Significance of Screen Grids?

[Justin Thomas]

Hi everybody,


I was just wondering what is the significance of the Screen Grids in a guitar amp, and why they are run at various voltages in a particular amp?
For example:

79 Champ, with between 405V-420 on the plates, approx. same on the screens, depending on wall voltage. A tech had replaced the first 1K dropping resistor w/a 10K, so this particular amp has about 355V on the screen.

Also, the cathode bias resistor was a 1K. The cathode resistor is to keep the plate dissipation down to avoid meltdown, correct? Higher value--less current---less Watts?

So what effect would the screen voltages have? I tried the screens with 355V and 405V and it didn't seem to make too much difference in anything? Is this anything to worry about? And the amp both ways was putting out between 13-15W, never redplating or otherwise cooking.

I'd like to clarify this BEFORE trying a 7591 in it...I've got lots of those...

Thanks advance!
Justin

[markbul]

Hi everybody,


I was just wondering what is the significance of the Screen Grids in a guitar amp, and why they are run at various voltages in a particular amp?
For example:

79 Champ, with between 405V-420 on the plates, approx. same on the screens, depending on wall voltage. A tech had replaced the first 1K dropping resistor w/a 10K, so this particular amp has about 355V on the screen.

Also, the cathode bias resistor was a 1K. The cathode resistor is to keep the plate dissipation down to avoid meltdown, correct? Higher value--less current---less Watts?

So what effect would the screen voltages have? I tried the screens with 355V and 405V and it didn't seem to make too much difference in anything? Is this anything to worry about? And the amp both ways was putting out between 13-15W, never redplating or otherwise cooking.

I'd like to clarify this BEFORE trying a 7591 in it...I've got lots of those...

Thanks advance!
Justin

Read this description of tube design and function:

http://www.vacuumtubes.net/How_Vacuum_Tubes_Work.htm



Screen voltage is very important to tube function. Screen voltage affects plate current, so that's the heart of your tube operation. Too much screen voltage can cause arcing in the tube, and can also cause the screen to exceed its dissipation limit. If the screen gets too hot, it will be damaged and the tube will fail. Some old designs ran screens too hot and cut down tube life. Within safe operating limits, screen voltage will also affect output power and tone. For most of us, it's probably best to run the screens at a safe level and manipulate other aspects of the entire circuit to taste.

[R.G.]

The screen grid is effectively a second input, but a much lower gain one, as befits its much more distant placement from the cathode's source of electrons.

It's usually used as an overall gain control. The screen grid can control how much current flows per unit of voltage change on the signal grid. The variation in voltages on the screens reflects the varied conditions that pentodes are set up for in various circuits. Often the screens are tied to nearl the plate voltage for as much gain as one can get out of the tube.

[danf]

6V6 screen grid resistor

Justin

I experimented with screen grid resistance on a single-ended 6V6 amp and found I like the sound no screen grid resistor better. I used a 5W potentiometer to dial in the resistance, so it was an easy comparison. Although I liked zero resistance on the screen grid, I ended up putting in a 100 ohm resistor - the sound difference between 0 ohm and 100 ohm was only slight, and it offered some protection. At 470 ohms and higher, I wasn't liking the sound nearly as much. The consensus on the amp discussion board at the time I was experimenting was that, as long as the 6V6's screen grid dissappation was within maximum (it was) all was well.

Dan

[markbul]

Dan

You checked the screen dissipation at idle? What will it be at full power?

[Bruce / Mission Amps]

What protection do you suppose the screen grid resistor is affording your tube?

[danf]

Screen grid resistor experiments

I dug up the old discussion posts, from when I was experimenting with screen grid resistors (this goes back almost 5 years – apologies for reposting it?).

Bruce – I put a 100 ohm resistor on the screen. I was happy with 100 ohms, 10 ohms or with no resistor, but felt just a little better with the 100 ohm resistor. The words “meltdown” and “screen grid glow” were mentioned in the discussions when I first posted this question. The 100 ohm resistor would help keep the screen dissipation down a bit, in case of sudden current increase.

Markbul / Justin - On the amp I was building (a single-ended 6V6), the screen grid voltage was 8 volts higher than the plate. Voltage readings to ground were 307vdc plate, 315vdc screen, and 18.5vdc cathode. I was concerned so I dropped the screen voltage using a 10k resistor. It sounded terrible, prompting me to experiment with a 5W potentiometer to “dial-in” various screen grid resistances. Here’s what I found (this is from the old post):

----------------------------
Screen grid resistor:

1) 10k ohm: sounded terrible.
2) 1k ohm: not nice at all, artificial sounding, harsh distortion.
3) 470 ohm: some improvement, but still harsh and ugly sounding.
4) 220 ohm: starting to sound better, good clean tone with a smoother transition to better-sounding distortion.
5) 100 ohm: nicer still, guitar and amp sounding great, good clean sound, really nice distortion, still has some amount of compression.
6) 10 ohm: loving it
7) no screen resistor (like the 5F2 schematic): loving it

Screen current and dissipation: From what I could tell by measuring voltage drop with a digital volt meter across the resistance, at max guitar and amp volume the screen dissipation peaked, during instantaneous hard chord whacks, at under 1.5 watts, using any of the screen resistors 1k and lower. Normal playing was much lower. This seems well within the 2.2 watts max screen dissipation in the 6V6 specs.
----------------------------

As mentioned, I ended up disregarding the fact that the screen voltage was higher than the plate voltage (the replies I received convinced me it was less important) and just made sure the screen’s dissipation was within spec. No problems have ever come with the amp.

Dan

[Shea]

Funny, my experience with my last amp leads me to virtually the opposite conclusions. It runs 6V6s. Rather than bore you with a long story, here are the essentials:

470R screen resistors, screen voltage < plate voltage: sounded good.

470R screen resistors, screen voltage > plate voltage: sounded bad - got weird, buzzy distortion.

1K screen resistors, screen voltage < plate voltage: sounds good.

Maybe there are other factors at work, but it seems that keeping the screen voltage lower than the plate voltage was more important than using the lowest possible value of screen resistors. I wonder if 470R or 1k screen resistors would sound better in your amp if you had a separate power supply node for the screens that keeps the screen voltage below the plate voltage.

Shea

[danf]

Shea

Did you also try screen grid resistor values lower then 470 ohms? In my amp, I was thinking it was the effect of the value of the resistor, more than the voltage differences between the plate and the screen. (If I remember correctly, in my amp, for all values of the screen grid resistor, 0 through 1k, the screen voltage was higher than the plate voltage. I'll double-check that.) I'll look at my notes from that test and report back.

Dan

[Bruce / Mission Amps]

Shea

Did you also try screen grid resistor values lower then 470 ohms? In my amp, I was thinking it was the effect of the value of the resistor, more than the voltage differences between the plate and the screen. (If I remember correctly, in my amp, for all values of the screen grid resistor, 0 through 1k, the screen voltage was higher than the plate voltage. I'll double-check that.) I'll look at my notes from that test and report back.

Dan

Dig it back out and measure the DC voltage across (use two clip leads) the 100 ohm screen resistor when the tube is idling and then while you are strumming away.
Keep in mind that sometimes the actual supply voltage for the screen is after the OT B+ node.
If the OT is small and a large DC resistance, can be more of a drop then a 470 to 1K screen resistor.

[Shea]

Shea

Did you also try screen grid resistor values lower then 470 ohms?

Nope, I never did. The screen & plate voltages are pretty high for 6V6s, and in fact one of the 6V6s shorted out when I was still running 470 ohm screen resistors. So I never would have considered going lower.

When I was originally running 470 ohm resistors and it sounded good, the plate voltage was running over 440, sometimes over 450 (the voltage coming out of my outlets fluctuates), and the screens were a couple volts lower than the plate voltage. That's when a tube shorted.

I lowered the plate voltage down to the 420 - 430 range. With the 470 screen resistors in there, the screen voltage was now higher than the plate voltage. So, the screen current must have been so much lower than before that the voltage drop across the screen resistors, plus the drop across the choke, was less than the drop across the primaries of the OT. That's when it sounded weird and buzzy. So I put in 1k screen resistors, which lowered the screen voltage below the plate voltage, and now it sounds dandy again.

Sometimes you can't really extrapolate what happens in one amp to a totally different amp, but for whatever it's worth, that was my experience.

Shea