Fender Screen Resistors-interesting symptoms

[Mars Amp Repair]

I have a young tech that I occassionally give aid to. He's pretty sharp & it's my opportunity as an older more experienced tech to impart my knowledge before I'M too old . So he calls me with a typical 6L6 Push-Pull Fender amp that he's replaced the typical screen and cathode resistors. He's only getting 5watts out of it. He's already (out of desparation) replaced the output xformer.
He's checked all the voltages around the power tubes and all seems reasonably normal but the highest he can get the idle current is about 5ma.
So I double check the plate, screen, control grid, & cathode voltages and aside from the really low idle current, I find nothing obvious.
I decide to see if we can raise the idle current to at least get the power tubes closer to their normal idling current buy bridging the 27K resistor to ground from one side of the bias pot with I think a 1K resistor.
We were able to raise the idle current to something like 15ma and the power tubes began to get hot BUT the voltage on the grids in order to even draw 15ma as nearly 0V! Normally of course, that would have had the tubes red-plating.
Upon driving the amp with a 1K signal the output still was very low and had crazy cross-over distortion. The 1KHz drive to the grids was more than adequate but also had the same cross-over distortion. I attributed that to possibly being a result of the neg feedback.
Upon knowing the issue had to be at the power tubes I rechecked the voltages on the 6L6's.
Low and behold the screen voltage was way down at like 40V. What the hell I proclaimed out loud.
I found the screen resistors on the board & immediately found the issue,
He had installed 470K resistors instead of 470ohm resistors.
The reason I dragged out this repair story is just to share the experience as since I've only seen completely burned up screen resistors that I would have replaced immediately given just their charred appearence, I've never experienced how an output stage would respond if the screen resistors were just too high a value to be able to handle the current the screens draw at idle OR the reaction the wrong values created when put under load.
Anyway, there ya have it. A great new experience for me and a great lesson for a newbee...thanx for listening...glen

[nevetslab]

I assume the screen resistors were color coded....YEL-VIO-YEL instead of YEL-VIO-BRN, both with the tolerance band. Not sure if that's in the realm of having color blindness. I too am used to seeing the screen resistors charred, though many times, no discoloration....just measures open. Interesting results virtually shutting the screen potential down that low. Shame he had swapped out the O/T. Hope he was there with you for the step-wise discoveries. As we all have learned....transformers are usually the last thing to go, unless it's blowing fuses every time, pulls high current (such as shorted turns in either P/T or O/T), or no voltage at all (open Primary or open internal thermal switch). Good story.

[tubeswell]

Vg2 has as much effect on pentode or tetrode tube current as Vg1. If Vg2 is low enough, if can turn the tube off - even if your -Vg1 is 'normal' (But we know this when Rg2s go open. Right?)

[Jazz P Bass]

The screen on a 6L6 or EL34 is more like the plate.

[Helmholtz]

The screen does both: As a grid it controls the plate current but it also acts like the plate of a triode.
Once the electrons have passed the control grid (g1) they mainly see/get accelerated by the screen (not the plate) potential as the plate is in the "potential shadow" of the screen.

[Leo_Gnardo]

Vg2 has as much effect on pentode or tetrode tube current as Vg1. If Vg2 is low enough, if can turn the tube off - even if your -Vg1 is 'normal' (But we know this when Rg2s go open. Right?)

Indeed, some amps switch the screen grid voltage off as a standby mode. Sure is a lot less stress on the standby switch compared to switching all the B+ off and on.

He had installed 470K resistors instead of 470ohm resistors.

Another story that bears retelling. I had to "fix" the work of another local amp tech who either works in the dark or is color blind. Found 4K7 screen grid resistors instead of 470. Red may look much like brown to the color blind. Or... those who work in the dark. Anyone who's unsure, it never hurts to break out an ohmmeter and check to make sure.

[Enzo]

Or someone reached into his 470 ohm drawer and grabbed a resistor, and didn't check to discover someone had put 4k7 in that drawer by mistake.

[nevetslab]

While working at BGW Systems, being a manufacturer, you end up with resistors from numerous vendors. Color Coded 1% metal film, as well as the 5% carbon film, once parts have left the stock room, where the reels or bags of parts have been clearly marked, they get kitted up for the production line, now in plastic bags. The production ladies then divy up the parts for their individual work stations. I swear, trying to tell the difference between a BRN, RED or ORG color band, as different mfgrs use different pigments in their banding of the resistors......you really don't know without an ohmmeter. AND, oh yes, there were a few batches of PCB's that told that tale once assembled and passed thru the flow solder machine to eventually hit production test when assemblies weren't working right.

We even briefly had an Optical Comparator brought in for evaluation to check assembled PCB's with the Production Standard assembly. Still left you uncertain as to stripes being BRN, RED or ORG. knowing the color code/values for 1%, 2% & 5% helped answer that, but..............

[Enzo]

Or someone gets fooled looking at a four band resistor code, reading it like a 3 band.

[g1]

So I double check the plate, screen, control grid, & cathode voltages and aside from the really low idle current, I find nothing obvious.

Low and behold the screen voltage was way down at like 40V. What the hell I proclaimed out loud.

I think one of these statements must be incorrect? Or is it possible for a high value screen resistor to give a false 'normal' reading as you initially reported?
I've never seen it, but if it can happen I'd like to be aware of it?

[Enzo]

If you have 470k screen resistors, if the sockets are empty, the B+ on screen pin will read OK, but with tubers in place, the current on the screen will drop a ton of voltage.

[Helmholtz]

If there was any idle current the tubes must have been in.

[Dave H]

I think one of these statements must be incorrect?

That's what I thought too.
Perhaps an autoranging meter was used and 40V was read as 400V?

[gui_tarzan]

I did that on an early repair (my own amp). It's embarrassing when you look at it after you have some experience, the color on the resistors I used at the time was almost brown but they were in fact 470K.