Hi guys,
Been thinking about this for quite some time now. What are the causes that damages/destroys an output transformer?
If it's not a good idea to run a tube amp without a speaker, what happens inside the amp? For example, if someone accidentally trips on the speaker wire and nobody notices that one of the channels of a tube amp is running without a load, will the opt be damaged? How long does it take for this to happen? Why?
TIA,
JojoD
Been thinking about this for quite some time now. What are the causes that damages/destroys an output transformer?
If it's not a good idea to run a tube amp without a speaker, what happens inside the amp? For example, if someone accidentally trips on the speaker wire and nobody notices that one of the channels of a tube amp is running without a load, will the opt be damaged? How long does it take for this to happen? Why?
TIA,
JojoD
Yes, open output can kill output transformers. Happened to me once during some testing on a DIY guitar amp.
Well, I switched on and played some guitar while waiting for the tubes to warm up and the sound to appear in the speaker, but the last thing never happened... I guess it died quick and painless, when I checked the primary winding was an open circuit.
That was very quick, not enough warning to stop it from being dead. Was the opt tranny the only casualty? Power tubes ok?
Hope someone explains the forces within which causes such opt failures.
Hope someone explains the forces within which causes such opt failures.
I soldered directly on the output xformer's secondary a 100-200 Ohm/2W resistor, to avoid such problems.
Transformers are inductors, going open can cause a very high voltage to develop. This may destroy the insulation. This effect may be stronger with better cores.
A gassy tube going into runaway is another possible OPT killer.
A gassy tube going into runaway is another possible OPT killer.
Another thing that can kill an output transformer is excessive plate current. Usually caused by bad biasing. Leaky coupling caps that couple from a driver's plate can misbias the control grid and make the tube turn on hard and run too much current thru the plate and thus the transformer. So replace any wax paper caps you see on sight, no questions asked, they are all bad as they are way beyond their service life today. They absorb moisture from the air, and the acid in the paper causes corrosion of the metal foil inside and thus leakage paths. Use your favorite poly-whatever caps of similar capacitence and same or higher voltage rating. No ceramic caps, they are slightly piezeoelectric and vary slightly in capacitence vs voltage across them, and introduce distortion products that you don't want in your hifi amp.
Loss of speaker load will present just the primary inductance of the OPT to the tubes.... This inductive load has NO DAMPING....
It presents "idealy" an infinite flat load line..... with a resonant frequency...SO there will be no current swing...but instead will produce a BIG voltage swing....
Therefore it will have a HUGE flyback voltage due to no damping co-efficient that the plate load "would" have presented...
These very high flyback voltages WILL break down the wire insulation as well as "punch-through" effect where the AC voltage gradient will burn through the insulation layer.... Once this happens you get a shorted turn somewhere in the transformer...after that , its all down hill !!!!!
I treat amp/transformer safety the same as gun safety at the range....You check EVERY connection and you check it AGAIN before turning the amp on...NEVER get complacient and never change your set-up routine......
Chris
It presents "idealy" an infinite flat load line..... with a resonant frequency...SO there will be no current swing...but instead will produce a BIG voltage swing....
Therefore it will have a HUGE flyback voltage due to no damping co-efficient that the plate load "would" have presented...
These very high flyback voltages WILL break down the wire insulation as well as "punch-through" effect where the AC voltage gradient will burn through the insulation layer.... Once this happens you get a shorted turn somewhere in the transformer...after that , its all down hill !!!!!
I treat amp/transformer safety the same as gun safety at the range....You check EVERY connection and you check it AGAIN before turning the amp on...NEVER get complacient and never change your set-up routine......
Chris
I tend to agree with cerrem, an open load will cause a high voltage to appear across the primary of the OPT. If the amp is driven hard enough so that one of the output tubes cuts off (common in guitar amps) the voltage rise is theoretically infinite. In practice the voltage will rise until SOMETHING breaks down (kilovolts). The common things that fail are the output tubes, the sockets, and the OPT (I have blown them all). Once an arc starts (even a small wimpy arc) it is now fed by the full energy of the power supply. I have seen output transformers catch on fire (think early Ampeg SVT). In powerful amps the death is instant and often violent, sometimes taking out the OPT, one or more tubes, and the power transformer.
It has been stated that this will not happen in an SE amp since the current in the output transformer is constant. This may be true under normal conditions, but again if the amp is driven hard enough to cut the tube off even for a microsecond, an arc can start. Believe me, I have blown more than one SE OPT (and a tube socket) from open or incorrect loads.
This may be true, but 50+ year old tinkerers that often act like kids can blow up a lot of stuff too. How do I know this?
There things that can be done to help the situation, but nothing can make a tube amp totally foolproof. I have used all of these with success, but each must be tweaked to the particular amp design and power level. For all except the resistor, the trick is to adjust the value of the protection device such that is does NOT conduct when the amp is operated with the highest possible load impedance at full rated power, hard clipping if that can be expected in normal operation (guitar amp, or teenage kids).
1) The previously mentioned big resistor. I use a 100 ohm 5 watt wirewound resistor directly across the speaker terminals on low powered (typically SE) guitar amps. This will cost some power but it works for most cases, and doesn't affect the sound even at full clip with weird loads. If you intentionally mis load the amp (guitar amp) to get a particular sound, a resistor may affect the sound.
2) Wire a "transient protection diode" of the bidirectional type directly across the speaker. The diode must be sized so that it NEVER conducts in normal operation. If it does the sound will be bad (no, gross) and the diode will get hot. In higher powered amps many diodes can be paralleled. The poor diode will have to eat the entire output of the amp if the speaker is removed, but if the diode fails they usually short, saving the OPT. Use the 1.5 KW (peak) rated diodes. I use the 24 volt flavor in my "Turbo Champ", an 8 watt SE guitar amp.
3) CP Claire, Seimens and others used to make gas spark arrestors for use in phone line equipment. These come in a few different voltage ratings. They seem to be able to take the output of a 50 watt guitar amp for several minutes before MELTING! I don't know if they are still common.
4) I have seen people connect zener strings and even VR tubes across the OPT primary. I tried this with a zener string, and it burst into flames instantly when I unhooked the speaker on a 50 watt P-P amp. The VR tube idea may work, but I have not tried it.
5) this is unpopular, but large amounts of negative feedback from the OPT secondary can tame the fire gods.
I will often use multiple methods together to design an almost foolproof amp. In a high power guitar amp I will use a big resistor and 2 or 3 gas discharge tubes across the speaker. I will also wire a bidirectional zener or a neon bulb from the 16 ohm tap back through a voltage divider to provide a large amount of negative feedback only when the bulb (or zener) fires. This is independent of the main feedback loop and applied to the driver or pre driver stage.
Those who know me (or have followed my experiments) understand that I tend to test anything that I intend to sell under extreme (often destructive) conditions. I believe (and it has been confirmed) that a customer (especially for a guitar amp) will find new and unique ways to blow things up that I could never imagine (or test for). All of my HiFi amps must live for a few hours with me playing guitar through them with about 10 db of overdrive. Don't try this with your expensive 300B amp.
As mentioned before an improperly set up amp (or one with bad caps) can slowly cook an OPT or power transformer to death. This will usually fry output tubes first. Sometimes a bad or low quality output tube can arc causing the death of an OPT, but this is not common. Old (vintage) OPT's can absorb moisture and loose insulation quality from years of heat soak, and just fail, even though there is nothing else wrong with the amp.
It has been stated that this will not happen in an SE amp since the current in the output transformer is constant. This may be true under normal conditions, but again if the amp is driven hard enough to cut the tube off even for a microsecond, an arc can start. Believe me, I have blown more than one SE OPT (and a tube socket) from open or incorrect loads.
This may be true, but 50+ year old tinkerers that often act like kids can blow up a lot of stuff too. How do I know this?
There things that can be done to help the situation, but nothing can make a tube amp totally foolproof. I have used all of these with success, but each must be tweaked to the particular amp design and power level. For all except the resistor, the trick is to adjust the value of the protection device such that is does NOT conduct when the amp is operated with the highest possible load impedance at full rated power, hard clipping if that can be expected in normal operation (guitar amp, or teenage kids).
1) The previously mentioned big resistor. I use a 100 ohm 5 watt wirewound resistor directly across the speaker terminals on low powered (typically SE) guitar amps. This will cost some power but it works for most cases, and doesn't affect the sound even at full clip with weird loads. If you intentionally mis load the amp (guitar amp) to get a particular sound, a resistor may affect the sound.
2) Wire a "transient protection diode" of the bidirectional type directly across the speaker. The diode must be sized so that it NEVER conducts in normal operation. If it does the sound will be bad (no, gross) and the diode will get hot. In higher powered amps many diodes can be paralleled. The poor diode will have to eat the entire output of the amp if the speaker is removed, but if the diode fails they usually short, saving the OPT. Use the 1.5 KW (peak) rated diodes. I use the 24 volt flavor in my "Turbo Champ", an 8 watt SE guitar amp.
3) CP Claire, Seimens and others used to make gas spark arrestors for use in phone line equipment. These come in a few different voltage ratings. They seem to be able to take the output of a 50 watt guitar amp for several minutes before MELTING! I don't know if they are still common.
4) I have seen people connect zener strings and even VR tubes across the OPT primary. I tried this with a zener string, and it burst into flames instantly when I unhooked the speaker on a 50 watt P-P amp. The VR tube idea may work, but I have not tried it.
5) this is unpopular, but large amounts of negative feedback from the OPT secondary can tame the fire gods.
I will often use multiple methods together to design an almost foolproof amp. In a high power guitar amp I will use a big resistor and 2 or 3 gas discharge tubes across the speaker. I will also wire a bidirectional zener or a neon bulb from the 16 ohm tap back through a voltage divider to provide a large amount of negative feedback only when the bulb (or zener) fires. This is independent of the main feedback loop and applied to the driver or pre driver stage.
Those who know me (or have followed my experiments) understand that I tend to test anything that I intend to sell under extreme (often destructive) conditions. I believe (and it has been confirmed) that a customer (especially for a guitar amp) will find new and unique ways to blow things up that I could never imagine (or test for). All of my HiFi amps must live for a few hours with me playing guitar through them with about 10 db of overdrive. Don't try this with your expensive 300B amp.
As mentioned before an improperly set up amp (or one with bad caps) can slowly cook an OPT or power transformer to death. This will usually fry output tubes first. Sometimes a bad or low quality output tube can arc causing the death of an OPT, but this is not common. Old (vintage) OPT's can absorb moisture and loose insulation quality from years of heat soak, and just fail, even though there is nothing else wrong with the amp.
So treating our tube amps with great care and attention especially to the connection of the speaker is of great importance.
What I am thinking of now is if the load is removed, how long would it take more or less to kill the opt?
10 seconds? 30 seconds? a minute? Does it make it worse if the volume is up when the speaker is disconnected?
What I am thinking of now is if the load is removed, how long would it take more or less to kill the opt?
10 seconds? 30 seconds? a minute? Does it make it worse if the volume is up when the speaker is disconnected?
Does it make it worse if the volume is up when the speaker is disconnected?
Yes. You might (I wouldnt be betting on it) be ok if no signal.
Yes. You might (I wouldnt be betting on it) be ok if no signal.
Thanks Tweeker, this goes to show that we must really check our amps before firing it up. Better sure than sorry is in order. 😀
Most early tube amp designs recommended a 1K resistor across the sec. Most LS in the 1960's were 15 ohms, so it had some effect swallowing around 1/2 W on most amps. It does influence the stability when the sec load is somewhat capacitive.
An amplifier which has been badly designed (many about) which has a normal amount of NFB (15-20dB) can oscillate under o/c conditions or even worse still a defective o/c LS creating an under- damped reactive circuit from the crossover network. I've come across this quite often. The 4040 amp problems in another thread are only the start of serious troubleshooting.
Deteroriation of the winding insulation takes place over time. Excessive temp, VA stresses and coupled with all too tight pile-wound windings all contribute to failure modes. The mean time for failure (MTBF) for a well designed o/p tranny should be around 20yrs. Modern-day insulation materials have improved to skip layer/layer insulation and is often the cause of many internal s/c's.
A overstrain caused by o/c secondary could create flashover on the primary if the insulation is poor or windings pile-wound over taught and or/also flashover in the o/p tubes. Tubes are more robust than most think.
The ideas of zeners across any windings should be scrapped. Any diode acts as a capacitive varactor under high voltage and this can cause serious ringing and destroy series resistors in UL circuits.
Look after these o/p trannies....( ther'e not just wound anyhow...ther'e expensive and in a decade there won't be the experts and skills around to know how to really wind them to a spec how they used to be..
(I'm ignoring the worse ones ..toroids)
richj
An amplifier which has been badly designed (many about) which has a normal amount of NFB (15-20dB) can oscillate under o/c conditions or even worse still a defective o/c LS creating an under- damped reactive circuit from the crossover network. I've come across this quite often. The 4040 amp problems in another thread are only the start of serious troubleshooting.
Deteroriation of the winding insulation takes place over time. Excessive temp, VA stresses and coupled with all too tight pile-wound windings all contribute to failure modes. The mean time for failure (MTBF) for a well designed o/p tranny should be around 20yrs. Modern-day insulation materials have improved to skip layer/layer insulation and is often the cause of many internal s/c's.
A overstrain caused by o/c secondary could create flashover on the primary if the insulation is poor or windings pile-wound over taught and or/also flashover in the o/p tubes. Tubes are more robust than most think.
The ideas of zeners across any windings should be scrapped. Any diode acts as a capacitive varactor under high voltage and this can cause serious ringing and destroy series resistors in UL circuits.
Look after these o/p trannies....( ther'e not just wound anyhow...ther'e expensive and in a decade there won't be the experts and skills around to know how to really wind them to a spec how they used to be..
(I'm ignoring the worse ones ..toroids)
richj
Yes, as mentioned it is usually safe with no signal (if the amp is stable).
If the amp is playing loudly the failure can be instant and catastrophic. I have seen an amp instantly burst into flames when the bass player tripped over the speaker cord (the previously mentioned Ampeg SVT). An autopsy revealed a severely toasted OPT, 2 fried 6146's, and a blown power transformer (open secondary) there were burn marks on several of the tube sockets. This amp runs 6 6146 transmitting tubes in PPP making about 300 watts RMS. Ampeg later switched to 6550's after a few more flaming amp incidents including one at a televised Rolling Stones concert.
At lower power levels it can take a while for the insulation to break down, but something will usually blow once an arc starts.
I had a guitar player complain that his amp would make a loud squeal sometimes when he was cranking it to eleven. I tested the amp on the bench at full power for a day without incident, so I took it back to him and asked him to show me. Yep, it screamed. I repeated the experiment with the amp opened up. When the amp screamed I could see a briliant blue arc jumping from pin 3 (plate) to pin 2 (filament - fortunately grounded) of one of the 6L6GC's. In this case nothing was damaged. The culprit, 16 ohm speakers on the 4 ohm tap (he liked the distortion this made). The OPT, and the cheap Chinese 6L6's survived. I put his 16 ohm speakers on the 8 ohm tap and put heat shrink on the plate leads. The amp is still playing with the same tubes and that was about a year ago.
This was an amp that I had made with an open frame budget OPT. I have personally blown these same OPT's (often involving smoke and flames). In this case the socket was the point that required the least voltage to break down, so that is where the arc started. If it had been a cheap plastic socket (it was ceramic) it would have fried. If pin 2 was not grounded the power transformer could have been fried, or the arc might have started somewhere else.
This amp had operated for two years without incident until the user decided that he only needed one speaker cabinet. His 8 ohm load became 16 and he liked the sound it made with the impedance switch on 4 ohms. I will also guess that his speakers were higher than 16 ohms at the frequency (always the same note in the same song) that started the arc.
In theory this is true, however I have found no ill efects from putting 3 or 4 of the big transorbs across the speaker leads in a GUITAR amp. The impedance is low enough to ignore the 1000+ pF capacitance. I wouldn't do this in a high end DHT HiFi amp (or any HiFi amp) but this is not as much of an issue in HiFi, because the amp is not moved around on a daily basis. I am sure that there is a case somewhere where this may cause instability, and that is why I said that all of these ideas must be tested on a case by case basis.
Momentary shorting of the output in a tube amp is not usually a problem. Operating a tube amp into a short at full power for an extended period of time will cause excessive dissipation in the output tubes leading to a failure.
On a similar note, there used to be an urban myth that playing bass through a Bandmaster will blow it up. I am sure that there are all kinds of other variables involved, but I can say that this can be TRUE. A Bandmaster has a dinky little OPT that does not have enough inductance to handle the lowest notes on a bass guitar. At these frequencies the transformer itself looks like a short causing the 6L6's to glow bright red. The amp was shut off before it fried.
tubelab.com said:
I noticed that as the opt tranny increases in physical size, it also increases it's power handling capabilities but the impedance also decreases. so is it a good idea to get a low impedance opt which can be used for say 4 EL34 in PPP and just use 2 EL34 in PP?
Does big opt size (physical size) = lower bass?
Thanks!
You don't really want to change the impedance... at least not without a well thought out reason. The impedance of the OPT should be matched to the tubes (and loads) used in the circuit.
Consider a transformer like a gearbox in a car or sprockets on a bike. You want a certain ratio so the tubes can do their best.
Now... all other things being equal (impedance for one); a smaller transformer means less inductance and does mean less bass. So... a larger transformer can mean more bass. This is a very general statement; with lots of exceptions.
🙂
Consider a transformer like a gearbox in a car or sprockets on a bike. You want a certain ratio so the tubes can do their best.
Now... all other things being equal (impedance for one); a smaller transformer means less inductance and does mean less bass. So... a larger transformer can mean more bass. This is a very general statement; with lots of exceptions.
🙂
- Status
- Not open for further replies.