Hi,
I am curious about the use of protective diodes in the output stage of valve amplifiers. I have seen flyback diodes in use with relay coils but the topology was different. In particular, I am looking at this schematic, which I don't really understand:
http://www.blueguitar.org/new/schem/fender/blues_deville.gif
Here is my analysis so far, assuming no load is connected to the secondary:
If the grid voltage is increasing, V4 is becoming more conductive. The voltage across the transformer primary is more positive than normal due to the changing current. Perhaps this is more voltage (but less current) than the transformer primary usually sees but I don't see what issue would occur as I assume the power dissipated by the transformer is still lower than normal operation, nor do I see the diode CR24 doing anything in this case. It seems to me that the rate of change of current is limited by inductance causing the plate voltage approaching 0, at which point current would no longer flow through the valve (except grid leakage?). So presumably the plate voltage cannot go (much) below 0, even without the diode, even if the grid voltage changes rapidly.
If the grid voltage is decreasing, V4 is becoming less conductive. The voltage across the transformer primary is less positive than normal, possibly resulting in a plate voltage which is above B+. I also don't see what the diode does in this case.
Based on my understanding, if I wanted to protect against an open circuit, I would instead think about connecting the diode cathode to B+ and the anode to CP1, much like what I've seen with relays, ensuring that the plate voltage never increases above B+. It seems like there is something I am not understanding about the open circuit failure modes, or that CR24 has nothing to do with them. Or is the plate going slightly below ground a huge issue? Is the plate voltage being above B+ not an issue at all?
I am curious about the use of protective diodes in the output stage of valve amplifiers. I have seen flyback diodes in use with relay coils but the topology was different. In particular, I am looking at this schematic, which I don't really understand:
http://www.blueguitar.org/new/schem/fender/blues_deville.gif
Here is my analysis so far, assuming no load is connected to the secondary:
If the grid voltage is increasing, V4 is becoming more conductive. The voltage across the transformer primary is more positive than normal due to the changing current. Perhaps this is more voltage (but less current) than the transformer primary usually sees but I don't see what issue would occur as I assume the power dissipated by the transformer is still lower than normal operation, nor do I see the diode CR24 doing anything in this case. It seems to me that the rate of change of current is limited by inductance causing the plate voltage approaching 0, at which point current would no longer flow through the valve (except grid leakage?). So presumably the plate voltage cannot go (much) below 0, even without the diode, even if the grid voltage changes rapidly.
If the grid voltage is decreasing, V4 is becoming less conductive. The voltage across the transformer primary is less positive than normal, possibly resulting in a plate voltage which is above B+. I also don't see what the diode does in this case.
Based on my understanding, if I wanted to protect against an open circuit, I would instead think about connecting the diode cathode to B+ and the anode to CP1, much like what I've seen with relays, ensuring that the plate voltage never increases above B+. It seems like there is something I am not understanding about the open circuit failure modes, or that CR24 has nothing to do with them. Or is the plate going slightly below ground a huge issue? Is the plate voltage being above B+ not an issue at all?
If you look at the polarity of the two diodes, they stop the transformer producing negative voltages normally caused by running a valve amplifier at high volume. They are rated at 3000Volts with a 4Volt forward drop and it is possible to almost achieve that voltage when spikes occur.
So, the anodes will never go negative, running the risk of flashover or damage to both the valves and the transformer.
When one half of the push pull output stage pulls down on the anode, the anode of the other rises above HT. If the amplifier pulls hard at a parasitic frequency without a load, the back EMF is catastrophic as the valve pulling can set up a resonation that will keep going past the valve conducting and try to pull the voltage negative. Stripping the grids because the anodes will be hotter than the grids!
That is a way of protecting the output stage and the transformer.
So, the anodes will never go negative, running the risk of flashover or damage to both the valves and the transformer.
When one half of the push pull output stage pulls down on the anode, the anode of the other rises above HT. If the amplifier pulls hard at a parasitic frequency without a load, the back EMF is catastrophic as the valve pulling can set up a resonation that will keep going past the valve conducting and try to pull the voltage negative. Stripping the grids because the anodes will be hotter than the grids!
That is a way of protecting the output stage and the transformer.
Plate voltage going below ground is not an issue for the output valve (unless it goes a long way down) but as it is a PP it is likely that the other plate is going up. Basically the diode protects against running without a load so the magnetic energy stored in the transformer emerges as a spark discharge - most likely within the transformer so damaging the insulation. Under these conditions the voltage could easily go below ground. The diode then provides a safe way to dissipate the energy by heating all the copper a little instead of heating one spot on the insulation a lot.
When the external circuit (whatever it is) tries to stop the current flowing though an inductance the result is a voltage spike as the magnetic energy in the inductance is turned into potential energy in whatever capacitance is present. This is because the inductance really wants to keep the current flowing. The diode allows current to continue to flow so instead of the energy being dissipated in a spark it just heats up the winding a little.
A reverse diode to ground is fine, because under normal conditions it is never switched on. A diode to B+ would not work because under normal conditions the plate voltage could approach twice B+ anyway.
When the external circuit (whatever it is) tries to stop the current flowing though an inductance the result is a voltage spike as the magnetic energy in the inductance is turned into potential energy in whatever capacitance is present. This is because the inductance really wants to keep the current flowing. The diode allows current to continue to flow so instead of the energy being dissipated in a spark it just heats up the winding a little.
A reverse diode to ground is fine, because under normal conditions it is never switched on. A diode to B+ would not work because under normal conditions the plate voltage could approach twice B+ anyway.
The output transformer works as a voltage see saw:
1) start with both V4 and V5 plates plus transformer center tap all at +470V , which is +V
2) now V5 is driven and starts passing more and more current: its plate voltage will lower, until it eventually saturates, reaching some +60V , so it dropped 410V.
At the same time, V4 plate rises by the exact same amount, remember the primary is an autotransformer on its own, center tap is at fixed voltage, when one end goes down the other goes up (that´s why it´s calld a see saw effect 🙂 ) so it will reach 470+410V=890V .... scary, huh?
3) that´s nothing: typical speaker is quite inductive, so driving it will create inductive peaks, besides that whe tubes saturate voltage swing stops abruptly .... transformer itself is a huge inductor, ugly voltge peaks appear.
On an overdriven guitar amp you can easily surpass 1kV peaks, and that with speaker connected; let alone if Musician forgets to connect it (or cable is bad or speaker just died with an open VC) and "tests" his amp playing a few power chords at high volume, wondering about the silence.
Fact is that for V4 plate to reach above +940V (2 x 470V) V5 must go below 0V , reaching negative voltage respect to ground.
But now the protective diode (if fitted) turns on and clamps that voltage.
Yes, energy will be dissipated in the winding as heat ... no big deal .... transformer will survive that easily while >1200 or even 1500V peaks which are entirely possible with load disconnected are safely stopped before punching through wire enamel insulation and killing the OT.
Mind you, I have tested both with/without them in guitar amps, with a switch to quickly go back and forth, ear memory is very short, and there is a small difference, "without" sound has a little more "bite" , with them is slightly smoother, easy to confirm by looking at waveforms, where the squarewave front end peak is reduced, but personally prefer to use them and feel safer.
1) start with both V4 and V5 plates plus transformer center tap all at +470V , which is +V
2) now V5 is driven and starts passing more and more current: its plate voltage will lower, until it eventually saturates, reaching some +60V , so it dropped 410V.
At the same time, V4 plate rises by the exact same amount, remember the primary is an autotransformer on its own, center tap is at fixed voltage, when one end goes down the other goes up (that´s why it´s calld a see saw effect 🙂 ) so it will reach 470+410V=890V .... scary, huh?
3) that´s nothing: typical speaker is quite inductive, so driving it will create inductive peaks, besides that whe tubes saturate voltage swing stops abruptly .... transformer itself is a huge inductor, ugly voltge peaks appear.
On an overdriven guitar amp you can easily surpass 1kV peaks, and that with speaker connected; let alone if Musician forgets to connect it (or cable is bad or speaker just died with an open VC) and "tests" his amp playing a few power chords at high volume, wondering about the silence.
Fact is that for V4 plate to reach above +940V (2 x 470V) V5 must go below 0V , reaching negative voltage respect to ground.
But now the protective diode (if fitted) turns on and clamps that voltage.
Yes, energy will be dissipated in the winding as heat ... no big deal .... transformer will survive that easily while >1200 or even 1500V peaks which are entirely possible with load disconnected are safely stopped before punching through wire enamel insulation and killing the OT.
Mind you, I have tested both with/without them in guitar amps, with a switch to quickly go back and forth, ear memory is very short, and there is a small difference, "without" sound has a little more "bite" , with them is slightly smoother, easy to confirm by looking at waveforms, where the squarewave front end peak is reduced, but personally prefer to use them and feel safer.
I don't know if I am the only one but I have encountered issues with these diodes in practice. To me it appears the reverse breakdown voltage gets lower or something and they start to conduct on higher volumes. It's usually one that seems to do it, not both. You end up getting asymmetric clipping at higher outputs, remove the diodes and the output comes back to normal. I have encountered this roughly 3 times now in Fender amps.
Since the ends of the primary are at opposite polarity, signal-wise, only one diode will ever conduct at any time. Higher volumes is where the primary voltage swings are greatest.
Correct, that makes sense. What would lower their reverse breakdown voltage? Heat? Age? I assume Fender doesn't want them to conduct except for failure mode, I think the diodes are rated for 1kV.
How many do you use in series for each anode? I use 3 per anode and haven't seen any differences or ill effects.
I looked at the manual it says they are 3kV.
http://www.tangible-technology.com/schematics/fender/hrd/Hot Rod Deluxe.pdf
http://www.tangible-technology.com/schematics/fender/hrd/Hot Rod Deluxe.pdf
There have been a few threads here addressing output transformer protection methods.
Tim posted this to one of them - worth a read:
http://dalmura.com.au/projects/Output transformer protection.pdf
Cheers,
Ian
Tim posted this to one of them - worth a read:
http://dalmura.com.au/projects/Output transformer protection.pdf
Cheers,
Ian
- Status
- Not open for further replies.