Simulation says it should work. You should be able to generate a positive and a negative voltage with two Maida Regulators in series if the conditions below are met.
Conditions: The two Maida Regulators must be fed from separate transformer windings, separate rectifiers, and separate reservoir caps. Basically, the input circuits must be completely separate. You then put the two regulators in series on their outputs and ground the centre point between the two.
Tom
Conditions: The two Maida Regulators must be fed from separate transformer windings, separate rectifiers, and separate reservoir caps. Basically, the input circuits must be completely separate. You then put the two regulators in series on their outputs and ground the centre point between the two.
Tom
Yep. The two regulators are completely independent. The only thing they share is the ground connection. So you can rig one for the B+ and another for a negative bias voltage with minimal current requirements.
I'd certainly welcome that. Thanks.
Tom
I recently bought a third Maida Regulator. The first ones work like a charm so
Awesome. Thank you for your support. I'm glad you liked them.
Yep. 550-560 V - or even 550-600 V - will be fine.
I'd limit the input voltage to 800 V. Just be really, really careful that any variation in the mains voltage doesn't cause the input voltage to exceed this value.
Tom
Rev. 2.0 is much easier to solder, optimized layout, and offers about 20 dB better line rejection and lower output impedance.
I hope you still have some in stock
I have 40+ boards in stock right now. If you buy all 40+, I'll gladly negotiate a bulk discount...
Best way to buy is through my website: www.neurochrome.com
Not at all. Just let the Maida Regulator do its job. You'll have about 20 µV of residual hum and noise on the output. It's a state of the art regulator.
Tom
Hi Tom, i am new to using SS devices on tube equipment and am very seriously considering installing your MReg to see what kind of difference it will make in my already completed RH amp. If i do install one where would it be located in the circuit? My power supply is usually CLCLC or LCLC type. thank you for replying and have a happy holidays.
A well-regulated B+ will result in a reduction in hum and mains hum related IMD products. The much lower output impedance of the regulator is also likely to result in a much improved (tighter) bass.
The 21st Century Maida Regulator requires a DC voltage on the input. The voltage must be 15 V above the output voltage at all times (including during the valleys in the ripple voltage).
There is no need for filtering beyond a smoothing capacitor on the input of the regulator. In my DG300B, for example, I'm running a 50 µF smoothing cap and have about 50 Vpp of ripple on the input of the 21st Century Maida Regulator. I measure 20 µV RMS of hum+noise on the output of the regulator (~1 MHz measurement bandwidth).
Tom
The 21st Century Maida Regulator requires a DC voltage on the input. The voltage must be 15 V above the output voltage at all times (including during the valleys in the ripple voltage).
There is no need for filtering beyond a smoothing capacitor on the input of the regulator. In my DG300B, for example, I'm running a 50 µF smoothing cap and have about 50 Vpp of ripple on the input of the 21st Century Maida Regulator. I measure 20 µV RMS of hum+noise on the output of the regulator (~1 MHz measurement bandwidth).
Tom
Hi and thank you for the quick response. So for the power supply I would just need the initial B+ to be cap filtered then straight into the MR. Do you hear any difference in sound if the rectifier is tubed or SS? If I need to power the grid on a pentode and also 2 or 3 drivers and phase splitters, would the MR be able to handle the additional current demands? cheers, 808
Exactly.
I highly doubt it. The difference in sound between a tube rectifier and diode rectifier is likely due to the nonlinearities in the voltage drop of the tube rectifier. I.e. its output impedance is a nonlinear function of the rectifier current. A diode is too, but the diode drop may vary a few 10s of mV where the tube rectifier's drop will vary by 10s of V in some cases.
With the Maida Regulator in there, the output impedance will be low across the audio band, hence, the output voltage won't budge as more current is drawn.
In theory, the 21st Century Maida Regulator can supply 1.1 A if you can get the regulator to survive startup and handle the power dissipated in the cascode device. In practice, I'd stick to 400-500 mA as a reasonable limit.
Thanks for answering the questions. Do you also have something for -120V @ 0.1A?
You could configure a separate transformer winding for a rectifier + 21st Century Maida Regulator. Rig the regulator for 120 V and connect Vout to GND. Then the "GND" of the Maida Regulator will be at -120 V. This does require that the transformer winding is separate and not just a tap off of a multi-tap winding used for other circuits.
Tom
This is an epic, much kudos for first posting and then the effort taken to join in in in the previous forty odd pages of comments. For those like me not especially interested in regulator design it's quite instructional. Clearly this is more sophisticated than RCRCRC...
So first of all thanks!
That said, and I'm sorry for bringing this up again as it's been covered. But, am I correct in thinking its the Mosfet that gets the big hit at start-up? If so I'm liking part STW22N95K5 as a potential alternate (17.5A), particularly as the STW12NK95Z is NRND? Playing 'top-trumps' with the datasheets suggests it'd be *more* robust but I'm not sure what I'm looking at! More robust seems like a good idea even if they are a few more £.
Noted though that the correct part is in stock at Digikey. But I've become interested enough to ask. My amp looks like 120ma at 420v so the original is absolutely tailor made.
Thanks Again
Andy
So first of all thanks!
That said, and I'm sorry for bringing this up again as it's been covered. But, am I correct in thinking its the Mosfet that gets the big hit at start-up? If so I'm liking part STW22N95K5 as a potential alternate (17.5A), particularly as the STW12NK95Z is NRND? Playing 'top-trumps' with the datasheets suggests it'd be *more* robust but I'm not sure what I'm looking at! More robust seems like a good idea
even if they are a few more £.Noted though that the correct part is in stock at Digikey. But I've become interested enough to ask. My amp looks like 120ma at 420v so the original is absolutely tailor made.
Thanks Again
Andy
You're welcome.
Correct.
Maybe.
What you're looking for is the MOSFETs ability to handle short-term stress. You'll find that information in the Safe Operating Area (SOA) curve in the data sheet. I usually estimate the start-up current from R3, Vgs of Q1, and the drop-out voltage of U1. I then look at the SOA curve to figure out how long Q1 can survive the full input current at the calculated start-up current. I prefer Q1 to be able to survive for longer - preferably much longer - than 10 ms (normally the longest time specified).
To get a better estimate of the power dissipated in Q1 at start-up, I suggest running a simulation.
I also look at the gate capacitances of the MOS device. If they're comparable to those of the STW12NK95Z, the substitute device will likely work.
A few candidates for replacements of the STW12NK95Z have been mentioned earlier in this thread (only a few pages back). Right now Digikey has 55 in stock (down from about 400 more than a year ago), so they're still possible to get and don't appear to be flying off the shelf. Until Digikey runs out of stock, I'd stick with the STW12NK95Z.
Tom
Dear Folks,
I wonder where to find the last schematic state of this wonderfull 21st Maida regulator?
This therad grows and grows and it seemed not to be really necessary to read more than 400 Post to extract the last schematic state
I just want to use the regulator for 70mA@245V. Input voltage will be arround 330V.
Thanks a lot and best regards
Karsten
I wonder where to find the last schematic state of this wonderfull 21st Maida regulator?
This therad grows and grows and it seemed not to be really necessary to read more than 400 Post to extract the last schematic state
I just want to use the regulator for 70mA@245V. Input voltage will be arround 330V.
Thanks a lot and best regards
Karsten
Tom,
On the 21st Maida reg. PCB capacitor C1b has a value 10uf/450 for voltages lower than 400V.
For higher voltages you populate position, C1a, with a lower value polyprop cap instead. C1b is intended for an electrolytic cap due to size limitations.
Will using the lower value polyprop intended for higher voltages in a lower voltage set-up (200V) give inferior results? My reason for asking is that I am planning on an all polyprop PS.
On the 21st Maida reg. PCB capacitor C1b has a value 10uf/450 for voltages lower than 400V.
For higher voltages you populate position, C1a, with a lower value polyprop cap instead. C1b is intended for an electrolytic cap due to size limitations.
Will using the lower value polyprop intended for higher voltages in a lower voltage set-up (200V) give inferior results? My reason for asking is that I am planning on an all polyprop PS.
Attachments
Last edited:
Oh, it's in the thread somewhere.
It is also part of the design documentation that you will get when you order the board.This thread actually dates back to the time when I gave my circuits away. I've since then decided to take a more business-like approach. Neurochrome is my only source of income while I'm back at the university tooling up for my second career.
Sounds like a perfect job for the 21st Century Maida Regulator.
More capacitance is better, hence, I specify an electrolytic cap for C1 for voltages where electrolytic caps are available. If you'd rather have a polypropylene cap, just go with the one I specify for operation above 400 V.
Tom