Relays And Popping

[Paul Marossy]

12AX7 and kin run far cooler, less surface but far less heat. 200 deg F IIRC.

Yeah, I can touch those and hold onto them without getting burned. Power tubes OTOH will burn me (duh). I've accidentally touched some EL84s with the back of my hand once, that was not a pleasant experience. 

While I'm typing: _I_ do not like the one cathode resistor nor the stupid diode. _I_ would want to convert to two 1 ohm resistors and no diode. With one resistor, if one tube is dead you can't tell (except by extreme grid voltage and redplate in the other tube). With two resistors you can verify each tube is at a safe dissipation. However I sure know that PCB construction can foil such mods.

It could be modified to take two cathode resistors easily enough. I'm not sure what the purpose of that diode is. Never seen anything like it on any other amp schematics I have had to use for repairs, mods, etc.

[Gus]

R66  1 ohm resistor is marked fp1 ohm I am going to guess it is a fuse resistor and the diode is to keep it from opening (by clamping the max voltage to the diode drop) for short term output tube over current events.
Is the fuse a blue color? or does it look different?

[Paul Marossy]

R66  1 ohm resistor is marked fp1 ohm I am going to guess it is a fuse resistor and the diode is to keep it from opening (by clamping the max voltage to the diode drop) for short term output tube over current events.
Is the fuse a blue color? or does it look different?

I assumed that R66 was a flameproof resistor. I've never heard of a fuse resistor before. There is such a thing?
Anyway, it looks much like the blue flameproof resistors I have used in the past. Maybe you're right about that. But it sure seems like kind of a roundabout way of doing things (kind of like a lot of the rest of the amp).

[Gus]

google bing etc "fusible resistor" and "flameproof resistor"

[Paul Marossy]

google bing etc "fusible resistor" and "flameproof resistor"

I did last night. You learn something new every day!

[PRR]

> accidentally touched some EL84s with the back of my hand

I once rested the back of my hand on a 6550 idling at FULL 42 Watts dissipation for several seconds.

Yup, it is as hot as a cookie-oven.

I took a picture of the coke-bottle shape 1st degree burn on my skin (actually a small 2nd-deg blister but too small to show in photo). Hurt for a week.

> I'm not sure what the purpose of that diode is.

My guess is some green engineer "protecting" a cheap part which "can't" ever blow-up without some expensive part (tubes) failing first.

0.6V in 1r is 600 milliAmps. I forget what amplifier this is, but only the crazy-large amp tubes can flow 600mA. Unless they short plate or G2 to catode, and then they will blow the 1N4001 and then the 1r resistor. So it does not improve reliability, just increases MTTR (repair takes longer for finding and swapping-out the silly diode).

For more fun: unless available current is large, diodes can fail *short*. (They do tend to short, but in higher current circuits they soon blow open.) Now the 1r is moot, also useless. When the amp runs again, a bias-check shows ZERO tube current (because cathode resistor is shorted-out) and the tech turns the G1 voltage more and more trying to find some life in the tube; meanwhile maybe not noticing the dull red glow of broiling plates.

Diodes are REAL CHEAP at the factory, and warranty costs are high, so you do see them in places a DIY builder would never think to worry about. But this one seems to me to be a dud.

[R.G.]

That's pretty much the analysis that Kevin O'Connor went through in his "The Ultimate Tone" book on the 1R and the diode. It went into my thinking on an experimental device I made for protecting tubes and the PT.

I put a 1R to ground in the cathode of each of the output tubes for measuring current/bias, then a high voltage MOSFET in series with the resistor to the cathode. Some simple comparator and logic circuitry sensed max current on each tube, then latched off all the MOSFETs (and hence all tubes) if the current was over X. The logic messed with things like soft on, standby, LED for "over current on this tube", and so forth. The logic was all MML CMOS. Today, I'd use one PIC and do it all. Some PICs have internal comparators for just such emergencies. I'd probably put a thermistor over the tubes and one on the PT, just because it's easy once you're over the hump of deciding to do the design.