used to form a bridge together with the tube and supply the - . The rectifiers are AZ1 (electrically similar to 5y3: big voltage drop + not too high current rating).
Don't think the mistake could be located further in the circuit - will be checking everything out again (although I'm rather sure all is wired as should be...) / The diodes & rectifier tubes are wired correctly; tubes are tested etc.
If this rings a bell, any help would be extremely welcome 🙂
I just plugged in my newly built double mono preamp and after a few minutes one of the trafo's started smoking... (B+ measured as expected however)
I unwired it and tested it without a load; everything fine (all the voltages measured fine); seems it had been the plastic covering the windings that was smoking and not the internal isolation. (lucky!!!)
Rewired it again and after a few minutes it began to smoke again...
Now: my problem is that I really don't understand why the windings get so hot (which is the real illness, smoking plastic is just the symptom). Both trafo's are identical besides the plastic covering the windings, that probably explains why one trafo smokes and the other doesn't (they both seem to get quite warm instantaniously). I used one of the trafo's before in a el84se pentode integrated with silicon diodes that drawed a lot more current + the filament winding isn't being stressed either. A el84se draws around 100mA; this preamp (12b4) draws 30mA = 15mA for each trafo (= 15% of the el84 current draw but the trafo gets much hotter)!!!
I'll give a short description of the setup as I'm not an electrical expert and maybe the problem is clear to someone reading:
HT winding is 125-0-125V; each anode of the rectifier tube is connected to one of the 125V taps (because I need 250V), two diodes are used to form a bridge together with the tube and supply the - . The rectifiers are AZ1 (electrically similar to 5y3: big voltage drop + not too high current rating).
Don't think the mistake could be located further in the circuit - will be checking everything out again (although I'm rather sure all is wired as should be...) / The diodes & rectifier tubes are wired correctly; tubes are tested etc.
If this rings a bell, any help would be extremely welcome 🙂
First you must identify the windings! 
which you obviously haven't done yet. 
Have you got any data of the transformer? If yes, have checked the output voltage before you connect any load?
Do you have any sketch of your wiring?
which you obviously haven't done yet. Have you got any data of the transformer? If yes, have checked the output voltage before you connect any load?
Do you have any sketch of your wiring?
Hello,
The voltages are marked on the trafo. Without a load, they all measure as should.
What do you mean by 'First you have to identify the windings'?
I have no sketch of the wiring; the schematic comes from diyparadise.com & I carefully designed the ps myself...
The voltages are marked on the trafo. Without a load, they all measure as should.
What do you mean by 'First you have to identify the windings'?
I have no sketch of the wiring; the schematic comes from diyparadise.com & I carefully designed the ps myself...
Mmmh, you tell about TWO trafos ?
How are they connected together (if they are) ?
The center tap of the HT winding is floating ... of course.
Yves.
How are they connected together (if they are) ?
The center tap of the HT winding is floating ... of course.
Yves.
Hello Yvesm,
Anode a of rectifier tube one is connected to anode b of rectifier two, and vice versa (I know I did this right because I have another dual mono amp with identical wiring that runs perfect / That one didn't have a center tap in the HT winding though).
The center taps of both trafo's aren't connected to anything (or to each other)// is this the magical solution? Should I keep the other connections as described above when also connecting the center taps with each other?
MANY MANY THANKS!!!
Anode a of rectifier tube one is connected to anode b of rectifier two, and vice versa (I know I did this right because I have another dual mono amp with identical wiring that runs perfect / That one didn't have a center tap in the HT winding though).
The center taps of both trafo's aren't connected to anything (or to each other)// is this the magical solution? Should I keep the other connections as described above when also connecting the center taps with each other?
MANY MANY THANKS!!!
Hi !
HV center taps must be left floating.
I still don't understand how you connetc the TWO tranfos.
Are they completly independent or have something in common ?
Certainly a schemo would be helpfull !
Just a though: have you checked if center taps wasn't internally tied to something else ?
😕 😕
Yves.
HV center taps must be left floating.
I still don't understand how you connetc the TWO tranfos.
Are they completly independent or have something in common ?
Certainly a schemo would be helpfull !
Just a though: have you checked if center taps wasn't internally tied to something else ?
😕 😕
Yves.
Hello,
I'll explain a little better:
I use a bridge rectifier (one for each channel) that consists of two silicon diodes followed by a rectifier tube. The anodes of each tube are connected to the opposite anodes of the other tube (think this is for balancing, found it on audiodesignguide.com but can't seem to find it right now).
The center taps of the two transformers are 'floating' (not connected to anything) - I measured without load and they measure as should; no problem here. (125volts to each of the two 125v taps)
The two power supplies are dual mono, but after the last ps capacitor I connected the two grounds with each other + the circuit ground.
Hope this helps and you can help me!
I'll explain a little better:
I use a bridge rectifier (one for each channel) that consists of two silicon diodes followed by a rectifier tube. The anodes of each tube are connected to the opposite anodes of the other tube (think this is for balancing, found it on audiodesignguide.com but can't seem to find it right now).
The center taps of the two transformers are 'floating' (not connected to anything) - I measured without load and they measure as should; no problem here. (125volts to each of the two 125v taps)
The two power supplies are dual mono, but after the last ps capacitor I connected the two grounds with each other + the circuit ground.
Hope this helps and you can help me!
Klimon said:
Do you mean you use 2 tubes in // ? ? ? in each channel ? ? ?
The silicon diodes must have their anodes tied together and to the ground. Right ?
Yves.
I use one tube in each channel + anodes of the silicons tied together🙂
But: the anode 1 of tube 1 (left channel) is connected to anode 1 of tube 2 (right channel) & anode 2 of tube 1 is tied to anode 2 of tube 2 // quite sure the mistake isn't to be found here... (seems like I've put the power trafo secondaries in parallel - don't know if that's good or bad however)
What startles me is that B+ voltages (measured at the last cap of each channel) are very near what I expected & simulated (around 270V) --- If the voltage is right and the trafo's get too hot (= VAC too high), the current draw must be too high - but where is that current going???
Thanks for thinking along🙂
But: the anode 1 of tube 1 (left channel) is connected to anode 1 of tube 2 (right channel) & anode 2 of tube 1 is tied to anode 2 of tube 2 // quite sure the mistake isn't to be found here... (seems like I've put the power trafo secondaries in parallel - don't know if that's good or bad however)
What startles me is that B+ voltages (measured at the last cap of each channel) are very near what I expected & simulated (around 270V) --- If the voltage is right and the trafo's get too hot (= VAC too high), the current draw must be too high - but where is that current going???
Thanks for thinking along🙂
... please don't do any more "smoke" 😀 tests on the circuit and post a schematic, i think it will be much easier to answer your question then...
Greetings,
Andreas
Greetings,
Andreas
Hello,
thought about it a little, then removed the two wires that parallelled the secundaries of both trafo's (because that was clearly the error; and a stupid one).
Andreas; I didn't perform the smoke-test but a more sophisticated and safer version to verify the result: I put my finger against the trafo and waited; hoping it wouldn't get hot...
All runs cool as a popsickle now
Thanks for helping an idiot finding his way back home
thought about it a little, then removed the two wires that parallelled the secundaries of both trafo's (because that was clearly the error; and a stupid one).
Andreas; I didn't perform the smoke-test but a more sophisticated and safer version to verify the result: I put my finger against the trafo and waited; hoping it wouldn't get hot...
All runs cool as a popsickle now
Thanks for helping an idiot finding his way back home
You should remove those connections between the anodes of the rectifiers on the left and right channel. If I read you correctly they are effectively connected in parallel and small differences in the winding ratios will cause large currents to flow between them causing the overheating symptom you describe. 
The transformers should be totally independent of each other - it is considered very bad design practice to parallel power transformers as there may be small differences in the way they are wound which will then result in large circulating currents. This is not the same as paralleling windings on a single transformer designed for this type of connection.
I would remove those connections on the power amplifier as well incidentally. Connecting them together totally defeats the purpose of dual monaural construction - if that was the intent you would be better off with a single power transformer with twice the secondary current rating.
Kevin
Edited for improved clarity
The transformers should be totally independent of each other - it is considered very bad design practice to parallel power transformers as there may be small differences in the way they are wound which will then result in large circulating currents. This is not the same as paralleling windings on a single transformer designed for this type of connection.
I would remove those connections on the power amplifier as well incidentally. Connecting them together totally defeats the purpose of dual monaural construction - if that was the intent you would be better off with a single power transformer with twice the secondary current rating.
Kevin
Edited for improved clarity
That's exactly what happened... I'm learning all sorts of stuff firsthand... It did sound better when I still believed it was dual mono 
What about parallelling 2 independent windings in a transformer?
I have a 300-0-300V @ 60mA transformer and from PSUII I worked out the current requirement of the preamp circuit for the transformer to be exactly 60mA which is the limit of the voltage of the winding. I think I will use a bridge in which case I only need 0-300V. I am thinking about parallelling the 0-300 and 0-300 to double the current capacity but don't know if there is any risk since there must be differences in the windings.
Regards,
Bill
I have a 300-0-300V @ 60mA transformer and from PSUII I worked out the current requirement of the preamp circuit for the transformer to be exactly 60mA which is the limit of the voltage of the winding. I think I will use a bridge in which case I only need 0-300V. I am thinking about parallelling the 0-300 and 0-300 to double the current capacity but don't know if there is any risk since there must be differences in the windings.
Regards,
Bill
It can and will be dual mono if you have two separate power transformers with no connections between the secondaries. In any case it is a bad idea to connect the secondaries of separate transformers together.
In the case of a single transformer with dual secondaries this can be done if the transformer was designed for this kind of operation. Most toroids for example are designed for parallel operation of secondaries - when in doubt check the manufacturer's spec..
Kevin
In the case of a single transformer with dual secondaries this can be done if the transformer was designed for this kind of operation. Most toroids for example are designed for parallel operation of secondaries - when in doubt check the manufacturer's spec..
Kevin
Hi Bill,
Why not try using the two 0 - 300V windings to power separate power supplies for each channel. You will get the benefit of dual mono construction and should be operating well within the ratings of the individual windings without the possible risk of circulating currents causing overheating if the transformer was not designed for paralleled secondaries.
Kevin
Why not try using the two 0 - 300V windings to power separate power supplies for each channel. You will get the benefit of dual mono construction and should be operating well within the ratings of the individual windings without the possible risk of circulating currents causing overheating if the transformer was not designed for paralleled secondaries.
Kevin
Kevin,
I like your idea of dual mono. However it is a choke input PSU which means I must have two chokes. The chasis is too small for that. How about having two separate bridges for the two windings then bring the output to the same choke? Would it have the same problem as in the difference caused by the windings?
For my 12B4A preamp, the transformer runs very hot - about 39mA on the 380-0-380 winding, 2A on the 5VAC winding for the 5AR4, and 2.7A on the series of a 0-6.3VAC and a 0-5VAC (both rated 3A) windings to provide power for the LM317 regulated 6.3VDC filament running at 1.3A.
Since it runs so hot, I suspect the cause to be the series connection of the 0-6.3VAC and the 0-5VAC windings.
I am now removing the 5AR4 and will use 1n4007 so that I can save the 2A@5VAc filament to cool it down a bit. I will also use the 1n5822 to replace 1n54XX so that I can use the series connection of the 6.3-0-6.3@2A windings for the 12B4A filament (regulated 12VDC), instead of the 0-5VAC and 0-6.3VAC (regulated 6.3VDC). This will reduce the current consumption on the transformer and I think it is going to work.
My problem is that I am building the Aikido preamp and I have on hand a transformer 300-0-300VAC@60mA, 0-12VAC@2A, 0-15-18VAC@2A, 0-15-18@2A I don't want to waste. Since the Aikido will consume 30mA so one winding of 0-300 is just enough for the job (need to regulate and smooth it down to 220VDC) therefore, I am thinking about paralleling the 0-300 windings.
Another big headache I have is for the 6n1p filament supply 6.3V@1.25A. Using the 0-12VAC@2A simply can't do the work when the voltage is brought down to about 10V the current requirement on the transformer is about 2.7A. One thing that will work is to parallelling the 0-15V@2A and bright it down for the 6.3V filament. This wastes hugh amount of power and may get the transformer running hotter.
Now I am thinking about running the heater on series using a single 0-15VAC winding using LM317 as a constant current source. This would be far more efficient. Would I raise the heater voltage in the same way when using a constant current source?
Regards,
Bill
I like your idea of dual mono. However it is a choke input PSU which means I must have two chokes. The chasis is too small for that. How about having two separate bridges for the two windings then bring the output to the same choke? Would it have the same problem as in the difference caused by the windings?
For my 12B4A preamp, the transformer runs very hot - about 39mA on the 380-0-380 winding, 2A on the 5VAC winding for the 5AR4, and 2.7A on the series of a 0-6.3VAC and a 0-5VAC (both rated 3A) windings to provide power for the LM317 regulated 6.3VDC filament running at 1.3A.
Since it runs so hot, I suspect the cause to be the series connection of the 0-6.3VAC and the 0-5VAC windings.
I am now removing the 5AR4 and will use 1n4007 so that I can save the 2A@5VAc filament to cool it down a bit. I will also use the 1n5822 to replace 1n54XX so that I can use the series connection of the 6.3-0-6.3@2A windings for the 12B4A filament (regulated 12VDC), instead of the 0-5VAC and 0-6.3VAC (regulated 6.3VDC). This will reduce the current consumption on the transformer and I think it is going to work.
My problem is that I am building the Aikido preamp and I have on hand a transformer 300-0-300VAC@60mA, 0-12VAC@2A, 0-15-18VAC@2A, 0-15-18@2A I don't want to waste. Since the Aikido will consume 30mA so one winding of 0-300 is just enough for the job (need to regulate and smooth it down to 220VDC) therefore, I am thinking about paralleling the 0-300 windings.
Another big headache I have is for the 6n1p filament supply 6.3V@1.25A. Using the 0-12VAC@2A simply can't do the work when the voltage is brought down to about 10V the current requirement on the transformer is about 2.7A. One thing that will work is to parallelling the 0-15V@2A and bright it down for the 6.3V filament. This wastes hugh amount of power and may get the transformer running hotter.
Now I am thinking about running the heater on series using a single 0-15VAC winding using LM317 as a constant current source. This would be far more efficient. Would I raise the heater voltage in the same way when using a constant current source?
Regards,
Bill
Hi Bill,
You could try a separate bridge rectifier for each 300V winding, but you might just want to consider a conventional center tapped full wave rectifier instead, this into a choke input should be ok if anything is.
I would recommend shifting some of that filament load onto a separate transformer. (Probably no room)
I would recommend ditching the LM317 for the 6.3V regulator not because it is a lousy device (it's not) - it's just that an LT1084 will operate properly with less than 1V across it which would probably allow you to run the filaments off of the 6.3V winding with a schottky bridge. No need for the 5V winding and the reduced power consumption in the transformer will result in cooler operation. Depending on winding resistance and rectifier losses the raw dc should be around 8V at nominal line which is more than enough to account for the drop out voltage of the LT1084 and 10% low line. In the old days (and this I'm not recommending!) it was not unusual to run a filament winding slightly beyond its ratings in icas service if one of the other filament windings was not being used.
Incidentally my usual rule of thumb for rectified filament windings is dc load should be < 75% of the AC rms current rating of that winding to roughly compensate for the non-sinusoidal waveform's heating effects. Larger filter capacitors with their increased charging current will increase the heating effect, if there is a lot of voltage margin several volts of ripple may be tolerated on the unregulated side in return for lower peak currents in the winding and less heating.
For the Aikido I would use both windings configured for 15V into a full wave center tapped rectifier and regulate that either in constant current mode or voltage mode at 12.6V.
Kevin
You could try a separate bridge rectifier for each 300V winding, but you might just want to consider a conventional center tapped full wave rectifier instead, this into a choke input should be ok if anything is.
I would recommend shifting some of that filament load onto a separate transformer. (Probably no room)
I would recommend ditching the LM317 for the 6.3V regulator not because it is a lousy device (it's not) - it's just that an LT1084 will operate properly with less than 1V across it which would probably allow you to run the filaments off of the 6.3V winding with a schottky bridge. No need for the 5V winding and the reduced power consumption in the transformer will result in cooler operation. Depending on winding resistance and rectifier losses the raw dc should be around 8V at nominal line which is more than enough to account for the drop out voltage of the LT1084 and 10% low line. In the old days (and this I'm not recommending!) it was not unusual to run a filament winding slightly beyond its ratings in icas service if one of the other filament windings was not being used.
Incidentally my usual rule of thumb for rectified filament windings is dc load should be < 75% of the AC rms current rating of that winding to roughly compensate for the non-sinusoidal waveform's heating effects. Larger filter capacitors with their increased charging current will increase the heating effect, if there is a lot of voltage margin several volts of ripple may be tolerated on the unregulated side in return for lower peak currents in the winding and less heating.
For the Aikido I would use both windings configured for 15V into a full wave center tapped rectifier and regulate that either in constant current mode or voltage mode at 12.6V.
Kevin
kevin,
Thank you for your advice. I am now considering using the conventional center tapped full wave rectifier instead of a bridge for both the 12B4A and the Aikido. In that case the load on the transformer is brought more even because I use separate windings for each half of the cycle and I can avoid the differences in the winding resistence if parallelling them, etc, though for each half cycle one winding is still reaching its current limit as the current demand happens to match exactly the current rating of the winding.
Now I have to look for some diodes that have higher voltage ratings than the 1n4007 since the diodes for the 12B4A (Tx 380-0-380VAC) must exceed 1000V. What do you recommend? I think I may use Shottkey diodes but if according to M.J.'s book I will need to use multiples connected in series - expensive also increasing the impedance. I don't know if any new Shottkey diodes will do the job.
For the filament supply, thanks for suggesting the LM180X regulators. I have no problems with the filaments for the 12B4A and 5687. It is only for the 6n1p that I don't have lower voltage windings. I am now considering using the LM317 for a constant current load as per M.J.'s book so that I can use the 15V or 18V winding. That saves me a transformer.
Regards,
Bill
Thank you for your advice. I am now considering using the conventional center tapped full wave rectifier instead of a bridge for both the 12B4A and the Aikido. In that case the load on the transformer is brought more even because I use separate windings for each half of the cycle and I can avoid the differences in the winding resistence if parallelling them, etc, though for each half cycle one winding is still reaching its current limit as the current demand happens to match exactly the current rating of the winding.
Now I have to look for some diodes that have higher voltage ratings than the 1n4007 since the diodes for the 12B4A (Tx 380-0-380VAC) must exceed 1000V. What do you recommend? I think I may use Shottkey diodes but if according to M.J.'s book I will need to use multiples connected in series - expensive also increasing the impedance. I don't know if any new Shottkey diodes will do the job.
For the filament supply, thanks for suggesting the LM180X regulators. I have no problems with the filaments for the 12B4A and 5687. It is only for the 6n1p that I don't have lower voltage windings. I am now considering using the LM317 for a constant current load as per M.J.'s book so that I can use the 15V or 18V winding. That saves me a transformer.
Regards,
Bill
- Status
- Not open for further replies.