diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Tubes / Valves (http://www.diyaudio.com/forums/tubes-valves/)
-   -   Maida regulator (http://www.diyaudio.com/forums/tubes-valves/108596-maida-regulator.html)

Andrewc556 16th September 2007 05:40 PM

Maida regulator
 
I've just finished a power supply design using a Maida regulator and I'm having some difficulties. It does seem to be working properly except that there's still excessive ripple on the output. When I try to add extra capacitance before it or after it doesn't seem to change anything. Without proper equipment I'm at a loss to measure frequency etc.. of the ripple voltage, but I'm getting about 200mv of ripple from a 200V @ 80ma supply. Being new at these things I was wondering if there's any common mistakes or problems I could be making? I understand that I haven't been too specific about the power supply but any initial thoughts would be appreciated. Thanks.

SY 16th September 2007 06:20 PM

Which variation did you build? Can you sketch a schematic of what you did? I've built dozen of Maida regs and have seen most of what can go wrong...

EC8010 16th September 2007 10:00 PM

How do you know you have 200mV of ripple? What did you use to measure it? Photograph of your regulator?

Colt45 17th September 2007 06:15 AM

are you dropping enough volts through the reg? if it's not enough for the dropout, you'll still have ripple...

Andrewc556 17th September 2007 03:33 PM

2 Attachment(s)
I've attached a JPEG of the circuit. I pretty much ripped the circuit right out of Morgan Jones's "Valve Amplifiers", though changed a few comonent values. I measured the ripple with a pretty run of the mill handheld multimeter. To be honest, not dropping enough voltage on the regulator sounds like something I could have over looked. I ran it through a spice simulation and when the voltage is under the drop out of the regulator it gives a ripple waveform close in voltage to what I'm measuring. Unless any other glaring errors are apparent in the circuit, that's probably what's going on. Unfortunatley I won't even be close to the board until the weekend so I won't know until then. Thanks for the help!

SY 17th September 2007 06:06 PM

Did you have the regulator loaded down with 5-10mA minimum? Without that, it won't regulate properly.

HiFiNutNut 27th September 2007 01:33 AM

Andrew,

Any luck?


SY,

I am about to build one for a 300VDC B+ for my common cathod 12B4 preamp. I have got Morgan Jones's book.

Quote:

I've built dozen of Maida regs and have seen most of what can go wrong
So what can go wrong? Would you please recommend the most reliable, simple version / variation of the Maida reg with its schematic?

I would like to keep things as simple as possible, so if I don't have to build the relay part of the circuit I would avoid it. What worries me at this stage is how the Maida reg would behave when the amp is turned on while the tube is still cold if no relay is used. I presume when turning off there would be no problem.

Your advice will be much appreciated.

Regards,
Bill

SY 27th September 2007 01:54 AM

The number one thing that goes wrong is operator error- a slipped scope probe has killed quite a few pass devices in my lab.

Number two is insufficient minimum current for regulation. 5mA is typical, but some regs seem to need a bit more. That's what you have to pay attention to for startup- a divider string and bleeder combo that takes you over 6-7mA will insure that the reg will start cleanly even with cold tubes.

Number three is insufficient voltage across the regulator (too much is fatal, too little ruins performance). The regs that MJ used in his projects are exemplary, but the one he uses as the basic example has a big problem in this area- the Zener needs to be more like 12V for that circuit.

Number four is output bypassing. For stability's sake, I tend to use a small resistance in series with the output cap. Without it, I've had some oscillation with certain capacitance values.

Number five is insufficient attention to power dissipation ratings, especially the set resistor and the heatsinking of the pass transistor. There are some cool tricks (like a resistor bypassing the reg) to minimize the pass device dissipation- don't hesitate to use them.

Number six is layout. The resistor at the top of the divider needs to be VERY close to the 317.

Number seven is the adjustment potentiometer- running DC through a wiper is not conducive to reliability, so it's best to try to keep that current low. In the Maida reg from the Red Light District, I used the trimmer in series with a large resistor as a parallel trim to the set resistor- the set resistor used about 80% of the set current, reducing the stress on the trimmer's wiper.

HiFiNutNut 27th September 2007 02:56 AM

Wow! Thanks, SY.


Quote:

Number four is output bypassing. For stability's sake, I tend to use a small resistance in series with the output cap. Without it, I've had some oscillation with certain capacitance values.

In MJ's book, 470n is used as the output cap and there is a 2R7 in series.

Is your "small resistance" larger or smaller than 2R7? What do you think to be the optimal value for the output cap and the series resistor?

Regards,
Bill

SY 27th September 2007 03:12 AM

The optimal value will be different, depending on the use, but MJ's values are not unreasonable. I like to swamp the output for best stability, so I use caps in the 20-50uF range, with series resistors of 3-10R. I'll optimize it experimentally if it's critical, otherwise, it's just 47u/4R7 just out of habit and because those values always work at least acceptably.


All times are GMT. The time now is 02:54 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2