MOSFET Follies Clairification

[R.G.]

... the purpose of direct coupled followers is to drive the grid of the power tubes into conduction. Where does the current from this conduction come from/where does it go?

It comes from the + side of your new power supply and goes to the ground attached to the cathode of the power tube; from there it flows back to the center tap of the new power supply which you have so thoughtfully attached to the ground of the main power supply. This current only flows when the output of the follower is higher than the cathode of the power tube, so it comes in pulses at the top of the signal waveform. Presumably it flows to the other tube through the other follower on the alternate half cycle of the signal.

Meaning, when I choose a power transformer for this setup, should I also be accounting for the current the grids will be conducting when driven into AB2?

Yes, probably. Yes, you must. And "probably" because sometimes or many times the current will be small enough to ignore.

And if so, about how much current will that be per grid?

]
Good question. Tube data sheet will have some info on that. It depends a lot on how far you drive them into conduction. But it will be nowhere near the plate current, as the surface area of the grid in the electron stream is nowhere near the area of the plate. And it'll be comparable to but larger than the plate current of a driver tube. I'd call it 5-10ma to start with.

About the zener ripple rejection. The current rating for the zener is the current that's going to be dumped to ground as part of the regulation, correct? I'm making sure the transformer I'm looking at is up to the job before I get too far ahead of myself. I'm not used to counting milliamps, in my day job, 5 amps is the smallest increment I care about.

You're hugely in front of people who don't care about current at all in their day jobs! And in front of people who don't think about what they're designing.

A zener regulates by eating all the current it's allowed to by the current limiting device in series with it (usually a resistor) and the load which siphons off some of the current. If the load is zero, the zener eats all the current it can. When the load eats all the available current, the zener can no longer regulate and the voltage sags. Between the two, the zener holds the voltage approximately constant as the load varies.

So to design a zener regulator, you first figure out or estimate your load current and its variation. If the minimum is zero, it's zero and the zener has to be designed to eat all the maximum current. If there is a minimum load, the zener can be designed to eat only the min-to-max currents. You then pick a zener which has the right voltage, and has enough power rating so that zener voltage times the max current in the zener (which happens at minimum load current) is smaller than the zener power rating plus any safety factor you want.

Finally you calculate the limiting resistor, sizing its power rating for the max voltage times current it will see, and adding any safety factor.

If you're a don't-burn-it-up kind of guy, you'll then go back through the calcs for what happens when the load goes to truly zero (cable broken, maybe) and when the zener and/or load is shorted.

[defaced]

Excellent. Thank you for all of your help. I'm going to attempt to prototype this this weekend before I go and buy all of the "right" parts/etch boards, etc. I'm thinking the 6VA isolation transformer I have in mind won't be big enough. 50ma doesn't go to far I'm finding out, but I'll take some measurements and find out.

[defaced]

Ok, finally got time to build this up and get it tested. Of course there's problems. Full schematic of the driver and bias circuit attached.

How easy is it to damage these MOSFETs? I'm getting some weird bias voltages, both at the 220R resistors and the grid wires. At the two of the 220R resistors, I'm getting -3v DC or so, and on other two I'm getting +150v DC (I'm using a 6VA isolation transformer, but the secondaries are a touch high at 150v AC).

Right now I'm just powering up the assembly to see idle voltages, this hasn't been connected to my amp at all. If I remove the bias circuit and connect it to a simple transformer/rectifier/cap supply, I get proper bias voltage. So I'm thinking something in the driver circuit is toast, and well, that basically only leaves the MOSFETs. If I can find my camera charger, I'll get a picture of the build up here.

Also, for design purposes, it seems that 250v would be more than enough for the 0.047u coupling caps, yes? I'm currently using 400v, but I'd like to use the smaller footprint of the 250v ones if possible.

[tedmich]

Ok, finally got time to build this up and get it tested. Of course there's problems. Full schematic of the driver and bias circuit attached.

How easy is it to damage these MOSFETs? .

Ouch! Its very easy to trash them with errant static electricity and/or miss-wiring, I hope they're AOK!

Whats PSpice on this monster look like?

[defaced]

I have not modeled it. I no longer have access to Multisim and I have no time to learn another simulation program, it's taken me two months just to get this far. I'm just giving it a go based on what I'm hearing from the amp, and for the pursuit of learning.

Pics of the MOSFET board attached. Not the jumpers for the pos and neg voltage rails. I have some 2SK3564 MOSFETs, would those be an acceptable substitute for the IRF820s I'm using?

2SK3564(Q) Toshiba MOSFETs

and

IRF820PBF Vishay Semiconductors MOSFETs

[Tage]

Hey I want to thank you, Defaced, for this thread, and all the contributors, R.G. especially. I have been designing my own mosfet drivers, and this has re-affirmed a lot of my assumtions and corrected more than a few mistakes.

I was wondering what VA size you ended up selecting for your transformer?

[defaced]

You are welcome. I like to give back when I can since this place (and internet forums as a whole) have taught me most of what I know with regard to electronics.

The transformer pictured in the images above is the transformer that I am still using. I just looked it up and it's a 6VA transformer. I have not done detailed analysis of this setup (still) after two years.

[Tage]

alright, thanks.

I don't think you really need to analyze this circuit, it seems very well designed, my design will only use one mosfet per push pull side though, but your diagrams were very helpful, especially with using an auxiliary transformer.

[Steve Conner]

I strongly recommend a 12v Zener between gate and source of each MOSFET. The gate really is insulated, and if it gets more than about +/-30V away from the source, the device will fail sooner or later for "no apparent reason".

Some fets have the Zener built in, but not all of them, and the datasheet doesn't always say. In a linear circuit with enough voltage floating around to damage the gate, I just always use the zener, no exceptions.

Some applications need two zeners in series back to back, so the gate can swing down to -12v, but I doubt this is one of them.

[Tage]

Yeah I had it suggested to me to use FET's with build in Zeners, I knew about using them from building pedals, ala SHO, but I never realized there were ones with built in protection until I looked up a part from a input buffer in a boss bd2. Now I use them all the time in effects as it's one less component when I'm crammed for space or just want something small.

Would you ever bother to use an extra Zener even with built in ones, Steve?

[Steve Conner]

Yes, I probably would. At least I would in a design that I was planning to sell or publish. My reasoning is:

1. Some popular parts are sourced by several manufacturers. Maybe IR's IRF820 has a protection Zener, but ST's clone of it doesn't. So then I have to tell my purchasing department (or my readers) to only buy the IR ones. If I use my own Zener I don't have to even care.

2. New MOSFETs come out all the time and old ones get discontinued. So, I get a call from the factory saying that they won't be making any of my widgets because the IRF820 is back ordered for 3 months. Shopping around I notice that Mouser have lots of the new OMG44N50 which offers twice the performance for half the price, but it doesn't have gate protection. Just as well I left a spot on the PCB for that Zener!

To sum up: the time saved shopping for FETs exceeds the time spent adding the Zener. This may not be true any more for high volume manufacturing, but I don't do that.

[Tage]

Yeha this is a one off thing for me, but I understand exactly were you are coming from. I also have ensured that the FET's I'm using have the built in Zeners.

[exclamationmark]

On a side note... can you parallel source followers without any problems, or will they do that current hogging thing that non-matched devices tend to be fond of doing?

[R.G.]

On a side note... can you parallel source followers without any problems, or will they do that current hogging thing that non-matched devices tend to be fond of doing?

?????

Whatever for?? MOSFETs tend to come in relatively huge current and power ratings compared to tubes. Why would you need to parallel them in a tube amp?

The straight answer is that it depends on the application. MOSFETs do not current hog like BJTs do; their semiconductor physics are different. However, they are still prone to differences in gain and cut-in threshold. So for things like power amplifiers, some attention to ballasting and matching is useful

[exclamationmark]

?????

Whatever for?? MOSFETs tend to come in relatively huge current and power ratings compared to tubes. Why would you need to parallel them in a tube amp?

The straight answer is that it depends on the application. MOSFETs do not current hog like BJTs do; their semiconductor physics are different. However, they are still prone to differences in gain and cut-in threshold. So for things like power amplifiers, some attention to ballasting and matching is useful

I'm doing some weird things with sweep tubes, which can easily draw 100mA+ peaks (and their grids are rated at something like 2 watt continuous dissipation). Normally I could get away with using something like a 20 volt supply to feed the mosfet, but I'm using a sort of strange topology where they are set up as pseudo cathode followers (which means the grid has to get up to something like +60 volts to get to full swing, leading to big instantaneous dissipation numbers when feeding multiple tubes).