Simple rig to match mosfets?

[phydauex]

To all of youse who match your own mosfets for amps like the ampeg SVT and others. What kind of a set-up do you use and particularly in regards to current matching for the Ampeg SVT4? I was reading on the Pass DIY site about his setup, but it seems overly "flexible" for what I need to do -simple current measure. Also, would it be more meaningful to try and make a rig that simulates the actual circuit, ie. @+/-80volts on the drains and using a .47 ohm resistor on the source and a preset gate voltage (3.6v)?

Anyway what do you do to match output devices?

I recently scored a bucket of unused IRFP240/IRFP9240 mosfets and would like to sort them for use, esp in the SVT that I'm finishing up.

Matt

[Jazz P Bass]

This link has a nice DIU tester.
Link: Power Amplifier: Zeus MOSFET Matching

[jmaf]

No specific to Mosfets, but maybe this will give you ideas: http://www.geofex.com/article_folder...h/fetmatch.htm

[MarkusBass]

Matt,

Some time ago I was using the circuit listed by Jazz P Bass (from Susan Parker). But I found out that MOSFETs matched in this way didn't seem to be matched when used later in SVT3Pro. So I started matching them directly in the amp. I usually have at least one MOSFET desoldered and I solder three short wires with clips, I mount a new MOSFET to a heatsink and connect it with clips. Of course the heatsink is isolated from the amp ground. In this way the MOSFETs are matched using exactly the same conditions (power supply, gate voltage). Works great and transistors in the amps fixed by me are matched much better than from AMPEG factory.

Mark

[Jazz P Bass]

So you are measuring what, I am curious.
The voltage across the ballast resistor?

[MarkusBass]

So you are measuring what, I am curious.
The voltage across the ballast resistor?

It's the current that has to be matched. So I can measure the current but, as you have noticed, measuring voltage across the ballast resistor is even simpler. So, yes, I measure the voltage across the ballast resistor. Of course I have to compare it with voltages on other ballast resistors because adding a new MOSFET can change the other voltages.

Mark

[phydauex]

Hey Mark, it worked like a charm. I did some rough grouping earlier today with a low voltage supply and measuring across a 10ohm resistor on the source. This gave me the bunches to install, then using a set of clip leads on each side got them all dialed in to with a few tenths of a mv per side and 1 mv side to side. Great idea! The 9240's seemed to be all over the place, while the 240's were much easier to get matched, I don't know if others have experienced that or not, maybe it was just my odd batch of devices.

Anyway, so I'm jumping up and down and getting ready to put it all back together once and for all and I turn it on to do one last check of everything. The relay on the other channel is clicking on and off like a stopwatch. What the hell?

###I've been looking at it and if I understand the circuit (big ??) the relays click on when there is dc on the rail, when I turn on the power, the relays starting clicking. Then if I put the meter from the center rail across the jumpers to the next trace(where coil and 3.3 ohm resistor are), the clicking stops as long as I have the meter across there. What does that tell you?

This amp has had numerous "mystery" (well some were things I missed), issues that have been resolved, I thought.... What would cause the relays to open&close. Maybe another solder joint come loose? Or maybe something shorting. It was doing fine all evening, then just when I'm ready to wrap it up, this starts.

I'm not liking this amp very much right now.

matt

[MarkusBass]

There are no mysteries in this amp if you understand the schematic. Have you actually measured whether there is DC on the output? I'd say that you incorrectly isolated the MOSFET from the heatsink (and you have V+ or V- on the output). But give us a chance and measure at least the voltage on the output. It's hard to guess what's going on if you don't provide information about current state of the amp .

Mark

[phydauex]

Ok, a little more info....

1) I was using the current limiter for the ac when the relay was clicking on/off.

2) Plugging straight into 120v, the relay is no longer toggling, but clicks on and stays on. There is 30mv in ref to ground at the relay (TP11), there should be 7.2mv. I'm assuming the output should be measured against ground. All other voltages in the power section measured against the output rail. Looking at the schematic, I see several places where +dc could be coming from: C213 (10uf across the base of Q208), D211 (again from base of Q208 to output- replaced), the zener string between B-Hi and the ouput (was replaced also). Is there any other section that could be letting +dc thru?

###went back to check things now that I have the bias set for both channels. Both A & B show about 29.4/29.6 mv on the front end of their relays, so maybe that's within reason? I don't know how much variance I should expect from the factory schematic, both channels were basically completely rebuilt, so maybe this is ok afterall.###

3) The mosfets are mounted on the insulator strip, as the old ones were, I didn't see any holes or tears in it to expose the heatsink.

4) The mystery part comes from spending a week trying to find why the bias rails were only about +/-1v (CH B) and all resistors/transistors/caps/diodes replaced or checked out good, only to have it return to normal(+/-3.6v) after remounting the board in the chassis when I gave up on it. Solder joints, bending the board just right? I don't know.

This was like that, because it was stable and voltages correct all the while I was getting the outputs back in, then all of a sudden it's doing something weird, that it wasn't before. I'm sure there is a specific cause for everything, but sometimes this amp seems to act in an arbitrary manner.

And to be honest I don't truely understand the circuit, I'm gaining a much better feel for it, but it's still a learning process.

[phydauex]

So, I believe it's fine now. I got it all back together and have been looking at the output on the scope and it looks swell. The light dc on the output that I had before is gone and both channels are at 7.5mv (down from 30mv). Now I didn't do anything but put it all back together and that voltage changed. That's the thing that has been troubling me all along. As long as it stays there, I'll be happy.

Also at some ppoint while testing the output, the bias dropped like in half. Everything was biased at 25mv and after I ran it a bit I checked it again and both channels were down to about 11.5/12mv. ????

I reset them back up which took all of a micro nudge on the pots, are they prone to moving with shock or vibration? I wouldn't imagine being a bass amp, but that's what I've experienced.

Anyway I appreciate all the advice and suggestions from the more knowledgable persons here, and the paitience real or imagined with some of us newbies.
Matt

[Jazz P Bass]

You really have to be aware of the ground schemes on some of these amps.
Dirty ground, signal, power.
Depending on the amp, sometimes they are not connected until the amp is fastened into the chassis.

[MarkusBass]

You really have to be aware of the ground schemes on some of these amps.
Dirty ground, signal, power.
Depending on the amp, sometimes they are not connected until the amp is fastened into the chassis.

You are right. I haven't seen such a bass amp but I got once an audio amp (I think it was Yamaha) with a DC on the output (-10V). The guy was trying to fix it for a month. And I found out the the ground connection in this amp is through a screw on the back panel to the chassis. And he just took the screw out. So he was trying to fix the amp which actually had no problem (appart from this little screw).

Mark

[MarkusBass]

1) I was using the current limiter for the ac when the relay was clicking on/off.

This explains a lot. I'm not sure what kind of current limiter it was. But if you think of a variac and how it works with a solid state amp, you have to be aware of the differences between solid state amps and valve amps. With valve amps when you increase the supply voltage all other voltages are increasing. So everything is safe and nice and you can easily test the amp. But this is not the case with a solid state amp. If you have a +/-15V stablilisators, which require e.g. 18V on the input, with lower voltages they will simply not work. The same story is with a relay. A 24V relay will not work with 12V. So in case of solid state amps and a variac you may assume that some parts of the amp will not work at all (and this is expected behaviour). That's why I don't use variac. It only prevents from blowing the output transistors but does not help you with debugging the amp.

Regarding your amp I'm only worried with this drop of the bias current. This shouldn't happen. It is possible that the current grows after warming up the amp but only 10-20%. The bias trimmer is very sensitive but I would check all the resistors around it.
Also take a look at the ground connections: from the transformer to the board and from the board to the chassis.
Did you test the amp with a dummy load at full power? This would answer whether the amp is OK or not yet.

Mark

[woodyc]

Here's a simple rig I made for measuring power mosfets...

It just has a few resistors and a cap.
And you just need to hook it to a DC power supply or two.
Oh, and a couple of voltmeters.
And of course you want to know if it can pass a clean signal, and what the gain is, so you need a signal generator and a preamp and a scope, naturally.
Ok, maybe its not so simple.

[pontiacpete]

How about a schematic for that little rig? Or inside pix?