These Diodes.???

[jpfamps]

A concern I have with "lifting" the cathode connection on fault condition, is that if the valve has shorted to the cathode, the cathode can float up to the HT, the cathode heater insulation breaks down and the HT is now imposed on the filament supply; the subsequent co-lateral damage can be spectacular!

I see evidence of this type of fault on a fairly regular basis, mainly it should be said in Fender amps that don't have an HT fuse......

It seems to me the best form of protection in a valve amp would be to shut down the HT completely on over current detection.

Failing this is an appropriately sized fast blow fuse in the HT after the main filter cap but before the OT centre tap seems to work very well.

[pdf64]

'I see evidence of this type of fault on a fairly regular basis, mainly it should be said in Fender amps that don't have an HT fuse......'

The evidence being that the 2 x 100 ohm heater ground reference resistors have blown open?
What gets me is that that fault should have pulled all the tube heaters up to VB+, which then could be expected to damage the heater cathode insulation of at least some of the tube compliment.
But I've not seen that, ie blown 100 ohm resistors and evidence of tube insulation damage; does that strike a chord with anyone?
Pete.

[JoeM]

Yes, I've seen the 100ohm resistors on the filament supply burn up with a 6L6 failure, but don't recall that any other tubes had problems due to the failure.

[jpfamps]

'I see evidence of this type of fault on a fairly regular basis, mainly it should be said in Fender amps that don't have an HT fuse......'

The evidence being that the 2 x 100 ohm heater ground reference resistors have blown open?
What gets me is that that fault should have pulled all the tube heaters up to VB+, which then could be expected to damage the heater cathode insulation of at least some of the tube compliment.
But I've not seen that, ie blown 100 ohm resistors and evidence of tube insulation damage; does that strike a chord with anyone?
Pete.

Had a Fender Blues Deluxe in last week with burnt 100 ohm resistors, melted insulation on the filament wire, and all the pre-amp valves had gone noisy, I assume due to the cathode heater insulation being compromised.

It's very common to see the hum balance pot smoked on late 70s Fender.

[pdf64]

Thanks, I have a theory that a proper PT heater secondary centre tap to ground was to be preferred over an 'artificial' centre tap using regular 2 x 100 ohm (1/2 watt) resistors, due to the ability of the heater circuit to withstand B+ fault currents whilst maintaining a solid ground reference (thereby preventing the heater circuit being pulled up to VB+).
I'd never seen such tube damage (in addition to the bad 6L6) but I only rarely do repairs.
Higher power 100 ohm resistors could be used to mitigate for the above failure mode I suppose, if a PT lacks a heater winding centre tap.
Pete.

[R.G.]

A concern I have with "lifting" the cathode connection on fault condition, is that if the valve has shorted to the cathode, the cathode can float up to the HT, the cathode heater insulation breaks down and the HT is now imposed on the filament supply; the subsequent co-lateral damage can be spectacular!

Hmmm. Good point. I've only ever seen one tube with a broken heater insulation, but that doesn't mean that the right failure won't make it happen.

It seems to me the best form of protection in a valve amp would be to shut down the HT completely on over current detection.

There are LED-to-photovoltaic optoisolators. If you put a 1kV MOSFET in series in the B+, and hooked its gate to the PV side, then putting current into the LED turns the MOSFET on, and no current turns it off. That could be arranged to happen faster than a fuse could clear. With a little circuits work it could be turned into a floating soft-start current limit as well as a shutdown.

[diagrammatiks]

protection diodes. Not exactly where I would put protection diodes.

A 5 watt 1 percent silicone coated mini wirewound is not that expensive and not that big.

If you look at a lot of the late 90s to mid 2000s marshall and fender schematics you'll see protection diodes everywhere.

It's like someone at the eagle computer just learned about them.

[Enzo]

It's like all the places they thought they would prevent having to pay for a warranty repair.

[Steve Conner]

First of all, I believe the designer of the Haze amps is British. I'm pretty sure he hangs out on one of the gear forums too.

Now, outsourcing. The Chinese factory will have an inventory of Chinese made parts that they stock. Those silicone-coated wire wounds might seem cheap, because as hobbyists we go by the prices on Farnell or Mouser. But say they were made in the USA and had to be special ordered by the Chinese plant, shipped internationally, got taxed and so on. They could end up much more expensive than a Chinese metal film resistor and diode. That might seem piffling, but when you're making 100,000 of something, all the pennies add up.

And yes, I was going to mention the HT on the heater winding thing.

[jpfamps]

There are LED-to-photovoltaic optoisolators. If you put a 1kV MOSFET in series in the B+, and hooked its gate to the PV side, then putting current into the LED turns the MOSFET on, and no current turns it off. That could be arranged to happen faster than a fuse could clear. With a little circuits work it could be turned into a floating soft-start current limit as well as a shutdown.

I like the sound of this.

As more amps are incorporating some form of HT regulation with a series MOSFET it wouldn't take much effort to have some form of shutdown on fault condition.

Having said that, the failure mode of the series pass element is often to go short circuit, which obviously will negate any attempt to shut off the HT, so perhaps a dedicated MOSFET to shut of the HT would be a better bet.

[mooreamps]

..... and I'm still going to say that using combination bias is a better and safer alternative...but you people will automatically shoot that idea down because it's coming from me, and not from others......

-g

[R.G.]

Having said that, the failure mode of the series pass element is often to go short circuit, which obviously will negate any attempt to shut off the HT, so perhaps a dedicated MOSFET to shut of the HT would be a better bet.

Yeah, it's always a sifting and weighing process. Fixing one exposure opens the door to others. The trick is in finding protection schemes where the protection scheme doesn't add other failure modes.

If you're interested, I'll do a schemo up for the shutdown circuit and post it.

..... and I'm still going to say that using combination bias is a better and safer alternative...but you people will automatically shoot that idea down because it's coming from me, and not from others......

Poor Gary. You just want to play with the big boys and nobody loves you.

Man up, Gary. Mother Nature doesn't care who says what. Technical ideas are either right (that is, they match the way the real universe works) or not. If they do, it doesn't matter much who says what, because the universe just works the way it does and the concepts which match this best will work; others won't. As I've mentioned before, you can't BS electrons. They will do what they do, no matter what you say.

I think you may be having some confusion here with the meaning of the words "better" and "safer". Unless you define how those are measured, then they have meaning only inside the head of the person saying them. If, for instance, "better" means "more power from the available power supply", then combined bias is definitely not better - it lowers the available power a little. That was, in fact, one of the technical motivations for going to fixed bias historically.

I'll repeat what I asked you back on March 25, in the preamp cathode bias capacitor thread:

In the spirit of channeling Spock and Data, I mean this purely out of curiousity and with all due respect.

Since the internet treats you so badly, why are you here?

It doesn't seem to work well for you. You don't want to participate in ways that the rest of the internet seems to want to work. Although I haven't done any looking, I have read here that your participation was banned at other forums.

What's positive enough about it to make it worth what you clearly regard as abuse?

[diagrammatiks]

First of all, I believe the designer of the Haze amps is British. I'm pretty sure he hangs out on one of the gear forums too.

Now, outsourcing. The Chinese factory will have an inventory of Chinese made parts that they stock. Those silicone-coated wire wounds might seem cheap, because as hobbyists we go by the prices on Farnell or Mouser. But say they were made in the USA and had to be special ordered by the Chinese plant, shipped internationally, got taxed and so on. They could end up much more expensive than a Chinese metal film resistor and diode. That might seem piffling, but when you're making 100,000 of something, all the pennies add up.

And yes, I was going to mention the HT on the heater winding thing.

Well ya. someone looked it over and decided that a few more 1n4007s was cheaper then a higher wattage resistor.

That's not to say that protection diodes aren't a good thing. I use my fair share...

It's just that they might not be the best idea on that particular situation.

[trem]

Kind of funny/impressive that they have a bias pot for each tube. All amps should be built that way.
Best

[Enzo]

It's just that they might not be the best idea on that particular situation.

We can sit around and come up with "better" ways all day, but in the real world, out of the hundreds of thousands of amps out there, how many times have you encountered these cathode diodes causing a problem? Or making some other problem worse? I haven't seen it.

Kind of funny/impressive that they have a bias pot for each tube. All amps should be built that way.

That may be impressive, but wiring the cathode circuit to pin 7 of the EL34s is maybe less than impressive. Not to mention wiring their heaters to pins 2 and 6.