• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Screen resistors-What value?

triode17

Super Member
I have been adding screen grid resistors to my amplifier designs as added insurance against tube failure. I know the common values are 100-470 Ohms (to 1.5k in git. amps). But how does one decide which value to use for a given power tube. Been using 470 for larger tubes, i.e. - 6550, EL34, 7591, but I am about to design a 6BQ5 type amp. Should I only use 100 there? Can someone explain this?
 
Register to hide this ad
In hi-fi amps, you want the minimum G2 stopper value that's consistent with freedom from oscillation under all conditions, particularly during power-up and power-down events. Higher values impair screen voltage regulation and may contribute to excessive distortion. Dave Gillespie has stated that 100R is appropriate in most cases, based on his own rather extensive investigation. I trust Dave will weigh in before long...
 
The point here is to ruin the circuit's Q so the parasitic tank (formed from stray inductance and the tube's Miller capacitance) oscillator is damped. Doesn't take much to ruin that circuit.

Stopper resistors also prevent arcing. A lesser-known, but very important, function.
 
In hi-fi amps, you want the minimum G2 stopper value that's consistent with freedom from oscillation under all conditions, particularly during power-up and power-down events. Higher values impair screen voltage regulation and may contribute to excessive distortion. Dave Gillespie has stated that 100R is appropriate in most cases, based on his own rather extensive investigation. I trust Dave will weigh in before long...

True.

In fact, one of the reasons you see high-value screen resistors in guitar amps, is precisely because of the distortion created. It can be an important part of the sound of a guitar amp, in breakup.

I usually use between 100 and 220 ohm screen resistors, in all my fixed-bias amp builds. As mentioned before, cathode bias amps don't need them normally- the cathode resistor actually acts as a damping influence, to enough of an extent to quell oscillation, in most cases.

I also tend to use input grid stoppers as well 1K is a good value for this.

I would also expect, that my practice of using 10 ohm cathode resistors in my fixed-bias amps, should also help slightly. They're primarily to allow easy bias adjustment- but they can actually slightly reduce distortion (see the old Genelex tube data sheets for an explanation of why), and reduce the tendency for oscillation as well, by adding damping...

Regards,
Gordon.
 
i typically use 100 ohms based on Dave G's suggestions. I have seen it work too. When I rebuilt my Pilot SA-260 (EL34/6CA7 UL, more or less similar to a Dynaco ST-70) the new tubes arced the first time I wound the volume up a little bit. I put the 100 ohm resistors on the screen taps and it has been absolutely fine since.
 
As mentioned before, cathode bias amps don't need them normally- the cathode resistor actually acts as a damping influence, to enough of an extent to quell oscillation, in most cases.
I don't build cathode biased amps but I wonder about this, because modern electrolytics, used for cathode bypassing, can have such low ESR that a tendency toward spurious oscillation is no longer attenuated.
 
Last edited:
I put them in all of my rebuilds, Cathode or Fixed BIAS. With the use of RUSSIAN OUTPUT TUBES it's CHEAP INSURANCE against arcing, as the Russian tubes aren't as robust in most cases, against screen arcing. Dave G has a good article on the TRONOLA SITE that addresses this, and why it happens. http://www.tronola.com/html/maximize_tube_life.html
 
Guys,

What about power ratings?

Screen current is very low: let's say 5 mA under static condition.
Under dynamic condition. let's say 3x that so 15 mA, so power of 22.5 mW. So a 1/2 watt would fit there no problem.

But I see some design with 2W screen resistor.

Any idea why?

Thanks.
 
i typically use 100 ohms based on Dave G's suggestions. I have seen it work too. When I rebuilt my Pilot SA-260 (EL34/6CA7 UL, more or less similar to a Dynaco ST-70) the new tubes arced the first time I wound the volume up a little bit. I put the 100 ohm resistors on the screen taps and it has been absolutely fine since.
ST-70s are particularly sensitive to this. The listed screen voltage is 415v, which is just 10 v below the maximum rating, at 117v in. At my nominal line of 122v, the screens are up at 435v, YIKES! I had two tubes of different brands red plate before I read Gillespie's article on Tronola. So, I added 470 Ohms to each screen grid, as close to the socket as possible, and I haven't had a red plate since. Another thing to do is to change your output tubes to KT77s. I use JJs and they work beautifully because they are designed for ultralinear use. The screen and plate ratings are both 800v!
 
The SA-260 calls for 425 at the plates and 428 at the screens, so even slightly more touchy than the ST-70 in that regard.

My 100 ohm are right at the socket, leads cut as short as I could get them. Its heat shrunk between the transformer lead and the socket pin.

I have EH 6CA7's in it currently, with a backup set of JJ EL34 and I think I own a set of EH EL34 too. When I manage to kill all of those, KT77's are on the list to try. I use that amp so infrequently that the tubes I have already will probably out last me.
 
Screen Grid resistors are used for a number of reasons, be it:

1. To eliminate garden variety (non-destructive) parasitic oscillations with tubes prone to such behavior,

2. To limit screen grid dissipation under conditions of high power output, and

3. To prevent output tube arcing in circuit/tube combinations prone to producing such behavior.

If the resistors are part of the original design and used/required only to address issues relating to #1 above, they are correctly called a "stopper" resistor, since the formation of parasitic oscillations is what they are stopping from happening. These resistors will typically be of a .5 watt rating, but can rated as high as 1 watt, with values as high as 470Ω depending on the tube they are serving.

Screen Grid resistors included in the original design to address the issues of #2 above will find values as high as 2.2K, with wattage ratings of 2 or even 5 watts not being uncommon. Values and ratings this high are almost exclusively reserved for high power true pentode tubes types like the EL34, where the screen grid is not aligned with the other grids in the tube, which allows screen grid current to soar under conditions of high power output and/or overload. In this case then, the resistors act to limit screen grid dissipation to protect the grid under those conditions, and therefore the tube as a whole. Resistors of this type can also act to address issues relating to #1 above as well. For applications like large sound distribution and guitar (production) amplifiers, oversized resistors were often used to favor ease of servicing. Back in the day when tubes were inexpensive and readily available, it was far easier and more economical to simply replace a tube than open a unit up to replace a screen grid resistor. Using a robust rated resistor then ensure that during a catastrophic event, it was the tube that burned up, and not the resistor. These types of resistors are often (and mistakenly) called stopper resistors as well, but are more correctly called screen dropping resistors.

Addressing item #3 above involves adding a screen grid resistor to a design where one was never originally included. Because they are added specifically to prevent any tendency towards output tube arcing, I coined the term "Screen Stability" resistor to identify their purpose, and the fact that they are not part of the original design.

Some designs certainly should have originally included them but didn't, with their resulting reputation of unexplained output tube arcing being near legendary. The Eico HF-89 is one such example. More common however, is the unit that was never particularly prone toward output tube arcing, but began arcing after the unit was recapped.

Many years ago now, I determined that such events were always a screen grid related event, with this element invariably arcing to either the cathode or control grid. The arcing began when modern (read that as: Low ESR) electrolytic caps were used to replace the original specification capacitors in the power supplies of favorite amplifier designs. The low ESR value of these caps, in conjunction with the screen grid itself, allowed a potent tank circuit to be created, that could be triggered to destructive levels anytime an errant transient came along to do so. Simply setting the needle down on a record was often all it took. An amplifier that was equally famous for output tube arcing after it had been restored was the Dynaco MK III. Adding Screen Stability resistors then to either the Eico or Dynaco amplifiers (and others like them) eliminates the tendency towards arcing.

Unless an amplifier is grossly abused (full or overload power developed with no load connected), virtually all arc events will occur at the screen grid of a tube. A plate arc can generally only occur when the back EMF from the collapsing field of an unloaded output transformer rares back from a full power signal applied to it. With no place for the energy to go, the plate receives a sucker punch of many thousands of volts in this scenario, giving way to an arc at the plate. In high fidelity applications however, this rarely if ever occurs other than by accident.

When adding Screen Stability resistors, the smallest practical wattage size is recommended so that the resistor will also tend to act as a fusing agent as well. A rating of .25 watt generally works just fine for amplifiers up to and including a power output rating of 35 watts when that power is developed from two output tubes. A .5 watt rating is usually sufficient for power levels above 35 watts developed from two tubes. Metal film resistors are highly recommended in this application, since they tend to open most quickly in case an arc event does occur. For the purposes of acting to prevent screen grid arcs, a value of 100Ω works well for virtually any tube type. Finally, and rather obviously, any resistance added to the screen grid circuit in an original design to address issues relating to #1 and #2 above, will also suffice to address #3 as well.

Dave
 
Thanks for the screen stability description.
One question : which type of resistors will work best when used as fusing agent ?
 
One question : which type of resistors will work best when used as fusing agent ?

Metal film resistors are highly recommended in this application, since they tend to open most quickly in case an arc event does occur.

I use carbon film because I've got a bag of them, but I will tell you they make lousy fuses. You'd be amazed how much abuse a CF resistor will suffer before it opens. I've bench tested them on the power supply and had them glowing red for longer than I'd be comfortable with if I needed a fuse element.
 
Hi Dave,

Thank you for your explanation on the subject: power ratings consideration were always omitted and you now closed it.
I didn't know some tubes could generate such high current to the screen that they need a 7W resistor to slow it down.

For us hobbyist, 100 ohm 1/4W and 1/2W based on the power rating of 35W output power for 2 tubes is the rule.

Thanks again,
Brice.
 
Thank Dave, this is exactly what I needed to know. I think I will stick with the 470 ohm for the EL34 and 7591 amps, and replace the 470s in the smaller amps with MF 100 Ohms.
 
Being a class A single ended amp, does that affect the value any? Since the screen current shouldn't swing like it would in an AB amp, it may not matter as much there.
 
Being a class A single ended amp, does that affect the value any? Since the screen current shouldn't swing like it would in an AB amp, it may not matter as much there.
No, different device. My SE Class A design uses a 6550 rigged as a triode by tying the screen to the plate through 100 Ohms.
 
Back
Top Bottom