Will this work to reduce screen voltage?

[brewdude]

I have an amp with fixed bias 6K6GT power section.  I have individual 1K screen resistors and ~296v on the screens—the listed max voltage for screens 285v. The anode voltage is ~310v and I have calculated dissipation at just about 70% of listed max.


I have found that if I adjust the bias to be cooler, the screen voltage goes up. If I adjust the bias such that the screen is at the listed 285v limit, the anode dissipation is pushed up to far higher than the typical 70% of max.


I am reluctant to test the limits of these old tubes.


I was wondering if I could just add a resistor from each screen to ground to form a voltage divider with the 1K screen grid resistor that would drop 10-15 volts?

[PRR]

> ~296v on the screens—the listed max voltage for screens 285v

That's 4%. And 6K6 must be Design Center ratings? These give huge margin for variations in build and use for mass production.
https://frank.pocnet.net/sheets/049/6/6K6GT.pdf

If you actually measure every amp you build, you can use Design Maximum ratings. Which are typically 10%-15% up on Des Center rating. But 6K6 did not live into the age of Max ratings.

But note that 6K6 voltage ratings are *exactly* same-as 6V6. 315V 285V. And later 6V6 listed MAX numbers:
https://frank.pocnet.net/sheets/191/6/6V6.pdf
350V 315V

And we know from decades of Fenders that 6V6 gives decent life in the 350V-400V zone.
__________________________________
6K6 is a re-badged type '41. And strangely this '41 sheet shows the same Center numbers, with "absolute" numbers too. Essentially same-as late 6V6 sheets.
https://frank.pocnet.net/sheets/093/4/41.pdf

[sluckey]

Try using a 15V zener in series with the 1K resistor. Anode points to the tube. More info from Aiken...

     https://ampgarage.com/forum/viewtopic.php?p=171268#p171268

[brewdude]

> ~296v on the screens—the listed max voltage for screens 285v

That's 4%. And 6K6 must be Design Center ratings? These give huge margin for variations in build and use for mass production.
https://frank.pocnet.net/sheets/049/6/6K6GT.pdf

If you actually measure every amp you build, you can use Design Maximum ratings. Which are typically 10%-15% up on Des Center rating. But 6K6 did not live into the age of Max ratings.

But note that 6K6 voltage ratings are *exactly* same-as 6V6. 315V 285V. And later 6V6 listed MAX numbers:
https://frank.pocnet.net/sheets/191/6/6V6.pdf
350V 315V

And we know from decades of Fenders that 6V6 gives decent life in the 350V-400V zone.
__________________________________
6K6 is a re-badged type '41. And strangely this '41 sheet shows the same Center numbers, with "absolute" numbers too. Essentially same-as late 6V6 sheets.
https://frank.pocnet.net/sheets/093/4/41.pdf

Thanks PRR, What I believe I should take away from this is that I shouldn’t sweat the extra 10-15 volts on the screens.  I suspect that I should still try to keep the anode dissipation within the 70% of max.(?)

[pdf64]

I have an amp with fixed bias 6K6GT power section.  I have individual 1K screen resistors and ~296v on the screens—the listed max voltage for screens 285v. The anode voltage is ~310v and I have calculated dissipation at just about 70% of listed max.
…

My take is that idle voltage is somewhat inconsequential, valve info often shows that at 0 current, much higher voltage can be tolerated. Of course your idle current will be a lot higher than 0, but a sliding scale might be applied. Whatever, unless the screen grids are idling at max dissipation, I suggest to not be concerned about their idle voltage. Of far greater relevance is the screen grid voltage at high power output.

[brewdude]

Try using a 15V zener in series with the 1K resistor. Anode points to the tube. More info from Aiken...

     https://ampgarage.com/forum/viewtopic.php?p=171268#p171268

Thanks sluckey,
I will reread through some of that thread again. 

[brewdude]

So, I measured 297v on the screens grid and about 3v across the 1k individual screen grid resistor.  I believe that this means the screens are drawing 3mA at idle.


When I turn the amp up to 11 and hit a hard E chord, the voltage across the SGR goes up as far as 11.2v but the the screen voltage also goes down to below 270v.

[tubeswell]

When I turn the amp up to 11 and hit a hard E chord, the voltage across the SGR goes up as far as 11.2v but the the screen voltage also goes down to below 270v.

As the screens pull more current (for a given bias voltage), the screen voltage will sag. Similar to the relationship between plate current and plate voltage

[pdf64]

So, I measured 297v on the screens grid and about 3v across the 1k individual screen grid resistor.  I believe that this means the screens are drawing 3mA at idle.


When I turn the amp up to 11 and hit a hard E chord, the voltage across the SGR goes up as far as 11.2v but the the screen voltage also goes down to below 270v.

I think that indicates that your amp will not be pushing the 6K6 screen grid voltage limit, as current and hence dissipation is fairly low at idle.
However, the screen grid dissipation seems to be nearly 3W when overdriven, a bit over the 2.8W (design centre) limit. Whether that's a problem may depend on how you intend to use the amp. It may be worth considering that almost all valve guitar amps do the same thing (ie exceed screen grid dissipation limits when heavily overdriven).

[Colas LeGrippa]

5 K instead of 1 K  on screen supply

[kagliostro]

I remembered I had something on my archive and maybe someone can be interested

here it is





Franco

[pdf64]

5 K instead of 1 K  on screen supply

Have I missed the schematic somehow?
Certainly it would be beneficial to reveal it.

[Colas LeGrippa]

Haha no, you haven't missed the schem but you don' need it to figure it out...
On my Lil' Gripper I had put by error a 10k dropping resistor in series with the 1K screen resistor on the unique 6CA7. It worked as is for a year until I opened the amp for a reason I don't remember and I took it off. A 5K will drop the voltage to a certain point, certainly not by half though.


Colas

[Colas LeGrippa]

I can't remember wether the valve wizard or Rob Robinette suggested a 3.9k in series  with the screen supply  instead of the traditipnal 500R or 1K resistor in order to drop its supply voltage under the plate voltage

[brewdude]

I was contemplating adding a shared 2.2k resistor to feed the individual 1k’s, but I’m now thinking maybe I’m still in the acceptable high range of operational limits…
Maybe, I will give it a try anyway.

[brewdude]

Here is the schematic... I think this is the most current, most likely, accurate version.

[ac427v]

Are you using the 1 ohm resistors to set bias? They usually connect the cathode (not the plate) to the ground.

[pdf64]

Here is the schematic... I think this is the most current, most likely, accurate version.

The 1k 5W in series with the choke is what I was going to suggest trying, if you wanted to bring things down a bit more. As it is, you might consider increases its value, or the value of the 120ohm 5W rectifier current limiting resistors, a bit.

[brewdude]

Are you using the 1 ohm resistors to set bias? They usually connect the cathode (not the plate) to the ground.

Yes, I know they are usually placed on the cathodes, but I’ve been pretty happy with placing them on the plates. 


I’m willing to be talked out of the practice with a compelling argument, but I like the convenience of measuring the plate current without also including the screen current.  I was thinking about using 1 ohm resistors at both the screens and plates of my next build.


I presume that the reason for using the cathode to measure current is that with one side at ground it is somewhat safer and could be measured relative the ground just like the plate and screen voltages.  And, one can use external test points.

[brewdude]

Here is the schematic... I think this is the most current, most likely, accurate version.

The 1k 5W in series with the choke is what I was going to suggest trying, if you wanted to bring things down a bit more. As it is, you might consider increases its value, or the value of the 120ohm 5W rectifier current limiting resistors, a bit.

Thanks for the suggestions.  I do not wish to reduce the preamps voltages so I’ve been reluctant to increase the choke/resistor. 
It never occurred to me to increase the rectifier’s limiting resistor.  I think if that were to be effective it would reduce all the voltages.


I may try adding the a shared 2.2k to the individual 1K SGR’s.

[sluckey]

I was thinking about using 1 ohm resistors at both the screens and plates of my next build.

I would not bother with the screens if you already have a screen resistor. Just measure the voltage drop across the screen resistor and do the math.

[Willabe]

I’m willing to be talked out of the practice with a compelling argument, but I like the convenience of measuring the plate current without also including the screen current. 

I presume that the reason for using the cathode to measure current is that with one side at ground it is somewhat safer and could be measured relative the ground just like the plate and screen voltages.  And, one can use external test points.

Yes, 1ohm R's on the K so you don't have to mess with the high B+dcv. Much safer.

[HotBluePlates]

Are you using the 1 ohm resistors to set bias? They usually connect the cathode (not the plate) to the ground.

Yes, I know they are usually placed on the cathodes, but I’ve been pretty happy with placing them on the plates. 


I’m willing to be talked out of the practice with a compelling argument ...

The first time you slip and your meter probe bridges Pin 2 & Pin 3, you'll know why it was a bad idea.

If you only ever clip-on your meter leads before power-up a good practice), then perhaps you'll never have a problem.

[pdf64]

Bear in mind that hanging a meter off an output valve anode temporarily creates the worst lead dress it's possible to conceive of, and oscillation (which may be ultrasonic) is a distinct possibility. If that does happen, it will screw up the readings obtained. Perhaps to a non obvious degree.

[brewdude]

I only use clipped leads.
I have not experienced any issues wrt augmented terrible lead dress.

[mresistor]

Brewdude  that transformer resistance method is not hard

[brewdude]

Thanks for the diagram.  For now, I will leave this amp the way it is with the 1R resistors at the plates, but I will consider this for the next build.


I decided to add 2.2k shared screen grid resistance feeding the individual 1k SGR's. 
I set the bias such that the screen voltage was 285v. 
This yielded 18.6mA w/ 310v @ the plates,
which works out to 5.8W,
which is about 68% of the 8.5W maximum plate dissipation for the 6K6GT's.

[sluckey]

I'm not gonna try to talk you out of putting 1Ω resistors on the plates. Heck I've considered doing that also. But I will tell you why I put them on the cathode...

It is safer and I occasionally put test jacks on the chassis. Three jacks is all you need to measure the current and a RV6 locking pot to adjust. There is no B+ on the outside of the chassis. Here's a couple examples...

     http://sluckeyamps.com/sunn/mod_2.jpg
     http://sluckeyamps.com/november/nov_05.jpg

Yes, I'm aware that the cathode current also contains the relatively small screen current and that causes your meter to always read higher than the actual plate current. But I consider that a good thing. The net results is that your idle dissipation will always be a bit lower than your calculated dissipation. IOW, if you meticulously adjust your calculated dissipation to 70%, you may actually be running at 65%. Err on the safe side and that's a good thing IMO. I am not so interested in hitting the "magic" 70%. More important to me is that the tube currents are equal/matched.

That OT resistance/voltage method is accurate but, just like the 1Ω resistors on the plates, it's best left to the people that know what they are doing. A slip of the meter probes can quickly turn your OT into a boat anchor! One spark is sometimes all that's needed on that tiny primary wire.

[brewdude]

I have built a couple amps with the test points and locking pot accessible from the outside.  It works great and I would not use it across a 1R on the anode.  I do not build to sell my amps and if I did I would place the 1R’s on the cathode—for sure.


I don’t mind the discussion about placement of the bias sensing resistors…
But, I’m more interested in screen operating conditions and limitations and what’s the best way to keep the tube happy.


I appreciate all the help and advice, this is a great forum.  I think I have the amp operating within the published design center max conditions.  I don’t think I could even tell the difference with the added resistance.

[mresistor]

Sluckey - very true..