Are MOSFETS reliable ?

[kg]

Well, I have a couple of amps full of FETs that have been working for more than 10 years. They show no signs of running out of electrons.

by contrast, i've a couple of >50 year old tubes still performing perfectly well, despite nearly daily use!

[Tilman]

Thank you very much R.G. !
Now I understand what you were saying and that brings me to the conclusion ( I am kind of slow...I know ) that a MOSFET is better suited for this application if the followers drain/anode voltage is close to the drivers anode voltage.
If a tube as a follower is used, the followers anode supply should be higher that the drivers supply voltage to make sure the followers grid/cathode can follow the drivers signal upwards easy enough.
Tilman

[Steve Conner]

by contrast, i've a couple of >50 year old tubes still performing perfectly well, despite nearly daily use!

In that case, we could conclude that tubes are more reliable, because there are no 50 year old MOSFETs. But somehow the military and space industries made the leap of faith to those touchy new semiconductors.

[redelephant]

I don't know about way off. However, if what you mean is that the tubes can't swing further in to conduction, they can do this with more grid drive. The same tube and OT will swing to higher currents with positive grid drive as long as the driving element can push the current for the grid. In transistor terms, the output tube "saturates" to a lower voltage and higher current.

Yes, agreed. It's just if the OT is high impedance as if designed for class-AB that the little extra you get is in my brief experience so small it's barely measurable, and certainly not noticable in use. So I think to really benefit from driving grids, the OT needs to be changed in accordance with design goals. That's why I think it would be interesting to measure and find out.

The MOSFETs work their stuff on the other end, on the tube that's driven on. In a class AB amp, one tube is (nearly) always off, so driving it off further makes no difference at all. It's the tube that's driven on that gives you the extra power and softer clipping.

Agreed.

What makes the clipping softer is that instead of the grid simply stopping at a fraction of a volt over the cathode voltage, it can be pushed a certain amount more positive. The grid's being clamped to 0V from the cathode can be set up to be as razor-sharp as any opamp clipping. The guys who used to make logic circuits out of tubes use to tie the grid of a triode to B+ with a 470K to 1M resistor to force the grid to cathode value by grid current. This was referred to as "running in clamp", and produces a razor-sharp entry into distortion.

I see this as logical too. Perhaps it's the tubes I use, or the way they are driven in the first place with a current that increases as the signal goes up, or a combinatino of the two. But in my attempt the clipping became harsher, and not softer. I wish I did some more testing, and I guess I probably will try it again, but it appeared to me that using MOSFETs to drive the grids made the tube clip harder. The tubes I use (6AS7s) are low impedance types that allow the plates to go pretty low, and maybe there just wasn't enuff left to make room for softer clipping. Or I could of course have problems with my MOSFET follower. It's possible I had high frequency oscillations as they started conducting the grid current (which btw peaked above 50mA), or some other strange artifact I didn't look for or pay attention to.

In my amps that have interstage transformer coupling between PI and output tubes, I think I get a softer clipping, which is probably due to the interstage tranny supplying power to drive the grids a little bit. Need to check that out as well...

Oh, so much to test and learn, so little time...

[jrfrond]

The original question hasn't yet been addressed: Are MOSFET's Reliable?

Here's my take on this: I've been working on amps utilizing power MOSFET's since the early 80's, and I have to say that I personally DO NOT trust them. I've seen too many GOOD designs go up in flames, quite literally, particularly when I was the Service Manager at Samson, and Hartke had the new 3500 and 7000 MOSFET designs. The returns were UGLY! There was no choice after awhile but to have Inter-M, the Korean vendor, adapt a bipolar transistor module for replacement, which is still the only replacement available.

Sure, I've seen some bipolar designs that were latent barbecue's too. Phase Linear comes to mind. However, bipolar designs are eminently more stable. It's ironic though, because in the early days of MOSFET amps, we learned that their negative temperature coefficient would simply shut down the output stage. And it did..... as long as it didn't cook first!

[Steve Conner]

Hitachi's advertising spiel about the lateral MOSFETs being short-circuit proof was wildly optimistic. They usually do cook before they shut down, and that is why they got the bad rep for reliability. In reality they need SOA protection like any other device.

Also, any design that uses vertical MOSFETs in the linear mode is not a good one. They are designed for switchmode use, and when used as linear amplifiers they are unstable and have lousy distortion performance.

[R.G.]

I also think it's time to address the original question:are MOSFETs reliable?

The answer is - no more, no less than any other thing in electronics.

There are no reliable parts in isolation. There are only parts which are used in ways which let them survive well - or not. Period. A solid lump of stainless steel may or may not be reliable, depending on the use to which it's put. It's not going to be very good for propping up one corner of a steel-melting furnace, for instance, nor for holding hydrofluoric acid or aqua regia. A layer of glass 20V thick will last **forever** if used to insulate voltages less than 20V.

If MOSFETs have a bad rep in the music world, blame the designers or the manufacturer's advertising, testing, and hype. It's worth noting that bipolar power devices have to be extensively protected in audio power applications because of their tendency to self destruct in second breakdown. If tubes have a reputation for being reliable under overload, it's because they have a setup which encourages this: longer thermal time constants because of bigger mass, hermetic sealing in a vacuum, and a tendency to stay high impedance.

A proper design with MOSFETs, bipolars, magnetic amplifiers, tubes, whatever, is all equally reliable because *the designer applied proper effort to forecasting usage and MIS-usage, and designed accordingly*, whether that was to make the resulting product so tough that it worked till it melted down or alternatively it quit hurting itself when misused by a clueless owner.

So - are MOSFETs reliable? Yes.
Are tubes reliable? Yes.
Are the design details needed for them different? Yes.
Can one horse run faster than another horse? Yes - the questions are **which one, what day, what conditions, etc.**? Details matter.

[kg]

there you go, rg being all wise and stuff.



to be honest, in all my tinkering, ratio-wise, i have fried a lot more semiconductors than i have vacuum state stuff. those pesky silicon devices seem AWFULLY finicky when not protected by dozens of housekeeping components. BUT, get the housekeeping right, and like steve says they'll last for a very, very, VERY long time.

[Tilman]

I think I have chosen a misleading thread title. My fault, sorry. My question actually was not : Are MOSFETs reliable per se. I think we all agree they are, if the design they are in is right.
My question was : Are they reliable as a follower to drive output tube grids in a musical instrument amplifier. We all know that musical amplifiers have to be kind of bullet- ; idiot proofed to survive the daily beat up.

Maybe I should have asked : Would Peter Traynor have installed MOSFETs at this position in his amplifiers or not.
I now think: Maybe. ;-)

Regards!
Tilman

[Steve Conner]

Well, I've seen schematics for a Yorkville guitar amp that used a high voltage MOSFET in the standby switch circuit. So maybe not Mr. Traynor himself, but there is a connection.

Tubes are certainly far more foolproof and resistant to abuse, which is why they are popular with hobbyists. Even if you wire up your new tube amp kit completely wrong, the parts have a good chance of surviving for another attempt. With a solid-state amp kit, put one transistor in backwards and half a dozen others go kablooie.

But that's not the same thing as reliable.

[redelephant]

I've never had problems with them. I can see in pwr amps where they are used at the limit of SOA there can be problems. But as small signal devices, drivers and buffers, and pwr amps to a few hundred watts they are just as reliable as any other device, as has been mentinoed several times.
That's my experience with them anyhow.

[kg]

honestly, tilman, i would have no compunction using mosfets as impedance buffers for driving output tube grids. just watch your max Vgs and go. you won't need much q-current, may not even have to heatsink 'em. because they're so efficient at pulling source up to drain you could feed them from a low positive voltage rail, like <50v. the negative rail only needs to be low enough to cut off the output tubes, probably around -50v (obviously depends on the tube choice and screen/plate voltages). i think most to-220 packages can handle 1w of free air dissipation.

by contrast my ultra-mega el84 pentode mode cathode followers need a LOT more voltage across them, and are considerably more complicated wrt number of components, socket mounting real estate, heater circuits, etc. doing it all again i would probably just go with mosfets...