I'm using a pair of Hammond 373BX transformers to create +-400V for an amp of mine. Both rails are full-wave rectified with 6cj3's and filtered with a CLCLC filter (100uf, 1.5h/56ohm, 100uf, 1.5/56ohm, 200uf). Now, the problem is that while one of them is dead silent and runs cool, the other hums loudly and heats up to at least about 50 deg. C.
The only difference between the two rails is that one only has to deliver about 100ma (that is the quiet transformer), while the other one is asked for almost 170ma. The question is, does anyone know any ways to go around that problem?
My understanding is that the charging current reaches over 700ma in peaks (that's what PSUD2 simulation tells me, I haven't actually measured it) and that saturates the core which causes the noise. But if I put a small series resistor from the rectifier to the first cap (say 100 ohm), and rearrange the filter like RCLC (100 ohm, 200uf, 1.5h/56ohm, 200uf), then the peak charging current should drop to approximately 550ma and possibly that would be enough to keep the transformer core from saturating. Of course, the filtering is going to be worse than in the original configuration, but I'm in as long as that makes the transformer happy.
Does anyone know how likely is such line of thought or maybe I'm looking for a problem in the wrong place? Are there some other things I could try to quiet it down and keep it from overheating at the same time? Gurus, speak up.
The only difference between the two rails is that one only has to deliver about 100ma (that is the quiet transformer), while the other one is asked for almost 170ma. The question is, does anyone know any ways to go around that problem?
My understanding is that the charging current reaches over 700ma in peaks (that's what PSUD2 simulation tells me, I haven't actually measured it) and that saturates the core which causes the noise. But if I put a small series resistor from the rectifier to the first cap (say 100 ohm), and rearrange the filter like RCLC (100 ohm, 200uf, 1.5h/56ohm, 200uf), then the peak charging current should drop to approximately 550ma and possibly that would be enough to keep the transformer core from saturating. Of course, the filtering is going to be worse than in the original configuration, but I'm in as long as that makes the transformer happy.
Does anyone know how likely is such line of thought or maybe I'm looking for a problem in the wrong place? Are there some other things I could try to quiet it down and keep it from overheating at the same time? Gurus, speak up.
I feel sorry for the rectifier valves, feeding 100uF. I wonder if they will fail prematurely...
The best way of reducing the charging current peaks is to go for choke input: LC, but of course the output voltage will be lower.
One thought: Are the chokes in good condition? Or are they old? If the choke is saturating, the charging peaks will be even higher than you calculated.
Presumably this is for a valve amp?
I've moved the thread to the Tubes Forum, leaving a redirect in the original forum, where there might be a better response
The best way of reducing the charging current peaks is to go for choke input: LC, but of course the output voltage will be lower.
One thought: Are the chokes in good condition? Or are they old? If the choke is saturating, the charging peaks will be even higher than you calculated.
Presumably this is for a valve amp?
I've moved the thread to the Tubes Forum, leaving a redirect in the original forum, where there might be a better response
these damper diodes are rated at over 2A peak plate current and 350ma average, so as far as I know I'm not even close to maxing them out. they do produce a faint whooshing/ringing sound, but mine seem to do that even when the power's only applied to their filaments.
I thought of the choke input, but it would bring the voltage down to under 300VDC which unfortunately won't do. I'll try the resistor input as a quick fix today and see if it makes any noticeable difference.
I thought of the choke input, but it would bring the voltage down to under 300VDC which unfortunately won't do. I'll try the resistor input as a quick fix today and see if it makes any noticeable difference.
To an old salt like myself , the overheating 50°C (25°C above t ambient) seems like a shorted turn syndrome in mains tranny..(we won't discuss the choke out at this stage)...I cannot discount the possibility that...a part shorted turn in biphase secs ......check rectifier anode voltages wrt.gnd. this will be high AC and might be ringing.. BE CAREFULL........they should be similiar readings....take previous postings advice but..... The winding resistances will be different but you could compare with good one as you mentioned that you have a pair of trannies with same voltages. Do a hardware swap A/B and recheck. I know it's rotten but it's the Simplest way.
Note down carefully what you find between the two trannies. remember that biphase secondaries will not give the same resistance per half as one half winding is wound above the other.....but its' the number of turns that counts.
The tube rectifiers slam-driving 100uF caps isn't on........are you sure youv'e read them right and not actually 10uF ?
rich
Note down carefully what you find between the two trannies. remember that biphase secondaries will not give the same resistance per half as one half winding is wound above the other.....but its' the number of turns that counts.
The tube rectifiers slam-driving 100uF caps isn't on........are you sure youv'e read them right and not actually 10uF ?
rich
if that were the case, the transformer would make a racket even without any load, right? yet, it only becomes noisy when both the rectifier and the output tubes wake up. I'll try swapping the transformer loads if a resistor before the cap doesn't do anything.richwalters said:
heh, I know that 100uf (it is 100uf: half of 100u+100u/500v bg wkz) sounds a little heavy on the tube rectifiers, but weren't tv damper diodes designed to be slammed with all their might thousands times a second? just checking
.
Okay problems on warm up.......the reason I mentioned winding problems, is that it once happened to me without the noise drama. However, your problem seems continuous, and mine didn't survive long. True SS diodes can be slammed, but remember if your core is really 50°C.... it's guts must be stewing' . uumm, I'm still thinking...those caps ?........rich
ok, so there were 2 parts to this problem:
1) the big 373bx transformer itself (still noisy),
2) a small 36vct transformer installed right underneath it used to generate +-15vdc @ 10ma.
originally I totally forgot about that smaller transformer, and only recently I discovered that it was contributing most of the noise. anyway, I replaced it with a 186f28 that I conveniently had lying around (100va, heh heh), so that source on annoyance is gone. but 373bx is still noisy and still gets hot. *sigh*
I measured AC on the halves of the secondary and got different values -- ~342vac on one half and ~335vac on the other. the other transformer measures ~342vac on both halves. so I guess it's a shorted turn after all. I suppose there's no nice and easy way to repair something like this, is there?
1) the big 373bx transformer itself (still noisy),
2) a small 36vct transformer installed right underneath it used to generate +-15vdc @ 10ma.
originally I totally forgot about that smaller transformer, and only recently I discovered that it was contributing most of the noise
. anyway, I replaced it with a 186f28 that I conveniently had lying around (100va, heh heh), so that source on annoyance is gone. but 373bx is still noisy and still gets hot. *sigh*I measured AC on the halves of the secondary and got different values -- ~342vac on one half and ~335vac on the other. the other transformer measures ~342vac on both halves. so I guess it's a shorted turn after all. I suppose there's no nice and easy way to repair something like this, is there?
This may sound whacky, but you said that one transformer is right below the other. If so and it is out of phase (or in phase) with its neighbor, the two clashing magnetic feilds may be bucking and causing heat and noise. Try to reverse polarity on one primary and see if it changes the noise level.
I found out that it hums even without HV secondary connected (having one filament winding is enough). yah, I can believe that they may have missed a turn and I guess that could technically put some unbalanced dc on the secondary. I'll look into that.Magura said:
and yeah, the transformers are (were) brand new before I put them in the amp.
yah, I thought about that too, so right now I'm running wires to an external LV transformer.Brian Donaldson said:
ok, I'm retarded. going on a hunch, I decided to replace the diodes with some new ones (they are $1.25 a pop, so I have a handful of them) and color me surprised when hum moved to another transformer! I looked down through the socket (amp's upside down for testing/taking measurements) and noticed that one of the diodes in guilty transformer's bridge is not glowing. I wiggled it, it lit up, and voila -- the hum is mostly gone.
isn't this lovely?
(admittedly, there's still some residual buzzing but it's very faint. I'll figure it out.)
so, now I suspect that the problem's not in the transformers or diodes but rather in the poor contact between the diode tubes and their sockets (full-wave becomes half-wave => tons of unbalanced DC current through the secondary => hum and heat). 6cj3 pins look rather oxidized and while the socket contacts are still clean, at some point they'll tarnish too. is it possible to treat them somehow so that neither would oxidize as much?
isn't this lovely?
(admittedly, there's still some residual buzzing but it's very faint. I'll figure it out.)
so, now I suspect that the problem's not in the transformers or diodes but rather in the poor contact between the diode tubes and their sockets (full-wave becomes half-wave => tons of unbalanced DC current through the secondary => hum and heat). 6cj3 pins look rather oxidized and while the socket contacts are still clean, at some point they'll tarnish too. is it possible to treat them somehow so that neither would oxidize as much?
Rewinding mains transformers......I've done this many times, but HT windings are usually next to primary; i.e well into bobbin and that means removing all the other windings on top. It's false economy re using unwound wire. If the tranny is vacuum varnished impregnated; core and iron; the going-is-going to be tough splitting laminations from the core, especially if they have been press machine inserted.
Warming the core many help. The best tool to wedge E & I laminations from each other is an old thin pocket sized steel ruler with an end slightly sharpened and a small hobby hammer and work alternate ends... Invariably one looses the first set. Once the bobbin has been separated from the core, you will need to set up a marndrel to count the turns.
It's not the nicest new-year-job for anyone.......especially on a cherished tranny......
rich
Warming the core many help. The best tool to wedge E & I laminations from each other is an old thin pocket sized steel ruler with an end slightly sharpened and a small hobby hammer and work alternate ends... Invariably one looses the first set. Once the bobbin has been separated from the core, you will need to set up a marndrel to count the turns.
It's not the nicest new-year-job for anyone.......especially on a cherished tranny......
rich
Your problem might be...
Are you are using American TV efficiency diodes in European sockets? The European B9D socket has exactly the same pin positions as the American Magnoval, but the European pin diameter is slightly larger, causing all sorts of problems if you accidentally use (for example) a 12CL3 in a socket intended for an E55L.
Are you are using American TV efficiency diodes in European sockets? The European B9D socket has exactly the same pin positions as the American Magnoval, but the European pin diameter is slightly larger, causing all sorts of problems if you accidentally use (for example) a 12CL3 in a socket intended for an E55L.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.