LR8: an LM317 on steroids to stabilize up to 450 V

[roberto]

Here it is, cheap and simple!
https://www.tme.eu/Document/49ed538122e ... 61/lr8.pdf

[bal704]

Looks interesting. Hopefully somebody who knows more than me can post a circuit using this in an amp application.

[Dr Tony Balls]

From the datasheet:

The output current limit is typically 20mA

There's the rub. Even a pretty simple 18W Marshall type is gonna draw 3X that. That said, there is a way to use one of these in conjunction with a power mosfet circuit to create a regulator that will work well with B+ voltages and currents. This link is provisional but explains a way in which it might be done: https://wtfamps.com/2018/10/17/the-lr8- ... -circuits/. Its a wormhole that I dont have time for right now but searching "LM8 tube amps" or "B+ regulator" produces a lot of information to digest.

[R.G.]

There are a few high voltage regulators on the market now, similar to the LR8. I ran into them while researching for the agile high voltage output for the tube matcher thing in another thread. They all seem to have some issues for a B+ regulator.

For a stabilizer, it's hard to beat an N-channel MOSFET and a zener string in simplicity. It's not a very good regulator, but it makes for high currents and relatively stable output. The output voltage is close enough for tube amps. Circuit image here:
https://www.circuitlab.com/circuit/qcv8 ... regulator/

For a regulator, it's hard to beat an N-channel MOSFET a bipolar transistor, a zener, and a handful or Rs and Cs. Circuit image here:
https://www.electronicsandcommunication ... esign.html
Notice that the last is for a bipolar pass transistor, as I could not quickly find the MOSFET version. There are a couple of issues with this for changing to MOSFET and high voltage. First, you need the standard 100 ohm series gate resistor to avoid self oscillation at many megahertz, and a gate-source 12V zener to protect the gate. Second, the remaining bipolar needs to withstand the entire bulk supply voltage. I would use a 2N5401 for the bottom transistor, but add a string of voltage-sharing resistors and 5401s to divide down the bulk voltage to a manageable 100V or so per transistor. Same concepts, just some add-ons.

These are not as sexy as having a one-IC B+ reguator, but they have the advantage that they're cheap and they work, as well as being highly available.

[Roe]

From the datasheet:

The output current limit is typically 20mA

There's the rub. Even a pretty simple 18W Marshall type is gonna draw 3X that. That said, there is a way to use one of these in conjunction with a power mosfet circuit to create a regulator that will work well with B+ voltages and currents. This link is provisional but explains a way in which it might be done: https://wtfamps.com/2018/10/17/the-lr8- ... -circuits/. Its a wormhole that I dont have time for right now but searching "LM8 tube amps" or "B+ regulator" produces a lot of information to digest.

Could work in preamps

[Triple Jim]

From the datasheet:

The output current limit is typically 20mA

There's the rub. Even a pretty simple 18W Marshall type is gonna draw 3X that.

Worse yet, that 20 mA is only good when Vin - Vout- = 12 volts. It tapers down to 0.5 mA when Vin - Vout = 450v.

[roberto]

I don't think that stabilizing screens and plates of the output tubes is a good idea.
On top of that, there's always the old trick to increase the current of voltage regulators.

[mhuss]

For a stabilizer, it's hard to beat an N-channel MOSFET and a zener string in simplicity.

The only problem with these is when they fail, they fail catastrophically, and you usually have to replace everything (he says, having unsoldered way too many small Zeners from crappy PCBs).

[R.G.]

True. However, there is much to be improved about the circuit. "Hard to beat" in my mind includes at least a partial consideration of the audience. This is a very mixed audience I can and have designed variations of that circuit that are much less prone to failure. But while there are some here who like a good delve into a complex circuit, there are others whose eyes glaze over after three non-resistor or cap parts.

More importantly, it's not a regulator in that it has no active correction for changes. It's a pretty good stabilizer.

I would add a MOSFET current limiter to it, so load faults can't make it evaporate, and probably worry more about the zener string. Actually, zeners are not needed, just a fixed voltage source. Zeners are just handiest. I've found some circuits that use the TL431 precision adjustable reference as a reference and error amplifier for a big MOSFET perched several hundred volts in the air. I'm not sure I trust that yet, though.

[mhuss]

The MOSFET supply you came up with for me a while back works a treat, but would be a bit of a bear to wire on a turret board.

[Kagliostro]

It works

K

[Synchu]

You can see a possible LR8 (preamp) application here.
https://diyaudioprojects.com/Tubes/EL84 ... lock-Amps/

I built two of these for my brother quite a bit ago and they are running just fine and sound fabulous.

From what I remember - it was PITA sourcing them in EU

Niki

[Kagliostro]

Not difficult to be find at 50c (61c with tax)

https://it.rs-online.com/web/p/regolato ... i/8293266/

Told that I must specify that I think it will be better used in an HiFi circuit

in a guitar amp a solid B+ on preamp tubes may be not what you would like to have

but this is my opinion

K

[roberto]

I really depends. I have one preamp where I have stabilized B+ and I like it because it gives you stable tone whatever voltage you have.
Drawback is that you don't have interaction on the power supply with the power amp, but no rack unit preamp has it.

Kagliostro highlighted the very advantage of it: it's so cheap that you can even use one for every (pair) of stage(s) and set the voltage you prefer for each one.

In a complete amp I would use it for the first stage, where S/N is more important.

Then Hi-Fi opens way more applications, of course.