Re grid stopper. I use 1k on both preamp and power tubes.
The value is not critical and you probably wouldn't hear any difference if you didn't fit them. My Chinese amp didn't have them but I am rebuilding it and will fit 1k..
More here:
https://www.aikenamps.com/index.php/grid-resistors-why-are-they-used
The only RF in doors we seem to have these days is 2.4G (wifi & Bluetooth) and 5G wifi which is too high to be a concern.
Not much VHF these days, but cell phones use from 800Mhz to a few Ghz so if I lived near a cell tower, or AM transmitter maybe it might be an issue.
Another excellent website:
http://www.valvewizard.co.uk/gridstopper.html
The value is not critical and you probably wouldn't hear any difference if you didn't fit them. My Chinese amp didn't have them but I am rebuilding it and will fit 1k..
More here:
https://www.aikenamps.com/index.php/grid-resistors-why-are-they-used
The only RF in doors we seem to have these days is 2.4G (wifi & Bluetooth) and 5G wifi which is too high to be a concern.
Not much VHF these days, but cell phones use from 800Mhz to a few Ghz so if I lived near a cell tower, or AM transmitter maybe it might be an issue.
Another excellent website:
http://www.valvewizard.co.uk/gridstopper.html
The grid stopper gets its name because it stops parasitic oscillations (there are also cathode, grid2 and anode stoppers - same deal). Receiving valves work pretty well up into the FM and old TV frequencies, and if we pretend (too much) that only audio frequencies will be amplified we get butt-bit by surprises at VHF that we weren't considering. Audio sized parts and wiring include all kinds of extra Ls and Cs, small for audio, sometimes big enough to matter for VHF.
The idea is to add enough (positive) resistance to damp down whatever demons that might lurk. It's a brute force approach, but that's all that's (usually) needed. Old fashioned bulk resistors are often recommended because they (carbon comps, at least) tend to still be pretty resistive at VHF. If you delve into this, you'll hear the term "negative resistance", which is a cool description of the issue.
All good fortune,
Chris
The idea is to add enough (positive) resistance to damp down whatever demons that might lurk. It's a brute force approach, but that's all that's (usually) needed. Old fashioned bulk resistors are often recommended because they (carbon comps, at least) tend to still be pretty resistive at VHF. If you delve into this, you'll hear the term "negative resistance", which is a cool description of the issue.
All good fortune,
Chris
Last edited:
I live within just a few miles of four TV and radio towers that are 1,000+ feet tall (very highly visible!) and several cell phone towers within less than a mile, and practically everyone has a wireless router now including me. I always thought that as far as input is concerned, RCA shielding and short runs, and proper grounding, would take care of it.
Since 6SL7 is "high mu" and given my VERY high RFI location, it seems that a grid stopper there of 330 Ohms or so would be a reasonable addition, although short well-shielded RCA cables and shielded input wiring internally plus a grounded chassis in theory ought to take care of the RFI.
I did add four 1k grid stoppers to the four output tubes of my new ST-35, after I built it, right at the EL84 pins, and it did seem to take away a little harshness. The ST-35 schematic doesn't call for them. I added them after-the-fact with a slight subjective improvement in sound quality.
I am a big advocate of wall outlet power filtration. I have been for decades. It doesn't have to be absurdly expensive. I use Furman AC line power filters on everything, with one for video components and another one for audio components. A pair of these works great for me, although they have gone WAY up in price since I bought my two. Each has two separate sections. They also have the added benefit of shutting everything off instantly if there is a voltage spike, which has happened several times over the years.
https://www.guitarcenter.com/Furman/PL-8C-Power-Conditioner.gc
Since 6SL7 is "high mu" and given my VERY high RFI location, it seems that a grid stopper there of 330 Ohms or so would be a reasonable addition, although short well-shielded RCA cables and shielded input wiring internally plus a grounded chassis in theory ought to take care of the RFI.
I did add four 1k grid stoppers to the four output tubes of my new ST-35, after I built it, right at the EL84 pins, and it did seem to take away a little harshness. The ST-35 schematic doesn't call for them. I added them after-the-fact with a slight subjective improvement in sound quality.
I am a big advocate of wall outlet power filtration. I have been for decades. It doesn't have to be absurdly expensive. I use Furman AC line power filters on everything, with one for video components and another one for audio components. A pair of these works great for me, although they have gone WAY up in price since I bought my two. Each has two separate sections. They also have the added benefit of shutting everything off instantly if there is a voltage spike, which has happened several times over the years.
https://www.guitarcenter.com/Furman/PL-8C-Power-Conditioner.gc
Last edited:
Audible sound improvement maybe due to the stoppers preventing parasitic oscillations.
It may be worth adding an anode stopper of about 47r to "discourage leakage inductance oscillation in the OPT" (ref: the Valve Wizard.co.uk)
I'll also be adding 47r in the OPT UL connection.
It may be worth adding an anode stopper of about 47r to "discourage leakage inductance oscillation in the OPT" (ref: the Valve Wizard.co.uk)
I'll also be adding 47r in the OPT UL connection.
Hiss from resistors is called Johnson Noise. it's a thing. Grid stoppers are also necessary. There's a balance somewhere in there. I've found that grid stoppers on an input stage were not necessary, while grid stoppers on an output stage were required to quiet oscillation, in the same circuit.
Stopper, stoppers everywhere! I hadn't given much thought to stoppers until this project came along.
Before I gutted and rebuilt my Dynaco Stereo 70 series ii in the 1990's, it had a problem that turned out to be common to all owners. The left rear output tube would run away, glow red, and blow out several resistors on the PCB every few months. Not good. It was always the left rear tube. Every time. My solution was to gut it and rebuild it with the same circuit laid out differently and wired point-to-point. It solved the problem for me. A long time later, others finally decided that it was a problem with the PCB/layout and adding screen stoppers to the UL solved the problem. So I already have had some bad experiences here. However, I will be using a regulated screen supply on this amp (post #115), and I'm not sure it's necessary to add an additional stopper at the output tube's screen pin, although I certainly could. It's the same regulated power supply as the Huaji Audio, which does not use another stopper resistor.
Given the RFI discussion in post #123 I think I'll play it safe and add grid stoppers on the input tube. Some sort of carbon I guess since consensus seems to be that this isn't a good place for metal film due to the possibility of inductance, or so I have read. Suggestions from Mouser or DigiKey or Antique Electronic Supply anyone? I'll be ordering from all three for my current projects. Adding resistors here can't hurt much, and without test equipment (at least for now), I'd rather be safe than sorry. I can always bypass them with a short wire later to see if it makes any difference.
All of my other amps have grid stoppers on the output tubes, so I am adding those to my schematic. I used Dale metal film 1k Ohm on my ST-35 and my Huaji Audio, and they seem to work fine. Wondering if carbon would have made a difference. Only one way to find out, and I don't feel like changing them now.
I haven't ever heard of this. Anyone else know about this resistor?
Before I gutted and rebuilt my Dynaco Stereo 70 series ii in the 1990's, it had a problem that turned out to be common to all owners. The left rear output tube would run away, glow red, and blow out several resistors on the PCB every few months. Not good. It was always the left rear tube. Every time. My solution was to gut it and rebuild it with the same circuit laid out differently and wired point-to-point. It solved the problem for me. A long time later, others finally decided that it was a problem with the PCB/layout and adding screen stoppers to the UL solved the problem. So I already have had some bad experiences here. However, I will be using a regulated screen supply on this amp (post #115), and I'm not sure it's necessary to add an additional stopper at the output tube's screen pin, although I certainly could. It's the same regulated power supply as the Huaji Audio, which does not use another stopper resistor.
Given the RFI discussion in post #123 I think I'll play it safe and add grid stoppers on the input tube. Some sort of carbon I guess since consensus seems to be that this isn't a good place for metal film due to the possibility of inductance, or so I have read. Suggestions from Mouser or DigiKey or Antique Electronic Supply anyone? I'll be ordering from all three for my current projects. Adding resistors here can't hurt much, and without test equipment (at least for now), I'd rather be safe than sorry. I can always bypass them with a short wire later to see if it makes any difference.
All of my other amps have grid stoppers on the output tubes, so I am adding those to my schematic. I used Dale metal film 1k Ohm on my ST-35 and my Huaji Audio, and they seem to work fine. Wondering if carbon would have made a difference. Only one way to find out, and I don't feel like changing them now.
I haven't ever heard of this. Anyone else know about this resistor?
Silly me. It's right on my own schematic with a 100R for triode operation. Not sure why I haven't seen it before for pentode operation. Does triode operation have some problem here that pentode operation does not?
Your schematic's 100r is not an anode stopper as it's not in the anode to OPT connection, its a screen grid stopper.
I was thinking of fitting a switch as you have but will put a 47r from pin 4 of the valve base to the switch common and then anode and UL tap to the switch but I can't find a switch suitable for such high DC voltages. (is as per The Valve Wizards schematic above)
I am also fitting a hum balancing 500r pot between the heater connections, centre tap to 0v. You can use two, 220r resistors.
I don't think the type of stopper resistors matters on the output tubes because the drive is volts. Whereas in a gain stage input tube, the Johnson noise is more relevant.
I wondered whether a wirewound 470-1k grid stopper would produce least noise.
On the subject of triode v UL, my research has produced a few facts:-
Triode - about 50% less power & gain, more distortion (different harmonic structure), but lower output Z
I suspect the Chinese like UL amps because they can use the higher power and lowish distortion as selling points.
I was thinking of fitting a switch as you have but will put a 47r from pin 4 of the valve base to the switch common and then anode and UL tap to the switch but I can't find a switch suitable for such high DC voltages. (is as per The Valve Wizards schematic above)
I am also fitting a hum balancing 500r pot between the heater connections, centre tap to 0v. You can use two, 220r resistors.
I don't think the type of stopper resistors matters on the output tubes because the drive is volts. Whereas in a gain stage input tube, the Johnson noise is more relevant.
I wondered whether a wirewound 470-1k grid stopper would produce least noise.
On the subject of triode v UL, my research has produced a few facts:-
Triode - about 50% less power & gain, more distortion (different harmonic structure), but lower output Z
I suspect the Chinese like UL amps because they can use the higher power and lowish distortion as selling points.
Last edited:
Johnson/Nyquist/thermal noise depends only on the resistance and is independent of resistor type. "Excess" noises all vary with current; no current into the grid, so no excess noise, so all types have the same contribution. In a "power amplifier" signal levels are already so large that noise contribution from a grid stop is not significant. Don't worry, be happy.
Good design is robust, forgiving, repeatable, reproducible, and device spread tolerant. Some of this comes from conservative Best Practices, but more comes from a hard-edged eye at assumptions.
All good fortune,
Chris
Good design is robust, forgiving, repeatable, reproducible, and device spread tolerant. Some of this comes from conservative Best Practices, but more comes from a hard-edged eye at assumptions.
All good fortune,
Chris
Thanks. My intuition about the electrons having to fight their way through a carbon composition or metal film causing more noise than a few cms of wire was rubbish 🙁
There are excess noises caused by voodoo within resistors, but they're all functions of current, so don't apply to grid stops. And at these signal levels, thermal levels of noise don't really matter. For a phono input stage, sure, if you're obsessive. But with signal levels >100 times larger, we instead worry about hum, ground path corruption, ground loops, etc. - eyes on the prize.
All good fortune,
Chris
All good fortune,
Chris
Novice mistake. I see. If I do put in the switch, I end up with both 100R screen stopper and 47R anode stopper in triode mode, but just the 47R anode stopper in pentode mode. Stoppers, stoppers everywhere! 😉
Seems like inductance in a metal film grid stopper resistor might fall into this category?
In the schematic in #132 the G2 stopper should be placed right at the socket, dangling immediately off pin4. Some of the need for stoppers comes from the wiring around valve sockets. At audio frequencies it all seems pretty innocuous, but receiving sized valves don't get the memo that we're only interested in audio frequencies, and may try go off reservation at much higher frequencies. It's then up to more sober elements to try to reign them in.
Inductance here is surprisingly not always a negative. (A resistively damped small inductor is a classic anode stopper in Ham radio transmitters). I think, and I'm treading beyond my limited range of security, that simply being resistive at the VHF at issue is all that's required. Not too much capacitive shunt and not resonant with surrounding reactances.
All good fortune,
Chris
Inductance here is surprisingly not always a negative. (A resistively damped small inductor is a classic anode stopper in Ham radio transmitters). I think, and I'm treading beyond my limited range of security, that simply being resistive at the VHF at issue is all that's required. Not too much capacitive shunt and not resonant with surrounding reactances.
All good fortune,
Chris
The 100r resistor goes to the anode - terminal 3 of the valve so is not in series with the 47r. (probably not make any audible difference anyway)
Just to add to the confusion of us all, on some schematics I have seen, the g3 resistor goes to the HT = B+ and not to the anode.
Don't forget the 500r pot between pins 2 and 7. You might as well add it to minimise any hum unless you are using DC for the heaters.
There is not one on g2 at present. Are you saying that I might want to add one? Do you have a favorite value? As with all of these various stoppers the values vary.
Hypothetically, how do these rank these in relative importance? Most schematics omit one or more.
input/driver tube grid stopper
output tube screen stopper
output tube grid stopper
output tube anode stopper
It seems that the most commonly included one is the output tube grid stopper, but I have read that certain tubes and designs often don't need one here.
Then I read that high-mu drivers often need the input/driver tube grid stopper. This one made sense to me in post #123 regarding high RFI areas.
The next least common one seems to be the output tube screen stopper.
The least common one seems to be the output tube anode stopper. I don't recall ever seeing one until yesterday.
We can't believe everything we read online, and it's very confusing for a novice. Then there are the original schematics for these amps, most of which don't use any stoppers at all as seen below. Yes, they are doing everything possible to reduce parts count, but they do seem to work at least. Thus my question how do these various stoppers rank in relative importance?
Last edited:
I make my life easy. I always use grid stoppers, 1k for g1, typically 100ohm for g2 (output tubes), which also can act as g2 current limiter, and almost never anode stoppers, only once when using GU-29 tubes.
I had the 100R screen stopper on my schematic, but only for triode mode for some reason.
So all I have to do is move it like this if I got it all correct. I might skip the triode-pentode switch, but it seems like an interesting option. They need different amounts of GNFB though, right? Am I going to have to switch the GNFB at the same time?
So all I have to do is move it like this if I got it all correct. I might skip the triode-pentode switch, but it seems like an interesting option. They need different amounts of GNFB though, right? Am I going to have to switch the GNFB at the same time?
More than I ever wanted to know about pentodes and screens, but a good read. I can't understand all of it, but I did get some of it:
SCREEN GRIDS in AUDIO and RF MODULATOR POWER TUBES
and here:
https://www.diyaudio.com/community/threads/screen-grid-resistors.71665/
The 2SC3150 in this regulated supply should be able to handle two EL34 screens, right? It comes from an amp where it supplies four EL84 tubes. It has a decent heatsink on it but not a huge one.
2SC3150 datasheet: https://rocelec.widen.net/view/pdf/vs96ddf8qi/ONSMS23400-1.pdf?t.download=true&u=5oefqw
SCREEN GRIDS in AUDIO and RF MODULATOR POWER TUBES
and here:
https://www.diyaudio.com/community/threads/screen-grid-resistors.71665/
The 2SC3150 in this regulated supply should be able to handle two EL34 screens, right? It comes from an amp where it supplies four EL84 tubes. It has a decent heatsink on it but not a huge one.
2SC3150 datasheet: https://rocelec.widen.net/view/pdf/vs96ddf8qi/ONSMS23400-1.pdf?t.download=true&u=5oefqw
Last edited:
Seems like the two screens only draw on the order of milliamps, so fine for the 2SC3150 with heatsink I suppose.
The regulated screen supply may be overkill, but many people do recommend it, and it's easy to implement with the little PCB shown above. It works great in my other pentode amp. The PCB should be here in a week or so if I decide to use it.
Some people just take the screens off the same power supply R-C as the driver tubes and skip the regulation. That's what is done on the original Chinese schematic. A much simpler approach. Some amps have a separate transformer winding for the screens with its own power supply. I don't have that. If anybody has any other simple screen regulation suggestions, please let me know.
A small separate transformer probably would fit, but I have no idea how many mA those two screens would need from a transformer on a continuous basis. I thought it was about 26 mA per tube maximum if I did the calculation right from an old Brimar EL34 tube spec sheet showing dissipation at 8W max for each screen. The most expensive option I guess, but very simple.
Last edited: