Regulated HV supply
Moderators: pompeiisneaks, Colossal
Regulated HV supply
Any of you electron wranglers out there use an SiC or Mosfet as a B+ regulator?
Tube junkie that aspires to become a tri-state bidirectional buss driver.
Re: Regulated HV supply
also IGBT.. even better
Re: Regulated HV supply
I've used MOSFETs for regulating high (600-1000V) voltages before, although not in tube amps. A lot depends on what the source voltage is, what the output voltage is, and the output currents, both max normal and transient.
Perhaps even more important is what the physical and economic ($) constraints are.
I would disqualify bipolars from the start. MOSFETs are too easy to drive and too cheap per unit of goodness to use bipolars unless you have some very specific reasons for going there.
IGBTs are bipolars that have MOSFET-like gate structures to drive them, eliminating some of the base drive hassles of bipolars. Unless you have some specific reasons for using IGBTs, discount these.
Silicon MOSFETs are very good, very available, very cheap per unit of goodness unless you (again) have some very specific reasons to not use them.
The minimum and maximum input to output voltage matters. If this is quite small, a few volts, it gets difficult to drive the pass device's gate/base well. With something like 5-10V or more between minimum input - output voltage, MOSFET gate drive gets relatively easy. If input - output voltage is big (50 to hundreds of volts) and current is high, then it becomes very difficult to get the waste heat out of your pass device with a linear regulator. As an example, if you are regulating 600V down to 500V at average currents of 300ma, then the power the pass device will have to dissipate is (600-500)*0.3 = 33W. That is modestly difficult for a single device in natural convection heat sinking, so you may be up for BIG heat sinks or multiple parallel pass devices.
If you have, for instance, 600V with 80V of ripple, then your regulated 500V will have as little as 20V of headroom, and your pass device will have to be able to conduct 300ma with only 20V of drop. If your "600V" sags by 20% AND has 80V of ripple, you can't regulate to 500V because 600V minus 20% is 480V, so there is no regulation headroom left.
The general circumstances of input voltage, ripple, tolerance, output voltage and output current are usually much, much more important that what you use for a pass device.
Perhaps even more important is what the physical and economic ($) constraints are.
I would disqualify bipolars from the start. MOSFETs are too easy to drive and too cheap per unit of goodness to use bipolars unless you have some very specific reasons for going there.
IGBTs are bipolars that have MOSFET-like gate structures to drive them, eliminating some of the base drive hassles of bipolars. Unless you have some specific reasons for using IGBTs, discount these.
Silicon MOSFETs are very good, very available, very cheap per unit of goodness unless you (again) have some very specific reasons to not use them.
The minimum and maximum input to output voltage matters. If this is quite small, a few volts, it gets difficult to drive the pass device's gate/base well. With something like 5-10V or more between minimum input - output voltage, MOSFET gate drive gets relatively easy. If input - output voltage is big (50 to hundreds of volts) and current is high, then it becomes very difficult to get the waste heat out of your pass device with a linear regulator. As an example, if you are regulating 600V down to 500V at average currents of 300ma, then the power the pass device will have to dissipate is (600-500)*0.3 = 33W. That is modestly difficult for a single device in natural convection heat sinking, so you may be up for BIG heat sinks or multiple parallel pass devices.
If you have, for instance, 600V with 80V of ripple, then your regulated 500V will have as little as 20V of headroom, and your pass device will have to be able to conduct 300ma with only 20V of drop. If your "600V" sags by 20% AND has 80V of ripple, you can't regulate to 500V because 600V minus 20% is 480V, so there is no regulation headroom left.
The general circumstances of input voltage, ripple, tolerance, output voltage and output current are usually much, much more important that what you use for a pass device.
Re: Regulated HV supply
Thanks R.G. always good to hear from an expert. I've followed your sandy state wizardry for decades? Man I'm getting old.
Glad to see you have valve interests also.
Glad to see you have valve interests also.
Tube junkie that aspires to become a tri-state bidirectional buss driver.