should power tubes glow orange while playing the amp?

[bolero]

wait a minute....I know this sounds like a stupid question with an obvious answer:



but as far as I know my amps are biased properly ( by an amp tech ) and the tubes don't glow at all when the amp is idling



however, when I crank the amp up and start playing, the top of the ( plate? --> the metal structure inside the tube ) starts to glow orange in the centre, and then spreads to the whole top surface of the (plate?). if I stop playing, the glow subsides...but will come back as I play aggressively.


the sides of the metal structures don't glow at all, just the tops.


both amps are 2x6V6 powered, that I normally love to play at full volume....and the tubes are old RCA, so I don't want to destroy them prematurely if I can avoid it.


thx!

 

[John Phillips]

If the plates themselves aren't glowing, what you're probably seeing is glow from the screen grids which are inside the plate structure. The top part of the structure is mica, not metal, and is semi-transparent, so you see light from inside the tube.

When the tubes are driven hard, the plate voltage falls significantly, but the screen doesn't, so it can become more 'attractive' to the electron beam than it should, and draw excess current, which makes it glow.

This is NOT GOOD, as you've probably already guessed - screen-grid failure is actually the most common cause of blown 6V6s when they're operated beyond their designed maximum ratings (which they are in many guitar amps).

It doesn't mean the amp is biased too hot, necessarily (although it could, since the greater the tube current, the lower the plate voltage relative to the screen) - but it means you probably need the value of the screen resistors increasing. The 470 ohms used by Fender and others is not usually enough - 1K or even higher is a good idea. You may need to experiment to find the right value.


BTW, I know that this is a controversial view, but IMO many (if not most) techs and players bias amps too hot anyway.

 

[arfy]

I have a ho-made tweed Princeton and I eventually wound up using a 11k screen grid resistor in order to get the screen voltage lower than the plate at idle. I've been wondering if I'm doing the right thing or what. Some tweed champs had no choke but a 10k resistor between where the plate connects to the power supply and where the screen is connected. There is a choke in my Princeton clone between the first and second filter caps, and both the plate and screen are connected at the second filter cap, I'm thinking of moving the plate connection before the choke. Will the choke drop the screen voltage some in comparison to what the plate is seeing if I do that?
I wound up using a cathode resistor almost twice as big as the stock 470 as well to get the idle power down to about 12 watts, with the 470 ohm resistor the idle current was almost 50ma, that seems like a lot, at 360 volts on the plate, isn't that 18 watts static dissipation? The GE 6V6 wasn't glowing at the plate or anything but that just seems like too much.
The amp's loud enough to play in a basement situation with bass and drums, I'd guess the Princeton output transformer puts out a lot more power than the Champ one.
Any criticism or suggestions would be appreciated.
thanks to all at this forum who teach us all a lot!

 

[John Phillips]

arfy, that all sounds fairly normal.

Low-powered single-ended amps have a big DC resistance in the OT primary, so there's a large voltage drop across it even assuming the idle current is in the right range. That means you need a much bigger screen resistor value to keep the screen voltage down below the plate (although 11K still sounds surprising).

Yes, the last Tweed Champs and Princetons use a resistor not a choke in the B+ chain, which helps - and moving the plate feed upstream of the choke will too (and increase power a little).

Does your current reading include the screen current? If so, you can probably knock off 5mA or so, but 50mA is still a bit too much. Max dissipation on a 6V6GTA is 14W (only 12 on a 6V6GT). Remember also to subtract the cathode voltage from the plate, since it's the tube voltage, not the plate-to-ground, which matters for calculating tube dissipation. Even taking all that into account and assuming a cathode voltage of 20V, your dissipation is still roughly (360-20V) x (50-5mA) = 15.3W. Close to OK, and not surprising the plate wasn't visibly glowing.

You probably didn't need to go as far as doubling the 470-ohm cathode resistor value - but you usually do for a later BF or SF Champ circuit, with 420V B+.

 

[bolero]

thx for the help John!


one of these days ( pretty soon I think ) I'll learn how to bias & adjust my own amps


fwiw the plate voltage on the 6V6's was 450v, not sure of the mA though

 

[John Phillips]

20mA would be about right.

Max. dissipation is 14W, and a Class AB amp needs to be biased to idle at no more than 70% of this, which is 9.8W. At 450V, that's achieved with a current of just under 22mA. That's the absolute maximum, a little less is a good idea.

You could actually go quite a bit less than this without affecting the performance. IMO.

 

[arfy]

For your help with my tweed Princeton project, I took your advice and moved the plate connection before the choke, this raised the voltage a little to about 380, and knocked the plate current down to 27ma, so I was able to sub in a smaller cathode resistor of 620 ohms, that knocked the voltage down to about 370, 370 minus the 27 volts I measured at the cathode gave 343 volts times .037 current = 12.69 watts, plenty for a 6V6 but not as high as most similar designs seem to go, it's got more clean volume and punch now, lots of changes from what started out as the single ended amp project from the Angela Instruments site. I'm getting to be an old guy who doesn't like too much volume but wants to turn up and this thing is really getting there. I put a pot in place of the resistor that goes to ground after the coupling cap from the second preamp stage and that makes a nice master volume which doesn't seem to mess with the tone.
Now if I can just make about my mind about what value cathode caps to add to the preamp...can the tweeking actually have an end?
Thanks again for sharing your knowledge.

 

[arfy]

I was checking over a friend's Deluxe Reverb that had been cutting out(a wire in the speaker from the voice coil to the connector thingy had frayed, old Utah speaker), and when checking the bias I also measured that the screen voltage was higher than the plate voltage when the amp was at rest, I'd guess the difference would get bigger when the amp is played and the plate voltage sags, so I changed the 470 screen resistor to bigger values until there was about a 5-6 volt lower voltage on the screen.
Am I being anal about this or is this something that Deluxes in general could use? The resistor value I evetually went with was about 10k, so the 470 wasn't doing much of anything.
Any input, suggestions, or criticism?

 

[John Phillips]

When amps like the DR were designed, tubes were tough and running excess screen current didn't seem to matter much. I don't think a lot of Fender's designs were really accurately thought-through in some ways either - no insult to the guys involved, they just had 'ways of doing things' that produced good-sounding and apparently reliable results.

But fast-forward 40 years and fit modern-production tubes and you can be in trouble. I used to think that tubes like the EH6V6 were simply poor quality, but after reading a lot of what some of the tube resellers say about the failures (which are mostly screen-related), I think in fact it's just that the amps push the tubes too hard, and that the old tubes would take it - they weren't really meant to, but they did.

So it may be a good idea to change some values (especially screen resistors) to make the amp less hard on the tubes.

Do you notice any tonal or volume change with the 10Ks?

 

[arfy]

I think it's a bit cleaner when pushed hard, less slop to the distortion the creeps in as it's turned up. I'll have more of an idea of how the Deluxe sound once some other issues with the baffle and speaker are fixed.
Getting the screen voltage down definitely gave my ho-made more volume and headroom, with a P12Q it's almost gig-level loud, surprises the heck out of me for an amp with a single 6V6 power section.