amps, resistors, and wattage

[Spivonious]

Can any of the amp experts tell me how I can tell what wattage a resistor is by looking at it? One of them in my Genz Benz is cooked and I'd like to replace it. I've read the color codes to determine it's 100 ohms +/- 5%, and looks like a larger than normal one (although not one of the big box shaped ones).

[Cygnus X1]

Pic?

Might be a 2 or 5 Watt metal oxide.

[DonP]

+1 to pics.

Schematic? If you look at where the resistor is being used it might help to guess at it's wattage.

[Chief_CrazyTalk]

If you take pics, be sure to include a ruler next to the resistor. Where in the circuit is this resistor?

Fyi if this resistor is cooked, there's a good chance it took another component out with it, more likely vice versa. Usually transistors.

[Spivonious]

I'm almost positive it's the screen resistor. It's in the power section and when it went before I got crazy bias reading and burned up a tube. I started getting these symptoms again so it hasn't burned up a tube yet.

Is there a way to test transistors? There are some on the board near it. I don't have a schematic, so it's hard for me to say what depends on what.

As I alluded to above, this happened before and both screen resistors were replaced (one for each EL34 tube).

Here's a photo of it on the board. It's the blue one.



Here's a pic of the whole board.

[Spivonious]

Also, it looks to me like some of the big caps have broken contacts. Could this have led to burning up the resisitor?

edit: They're 400V, 100uF, "KMG" caps. What does the KMG mean?

It's an El Diablo 60C combo, so I can believe the vibration knocking the caps loose.

[Cygnus X1]

Looks like a 1/2 watt metal film tied to the cathode (pin 8).
KMG is a brand name on the caps.

Those should not have broken leads unless they were bent over
in some way.
Are you saying the copper trace is lifting from the board around there?

[Spivonious]

It looks like one of the leads on a couple of caps is physically separated from the piece soldered to the board. I can't really get a good photo of it since it's a pretty small clearance. It's possible that I broke them when lightly wobbling the caps to check the connections, but if I did then I'm surprised they could put up with the vibrations of the combo.

For the wattage on the resistor, does it matter as long as I go high enough? It's just a heat rating, right?

Where do you guys go for your electronics needs? My local Radio Shack has a very small collection.

[Cygnus X1]

You could get a 1 watt just to be safe.
Keep in mind how it will fit on the board.

I can't say it is a cathode for sure, it just looks like it is tied
to pin 8.

I buy my resistors either in bulk or large quantities from
Weber or AES.
Shop around and fill out your order with some "wannagets", like tubes from AES.
Might as well get those caps while you're at it.
They will need to be replaced eventually anyways.
Tough getting good resistors from Rat Shack.
Many are 1/4 watt, and most are carbon film (yuk).

[DonP]

I'll just fill in/add to what Cyg has already said.

1) No transistors to worry about in this section of the amp.

2) screen grid resistors blow when a tube shorts and pulls mucho current through the screen grid. A low wattage resistor is used as a "fuse" in a lot of amps, because a resistor is cheaper to replace than a tranny. A screen grid resistor wouldn't be connected to the cathode, it would be connected to B+ just south of the OT tranny (usually after the choke or a filter resistor).

3) If we are talking the blue resistor in the pic, I'd say it's a 1 to 2 watt metal film. I'd make sure to get the correct wattage again, so that it'll burn up again if a tube fries. You can get a higher wattage resistor, but the next time your tube shorts, it might be a tranny you are replacing.

4) Those caps are your power filtering section, changing AC to DC. There's a lot of current going through the PCB, and current produces heat, and heat melts traces and solder. That's why older hard-wired stuff is prefered to PCB stuff, at least when you are talking high current. The solution isn't pretty - use copper wire as jumpers to handle more current, because the amp wasn't designed with traces fat enough to handle the current originally.

[Cygnus X1]

Thank you Don.
I'm still living in "cathode biased" land.

It's what I build, so I am biased.

[Spivonious]

Thanks Don! So you think it was a bad tube that caused the resistor to blow?

The first time it happened, it blew up a tube. Well, almost...the assembly inside of the glass rattled around loose. I got no sound at all with new tubes and had strange bias readings (there are test points on the back of the amp). The tech replaced both blue resistors under warranty and I was good for a couple of months. I also got a matched replacement tube for the one that blew up.

This time, I noticed severely overdriven sound and saw one of the tubes glowing red. Got strange bias readings again, although not as far off as the first time. It hasn't been powered up since. Do you really think it could be a bad tube? I'd be surprised since it happened again with a different tube.

Although it is true that the amp ran fine for two years before I replaced the tubes, and I've had two problems since. Could it be that the EL34s I chose run too hot for the design?

[Cygnus X1]

I could not find the bias scheme (didn't find the schem on Google).
You are on the right track though.

Root Cause Failure Analysis.
I would not think these would be cathode biased at all (I was joking in the prior posts)...as DonP pointed out, there is something else to look at further down the line.
You are right, an excessive draw in a bad tube...or a bad bias would "fuse out".
Just replace it, throw in some good known tubes, bias conservatively, and see what happens.

[Spivonious]

Yep definitely not cathode biased. There's a pot on the back next to the test points and a switch to go from EL34s to 6L6s.

[DonP]

Actually 100 ohms would be a closer to a common value for a cathode biased amp, and is a bit of an unusual value for a screen grid. I'm just calling it a screen grid resistor because that's what I read. I'd need the scematic or the amp in person (following traces) to tell for myself.

(Typically) Marshall uses a 1K 5 watt; Fender a 470 ohm 1 watt. 100 ohms seems too low for a screen grid.

But for cathode bias, you'd need a roughly 250 ohm at 10 watts. Cleary those resistors aren't 10 watt.