ampage Tube Amps / Music Electronics	For current discussions, please visit Music Electronics Forum.

ampage archive

Vintage threads from the first ten years

3 Watt screen resistors?

6/18/2000 9:23 AM
Steve Ahola	3 Watt screen resistors? Is there any problem using 3 watt metal film screen resistors? Those 5 watt "bathtub" resistors sure take up a lot of room... This is for an EL84 amp with a switchable B+ from 338vdc to 378vdc, but I was wondering about 6L6's as well with a B+ close to 500vdc. BTW certain boutique amp makers use carbon comp but only for their screen resistors... is there a reason for that? --Thanks! Steve Ahola

6/18/2000 9:23 AM

Steve Ahola

3 Watt screen resistors?

Is there any problem using 3 watt metal film screen resistors? Those 5 watt "bathtub" resistors sure take up a lot of room... This is for an EL84 amp with a switchable B+ from 338vdc to 378vdc, but I was wondering about 6L6's as well with a B+ close to 500vdc.

BTW certain boutique amp makers use carbon comp but only for their screen resistors... is there a reason for that?

--Thanks!

Steve Ahola

6/18/2000 9:29 AM
Stephen Conner	I've got an amp with 2 EL34s that run off 475V and it uses one 1K 5 watt screen resistor shared between both tubes. So I guess that would be the same as one 2.5 watt resistor per tube. The resistor doesn't seem to run too hot, so I don't think you'ld have a problem. Steve C.

6/20/2000 8:52 PM
MBSetzer	Carbon comps can sometimes take surges as well as overcurrent with less damage than other types of the same rating. Most likely you are looking at 3watt metal oxide which can take the HEAT better than the carbon (or especially metal film), maybe not an extreme voltage surge or continued overcurrent though . . . I used a 2watt 1K metal oxide on the screen of an individual 6BQ5 and it doesn't even break a sweat. Since it is soldered right to the tube socket with a fairly short lead, most of the heat comes conducted & radiated from the tube, but voltage drops across the resistor only add to that. It is easy to underestimate the wattage if you only consider the DC voltage drop at idle, DC must be checked at maximum volume, even then the AC voltage drop needs to be measured and added to the DC behavior to calculate theoretical heating watts. I have found it easier to just leave the slide rule in my pocket protector, and jam extensively at maximum volume before testing the resistor temperature using the human hand Incidentally, when it comes to heat, remember a 1watt 1K will dissipate the same calories as a 10watt 1K when operated at the same voltages & currents. The higher wattage units just handle it better and dissipate it without rising to as high a temperature themself. The amount ot heat released is exactly the same though, so when enclosing a resistor results in the chassis becoming too hot, replacing it with a higher wattage one may help the resistor survive but will not reduce the temperature reached by the chassis. I think you would be fine with either a single 3W on each screen, or just one 3W to isolate both from the screen supply, with the latter arrangement I might also add another 100R to each screen to isolate them from each other a bit. I used a 2W carbon comp on each tube in a 6L6 amp and it takes a while of jamming before they heat to a comfortably warm touch, I do turn the amp off and bleed the caps right before I touch them to avoid all the high voltage danger points. Hope this helps, Mike

6/20/2000 8:52 PM

MBSetzer

Carbon comps can sometimes take surges as well as overcurrent with less damage than other types of the same rating.

Most likely you are looking at 3watt metal oxide which can take the HEAT better than the carbon (or especially metal film), maybe not an extreme voltage surge or continued overcurrent though . . . I used a 2watt 1K metal oxide on the screen of an individual 6BQ5 and it doesn't even break a sweat. Since it is soldered right to the tube socket with a fairly short lead, most of the heat comes conducted & radiated from the tube, but voltage drops across the resistor only add to that. It is easy to underestimate the wattage if you only consider the DC voltage drop at idle, DC must be checked at maximum volume, even then the AC voltage drop needs to be measured and added to the DC behavior to calculate theoretical heating watts. I have found it easier to just leave the slide rule in my pocket protector, and jam extensively at maximum volume before testing the resistor temperature using the human hand

Incidentally, when it comes to heat, remember a 1watt 1K will dissipate the same calories as a 10watt 1K when operated at the same voltages & currents. The higher wattage units just handle it better and dissipate it without rising to as high a temperature themself. The amount ot heat released is exactly the same though, so when enclosing a resistor results in the chassis becoming too hot, replacing it with a higher wattage one may help the resistor survive but will not reduce the temperature reached by the chassis.

I think you would be fine with either a single 3W on each screen, or just one 3W to isolate both from the screen supply, with the latter arrangement I might also add another 100R to each screen to isolate them from each other a bit. I used a 2W carbon comp on each tube in a 6L6 amp and it takes a while of jamming before they heat to a comfortably warm touch, I do turn the amp off and bleed the caps right before I touch them to avoid all the high voltage danger points.

Hope this helps,
Mike

6/24/2000 6:54 PM
Steve Ahola	1k vs 1k5 vs 2k7 vs ??? for EL84's I've heard that changing the values of these screen grid resistors can fine-tune the tone and response of your output section. Any comments or suggestions on this? --Thanks! Steve Ahola

6/27/2000 9:03 PM
MBSetzer	Hi Steve: Just a brief comment . . . This is one of those things I found both listenable & measurable. I may be wacky sometimes but this is more accessible. Especially with EL84 you hear it easier. If you were only using 250VDC on both plate & screen, and their supplies are adequately decoupled by a choke or about a 5K or less power supply resistor, with good caps about 20mf on the plate & 10mf on the screen, then I would expect good NOS tubes or Sovtek EL84M to need no screen dropping resistor. This might be considered a traditional reference layout since lots of old circuits were built like this for NOS 6BQ5s and many still going strong. Most weren't guitar amps though. Behavior is hard to predict with pinpoint accuracy depending on the resistor & cap value variation, but basically the screen will try to draw more current when the amp is loud, then lots more when overdriven sufficiently. Without an extra dropping resistor between the screen supply node & socket pin, the plate may be able to draw down the B+ at the plate, below that being maintained by the discharge of the screen cap through the unlimited screen. At that point the screen becomes a more attractive place for the cathode's ex-electrons compared to the lower voltage of the plate. If this continues long enough the screen fries. Adding for example a 1K resistor as screen dropper to the same circuit would simply limit the current to safe values under forseeable conditions. This reduces the screen voltage inside the bottle though. At idle the plate & screen can still have about the same voltages as before, but as the volume is increased the screen drops proportionally. I interpret this changing behavior as bloom to overdriven playing. Remember with a pentode the plate current is more dependent on the screen voltage than plate voltage. This can sometimes happen without the screen dropper, but with it there you see more of a drop in screen voltages under transient signal conditions. Also, any signal current pulled through the screen is silent compared to that pulled through the plate which has to go through the OPT ands ends up making audible output. So as the screen draws more, it contributes to a kind of subtractive compression (by taking away some plate signal) that adds to any other introduced earlier in the signal chain, and recovers more gradually as the currents restabilize when volume is reduced. Since you control the balancing act partially by choosing the value of the screen dropper it does have some effect. If you are already operating the screen at idle near its limit, that is when a dropping resistor might be absolutely necessary. Then you have no choice but to accept the compression introduced by the lowering screen voltage as you crank it. Conversely, if you have the screen real cool at idle, it can maintain its voltage without a screen dropper as the plate itself compresses by saturation, drawing down B+, or both. You have room to heat it up a bit since it was not stressed to begin with. These seem to me to be the sound variables to juggle, arriving in the middle by ear, then measuring & testing for reliability. In addition, suppose there was a high resistance isolating the plate from the screen supply, but relatively small resistances between the screen & preamp. Still depending on the filter cap values, the plate might be drawn down while both the screen & preamp continue to enjoy higher B+ as they discharge the reserve stored in their filter caps. They don't just look like batteries, they sometimes act that way If the preamp was better isolated from the screen then their relative voltage could vary in either direction depending on how many tubes were in the preamp and their current draw, besides just cap size. OTOH, using smaller resistance between the plate & screen, both their caps are relatively available for the plate itself to discharge, these are the type of traditional circuit that stress the screen less. So there are other components nearby that also effect the dynamic sound of the different values of screen dropper. Especially the OPT, ain't it always the truth? If the DC resistance of its primary is high enough, you will get a big enough voltage drop with the plate drawing idle current, to where the screen can be quite a bit higher in voltage than the plate. The tube can last and sound good, but may suffer in overdrive. You can choose high ohms for the screen dropper to reduce its idle voltage to match the plate, it will drop lots below the plate when you crank it though. Middle of the road would be to choose ohms which leave voltage a little higher than the plate voltage, but just enough to almost follow the plate down as the volume is turned up. Little to no ohms and the screen will always be higher voltage than the plate, just much higher when playing. At the relatively low B+ I have been considering here, and with the typically small power transformer capacity usually used on EL84's, both plate & screen voltage can drop considerably when loud, even with the screen much higher than the plate, it's not going to draw as much at 225V as it does as 250. so there are going to always be lots of variables. I just thought I would add a reminder about my above rambling; most of the behavior I am describing this time is the dynamic response of the power tube I have observed to various changing signal levels, many times outside the linear operating range of it & other components. YMMV & some equations may not apply . . . IMHO dynamic signal analysis seems to be easier for me using a soldering iron rather than a slide rule As an artist doing lead sculpture, you could look at the tube as your subject material, and the components around it as your personal embellishments Maybe that would make me abstract or impressionist Hope this helps, Mike

6/27/2000 9:03 PM

MBSetzer

Hi Steve:

Just a brief comment . . .

This is one of those things I found both listenable & measurable. I may be wacky sometimes but this is more accessible. Especially with EL84 you hear it easier.

If you were only using 250VDC on both plate & screen, and their supplies are adequately decoupled by a choke or about a 5K or less power supply resistor, with good caps about 20mf on the plate & 10mf on the screen, then I would expect good NOS tubes or Sovtek EL84M to need no screen dropping resistor. This might be considered a traditional reference layout since lots of old circuits were built like this for NOS 6BQ5s and many still going strong. Most weren't guitar amps though.

Behavior is hard to predict with pinpoint accuracy depending on the resistor & cap value variation, but basically the screen will try to draw more current when the amp is loud, then lots more when overdriven sufficiently. Without an extra dropping resistor between the screen supply node & socket pin, the plate may be able to draw down the B+ at the plate, below that being maintained by the discharge of the screen cap through the unlimited screen. At that point the screen becomes a more attractive place for the cathode's ex-electrons compared to the lower voltage of the plate. If this continues long enough the screen fries. Adding for example a 1K resistor as screen dropper to the same circuit would simply limit the current to safe values under forseeable conditions. This reduces the screen voltage inside the bottle though. At idle the plate & screen can still have about the same voltages as before, but as the volume is increased the screen drops proportionally. I interpret this changing behavior as bloom to overdriven playing. Remember with a pentode the plate current is more dependent on the screen voltage than plate voltage.

This can sometimes happen without the screen dropper, but with it there you see more of a drop in screen voltages under transient signal conditions. Also, any signal current pulled through the screen is *silent* compared to that pulled through the plate which has to go through the OPT ands ends up making audible output. So as the screen draws more, it contributes to a kind of subtractive compression (by taking away some plate signal) that adds to any other introduced earlier in the signal chain, and recovers more gradually as the currents restabilize when volume is reduced. Since you control the balancing act partially by choosing the value of the screen dropper it does have some effect.

If you are already operating the screen at idle near its limit, that is when a dropping resistor might be absolutely necessary. Then you have no choice but to accept the compression introduced by the lowering screen voltage as you crank it.

Conversely, if you have the screen real cool at idle, it can maintain its voltage without a screen dropper as the plate itself compresses by saturation, drawing down B+, or both. You have room to heat it up a bit since it was not stressed to begin with.

These seem to me to be the sound variables to juggle, arriving in the middle by ear, then measuring & testing for reliability.

In addition, suppose there was a high resistance isolating the plate from the screen supply, but relatively small resistances between the screen & preamp. Still depending on the *filter* cap values, the plate might be drawn down while both the screen & preamp continue to enjoy higher B+ as they discharge the reserve stored in their filter caps. They don't just look like batteries, they *sometimes* act that way

If the preamp was better isolated from the screen then their relative voltage could vary in either direction depending on how many tubes were in the preamp and their current draw, besides just cap size.

OTOH, using smaller resistance between the plate & screen, both their caps are relatively available for the plate itself to discharge, these are the type of traditional circuit that stress the screen less.

So there are other components nearby that also effect the dynamic *sound* of the different values of screen dropper.

Especially the OPT, ain't it always the truth? If the DC resistance of its primary is high enough, you will get a big enough voltage drop with the plate drawing idle current, to where the screen can be quite a bit higher in voltage than the plate. The tube can last and sound good, but may suffer in overdrive. You can choose high ohms for the screen dropper to reduce its idle voltage to match the plate, it will drop lots below the plate when you crank it though.

Middle of the road would be to choose ohms which leave voltage a little higher than the plate voltage, but just enough to almost follow the plate down as the volume is turned up.

Little to no ohms and the screen will always be higher voltage than the plate, just much higher when playing. At the relatively low B+ I have been considering here, and with the typically small power transformer capacity usually used on EL84's, both plate & screen voltage can drop considerably when loud, even with the screen much higher than the plate, it's not going to draw as much at 225V as it does as 250. so there are going to always be lots of variables.

I just thought I would add a reminder about my above rambling; most of the behavior I am describing this time is the dynamic response of the power tube I have observed to various changing signal levels, *many times* outside the linear operating range of it & other components. YMMV & some equations may not apply . . . IMHO dynamic signal analysis seems to be easier for me using a soldering iron rather than a slide rule

As an artist doing lead sculpture, you could look at the tube as your subject material, and the components around it as your personal embellishments

Maybe that would make me abstract or impressionist

Hope this helps,
Mike

6/23/2000 3:50 PM
Don Symes	Re: 3 Watt screen resistors? Take a look at these puppies: http://rocky.digikey.com/USCatalog/V3/387.PDF The voltage rating is a little light, but they're small enough that a couple in series might do it for ya'

Page 1 of 1