• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Maida Regulator PCB

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
You bring up an interesting point. How much is too much?

A couple of years ago I tried to build a maida regulator to use as a choke substitute. I did it on perf board and the result was something of a hash. It fouled up the room with acrid smoke and scared the dog. I went back looking for the mistake but couldn't find one. It's still around here somewhere collecting dust. I was using it with a 56uf cap IIRC, and I wonder if that might have been the problem rather than a wiring error. I was just getting ready to revisit that project when I stumbled across this excellent thread.

I'm reconsidering whether the maida circuit is really such a good idea for that application. Specifically, I'm concerned that a power supply without reasonable levels of capacitance will not have adequate reservoir for a tube amplification application. You've obviously given this plenty of thought; what is your take on this please?

Let me ask you a question, why does it need a reservoir on the output? A good regulator will provide a low zout without capacitance on the output. Are you dealing with a topology that needs a capacitor on the output for some reason?
 
Let me re-phrase that.

Which devices are in MELF packages? I'm familiar with the SOD80 and related packages. Pad layouts can be made such that they support DO219, SOD123, or others as well.

Manufacturing instructed us to quit using them (MELFs) some eight years ago due to solder quality issues. Apparently the devices like to roll around.

If you are hand soldering them, it is not really an issue but I prefer other devices wherever possible.

They are all 15V zeners in my implementation, mouser part number: 78-TZM5245B-GS18
 
This...
Or, how much resistance is enough before a capacitor? Say, I have regulator feeding B+, plus 2 resistors w/ 2 caps behind the resistors for the next stage of the amp.

You're going to have to look at what the peak current will be at startup charging those caps through those resistors and see if that violates SOA for the pass device that you choose.
 
A good regulator will provide a low zout without capacitance on the output

A good stiff B+ is on my list of things that make a hi-fi tube amplifier sound good. I'm not certain that belief is universal, but it is a conviction to which I cling. Consequentially, the Maida regulator's problem with capacitance got my attention. As you point out, the low z characteristics may make my concern moot.

I can think of several applications for the Maida. My list includes a versatile HV power supply that would be suitable for a variety of builds. The capacitance issue is at odds with versatility. I have an Aikido preamp that needs some tweeking; the Maida would be perfect for that application, no question. It's the SET project that would most benefit from low ripple and it is that application that most raises the concern.
 
It all comes down to knowing the limitations of the circuit you are using. There are better regulators that have lower Zout at higher frequencies. Where the Maida wins out is in simplicity (no floating power supplies), efficiency (compared to shunt regulators), and it is pretty compact. I am putting five of these into one push-pull amp.

Tom Christiansen (tomchr) has a regulator for sale that has built-in current limiting so that it can start up into a capacitive load safely. It is probably a better performing regulator but requires a floating transformer winding to power the error amp. That could be a tool where top performance is required.

I have used Maida regulators for Push-pull output stages and the stiff supply really gives the amp that extra oomph during a hard-hitting bass note.
 
Received my boards. Very nice quality. Thanks.

Any recommendations for a not too expensive MOSFET with isolated tab? Say 600V rated and up? I just don't want to risk it at 500V+ with the shoulder washers etc.

Also, does anyone make a 317 or same-pinout equivalent with isolated tab?

Edit: OK, I see in #30 that you found an isolated 317. Can you share the info? Thanks.
 
Last edited:
Isolated LM317: Mouser p/n 513-NJM317F

FQPF4N90C is a pretty cheap 900V TO-220 isolated package(this one makes a good source follower to drive power tubes as well with low Crss). I use FQAF11N90C which is an isolated TO-3P package (similar in size to TO-247). I think they are a bit pricey but should be able to transfer heat better than the smaller part. Both are 900V parts. I bought a tube of the 4N90Cs since they work in so well in so many places.
 
A good stiff B+ is on my list of things that make a hi-fi tube amplifier sound good. I'm not certain that belief is universal, but it is a conviction to which I cling.

I agree with the sentiment, but lately I've been appreciating the simplicity of the aikido, which compensates for all PSU hash, hum, sag etc., so my new idea is to leave the PSU as a simple capacitor and just feedback compensations signals.

This then re-affirms the role of an amplifier as a voltage controlled power supply, and uses the power amp's active elements for the whole task.

It should be cheaper too ;)
 
I agree with the sentiment, but lately I've been appreciating the simplicity of the aikido, which compensates for all PSU hash, hum, sag etc., so my new idea is to leave the PSU as a simple capacitor and just feedback compensations signals.

This then re-affirms the role of an amplifier as a voltage controlled power supply, and uses the power amp's active elements for the whole task.

It should be cheaper too ;)

Class A line stages are one thing, but power amps are another. An interesting exercise I engaged in once was to calculate the necessary capacitance to keep half of a cycle @ 20Hz from sagging more than a couple of volts in a 100W stereo KT88 amp. Try it.
 
It's not really that bad. The key is having flux on the joint, a bit of solder on the iron's tip and touch the iron so it heats the component and pad at the same time.

Getting the first side locked down is the hard part. Even if the joint doesn't look all that great, move on to the second side (which will be much easer since the part can no longer move). You can always paint a little flux on the first side and re-flow the joint to make it look nice.

Don't forget to clean the flux. The cheap option is rubbing alcohol which requires some elbow grease. I have used 3M flux cleaner which works very well but costs more.

Let me know how it goes.
 
The problem with relying solely on PSRR is that when the power supply voltage fluctuates, the operating point does as well. When you apply a signal to the input you get IMD as a result.

It is possible to come up with many clever ways to null out power supply noise so that it does not appear on the output with no input signal. Coming up with an amplifier that is truly(nearly) unaffected when a signal is applied is another thing and usually involves lotsa feedback.

The best solution is good power supply and good PSRR.

I'd prefer not to turn this thread into a Regulated Vs. Unregulated discussion. This was meant to be a discussion about a particular type of regulator. If you want to start that discussion, please open another thread and I will happily join you there:).
 
In my experience, the easiest way to solder any surface mount device is to tin one pad with solder.

Then hold the device with tweezers, position it in place, and heat the lead and solder bump on the pad. When it flows remove the soldering iron and let it solidify.

Then let go of the component.

Now you can solder the other pads.
 
Jack,

I'm not sure what you are saying. Dominates all else in what sense?

I was just trying to say that focusing solely on giving circuits awesome PSRR or solely on power supply performance is not the way I would do things.

I also know that there are better regulators that have lower Zout at higher frequencies than this one. It is just a good compromise between performance/size/simplicity/efficiency(vs shunt regulator) IMHO.

TheGimp,

Yeah, there are lots of ways to do this. That sounds like a pretty good way. Basically there are three things you need to deliver with two hands, solder, component, and iron. You just have to figure out a place to store the solder so that it becomes a two-handed job. I store it on the tip, you store it on the pad.

Just a note: the joint should be a nice fillet. If solder is bulging out of the joint, there is too much solder. That's okay, just paint some flux on and touch the tip to the joint and some solder will be pulled to the tip and you should have a nice fillet joint (if you do this without flux it will form a peak as you pull the tip away and will make an improper joint).
 
Interestingly my ideas of nulling out supply noise in a power amp are discussed here by the master!

http://www.tubecad.com/march2001/2001_03.pdf

Not to turn this into a discussion about regulated vs unregulated supplies, but there are many ways to skin a feline. Assuming the end goal is to have the least amount of power supply garbage come through to the output, cancellation as per Tubecad or, much earlier, Loftin-White are options.

But a just as valid option is to designate a control loop to get rid of the garbage and have a clean supply. The Maida regulator achieves this goal quite nicely.

~Tom
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.