Yeah... I guess that's what I get for making my designs available freely to the DIY community. Duly noted... Thankfully, most people are honest, but sadly some people are not.
Getting the LT3080 to survive in a high-voltage environment is significant intellectual property. That part of my design was flat out copied.
As noted on the schematic, you're free to do that for NON-PROFIT USE ONLY.
Please... There are three resistors in that design that need to be power resistors and only a few components that actually see high voltage across them. I don't see anything in your layout worth writing home about. In case of the tranzorb (D3), one of the diodes (D1), the trimpot, and the Solen output cap, the components on the board don't even fit the footprints.
Tom
Hello Tom and thank you for all the effort put behind this regulator!
I have a working board bought from another member set up at 400V out with 450V in, recently I needed for a test a 50-150V supply at 1-2mA max. I tried to mod the circuit but it is not going any lower than 160V even with R9=47K and R5=1k. Am I doing something wrong or it is not doable?
Thank you!
I have a working board bought from another member set up at 400V out with 450V in, recently I needed for a test a 50-150V supply at 1-2mA max. I tried to mod the circuit but it is not going any lower than 160V even with R9=47K and R5=1k. Am I doing something wrong or it is not doable?
Thank you!
You're welcome.
Is this MaidaReg Rev. 1.0 or Rev. 2.0? It should say on the board which revision it is.
I'm guessing you mean R3 = 1 kOhm and R9 = 47 kOhm. That should give you about 100 V out, assuming you provide at least 115 V in.
You may be browning out the LT3080, though. The input voltage to the regulator IC (accessible on R3) should be at least 1.6 V above the output voltage. If this is not the case, decrease R3 to 470 ohm.
I have not used the 21st Century Maida Regulator at those low voltages, but there is nothing technically that prevents it from working. Simulation says you should get 93 V out with the pot centered with R3 = 1 kOhm (or 470 ohm) and R9 = 47 kOhm.
Tom
Would something silly like using a regulator like this on a 211 @ 1000/1100V be possible?
Provided all components are replaced where necessary for ones that can handle the voltage(if they exist). Maybe through hole for some extra spacing. At first glance not a lot low rson / high PD 1500/1700V n-channel fets, it might be a bridge to far.
The advantages are interesting for every amplifier stage, not just the input stage psu's and their "sensible" voltages I need regulators for.
(lots of sand in this amplifier, I really like your filament regulators to!)
Provided all components are replaced where necessary for ones that can handle the voltage(if they exist). Maybe through hole for some extra spacing. At first glance not a lot low rson / high PD 1500/1700V n-channel fets, it might be a bridge to far.
The advantages are interesting for every amplifier stage, not just the input stage psu's and their "sensible" voltages I need regulators for.
(lots of sand in this amplifier, I really like your filament regulators to!)
211 and other similar tubes require a separate dedicated anode voltage transformer. It makes no sense to regulate this supply, you would just burn off tremendous heat in the regulator.
The driver circuit runs from its own lower voltage supply, and may well be regulated.
I don't want to burn a lot of power, 1150 in 1050 out @ 70mA, something like that.
The same way you use it in a 300b amp, running a 211 @ 600V makes no sense
I need regulated supply's in the driver because it will be a dc coupled design, but the output stage will benefit the most from the Ultra-low output impedance swinging between 140mA and 10mA at full power
The same way you use it in a 300b amp, running a 211 @ 600V makes no sense
I need regulated supply's in the driver because it will be a dc coupled design, but the output stage will benefit the most from the Ultra-low output impedance swinging between 140mA and 10mA at full power
I'm already stacking supply's to make the amp dc coupled, the 211 ground will be at +320V compared to the input stage ground. And a 100V supply will stack on the 320V one providing the power for the input stage.
The 320V and 100V need to be regulated to provide a stable bias for the 211.
What's one more stack gona hurt
, but I'd prefer a single regulator.that design seems a lot more complicated and would be operating at it's limits it seems, but could be made to work.
But it would be easier to get the 21st Century Maida Reg. working at 1100V.
R9 560k 10W,
some high voltage caps(or normal ones in series)
upgrade D1, D5 and D6
STW9N150 1500V 8A 320W nfet or STP12N120K5 1200V and keep Vin under 1100V.
Last edited:
The biggest challenge with the Maida Regulator (21st Century or Mike's original circuit) is to make it survive start-up and the high di/dt pulses you get if a socket is loose and the B+ disconnects briefly. It's very easy to hit the SOA limit of the cascode device during start-up. This results in spectacular fireworks... Ask me how I know...
I wonder if a tube-Maida would be the answer. Use a tube for the cascode. They can handle both the voltage and the power and tend to be rather indestructible compared to silicon devices.
Tom
I wonder if a tube-Maida would be the answer. Use a tube for the cascode. They can handle both the voltage and the power and tend to be rather indestructible compared to silicon devices.
Tom
A tube cascode would be interesting, for exactly the reason Rod Coleman mentioned in this topic. I can lose the Ey500 rectifiers and still have a slow rise, but I also wanted to use ey500's/GZ34 rectifier tubes for the driver stage.
So all regulators with tube cascode(even when not necessary) and sand rectifiers would even be more energy efficient and still keep the benefits
So all regulators with tube cascode(even when not necessary) and sand rectifiers would even be more energy efficient and still keep the benefits
The issue is not the power during steady state operation. The issue is at start-up. 1100 V at whatever current the load draws on start-up means trouble. If the load has any capacitance, it'll behave like a short circuit at first. This kills cascode devices.
I suggest running a simulation to play with this.
(I'm in over my head
Then 1100 V circuits is probably not what you want to be working on... You only get one chance at life. Just saying...
Tom