Hello,
I've searched here and elsewhere for a good answer and have yet to find one.
How to use the Maximum Working Voltage rating for resistors? I read a debate over whether this describes total voltages (as in wire dialectric ratings) or voltage drop.
Here is my specific question:
I am working on a Tubelab SSE (designed by George Anderson) and he specifies 1/4 watt resistors in many positions (signal path, pre-drive and interstage) in his BOM. This would include the grid stoppers which are in series with the coupling caps--for which he specs 600v parts. The B+ in this amp is 450v.
If I use 1/4 watt resistors rated for 250-300v am I risking falures? Or is it only in the voltage drop?
Thanks! Any and all info is appreciated!
I've searched here and elsewhere for a good answer and have yet to find one.
How to use the Maximum Working Voltage rating for resistors? I read a debate over whether this describes total voltages (as in wire dialectric ratings) or voltage drop.
Here is my specific question:
I am working on a Tubelab SSE (designed by George Anderson) and he specifies 1/4 watt resistors in many positions (signal path, pre-drive and interstage) in his BOM. This would include the grid stoppers which are in series with the coupling caps--for which he specs 600v parts. The B+ in this amp is 450v.
If I use 1/4 watt resistors rated for 250-300v am I risking falures? Or is it only in the voltage drop?
Thanks! Any and all info is appreciated!
All your parts should be rated for 600V. It keeps the electricity out as well as in.
You risk the resistors outer casing coming in contact with 450V B+, almost twice the insulation rating of the resistors you are considering.
Happy New Year.
You risk the resistors outer casing coming in contact with 450V B+, almost twice the insulation rating of the resistors you are considering.
Happy New Year.
Voltage rating,
Its the maximum voltage across a resistor. That doesn't mean it will instantly flash over..but its not good to be close to the maximum.
That is the voltage across it from one end to the other.
Now here is the twist in the tail.
You are assuming that the volt drop across the resistor is constant, in many situations its not!
You have to take into account voltage swing in a circuit and peak values.
Then you would assume that's it..wrong some circuits create back emf and this can exceed the value.
Another situation can be under fault conditions when other components fail and put supply voltage across the part. Some situations work to take advantage of a safe fail.
And as above insulation break down from the component to Gnd.
Then there is wattage rating and working voltage...
One other problem is capacitor loading which can change voltage levels.
Regards
M. Gregg
Its the maximum voltage across a resistor. That doesn't mean it will instantly flash over..but its not good to be close to the maximum.
That is the voltage across it from one end to the other.
Now here is the twist in the tail.
You are assuming that the volt drop across the resistor is constant, in many situations its not!
You have to take into account voltage swing in a circuit and peak values.
Then you would assume that's it..wrong some circuits create back emf and this can exceed the value.
Another situation can be under fault conditions when other components fail and put supply voltage across the part. Some situations work to take advantage of a safe fail.
And as above insulation break down from the component to Gnd.
Then there is wattage rating and working voltage...
One other problem is capacitor loading which can change voltage levels.
Regards
M. Gregg
Last edited:
Yes, this can be confusing!
The voltage rating is the maximum voltage supported by the physical package, irrespective of the resistive element.
In most circuits the actual limit is due to the resistance and heating.
So a 1R resistor can never realistically be subject to the maximum voltage rating.
A 100M bleed resistor in a CRT HV circuit may well be made of 5 20MR resistors to share the voltage load.
The voltage rating is the maximum voltage supported by the physical package, irrespective of the resistive element.
In most circuits the actual limit is due to the resistance and heating.
So a 1R resistor can never realistically be subject to the maximum voltage rating.
A 100M bleed resistor in a CRT HV circuit may well be made of 5 20MR resistors to share the voltage load.
Just build it to the specification he states in the parts list. I built one of the first SSE and it is still going strong.
The voltage rating is the peak voltage across the resistor. You can put a resistor which is rated for 300 volts into a 1500 volt circuit if the voltage drop across the resistor never exceeds 300 volts.
whoa! Thanks for all the quick replies!!
This is an interesting topic. Everywhere I have looked, I have run into this debate of absolute voltage vs. voltage drop.
That being said, I have had my amp running with no issues for 6 months using 1/4 RN60's. I know there have been dozens of Simple SE's built over the last few years. If there were failures, I think they would have come up by now.
I'm putting together some upgrade parts. I suppose I'll just stick with the specified 1/4 watt rating, as I can get all the values I need in the resistor make/models I want to try.
Unless someone has a good argument to the contrary?
Thanks a ton!
John
This is an interesting topic. Everywhere I have looked, I have run into this debate of absolute voltage vs. voltage drop.
That being said, I have had my amp running with no issues for 6 months using 1/4 RN60's. I know there have been dozens of Simple SE's built over the last few years. If there were failures, I think they would have come up by now.
I'm putting together some upgrade parts. I suppose I'll just stick with the specified 1/4 watt rating, as I can get all the values I need in the resistor make/models I want to try.
Unless someone has a good argument to the contrary?
Thanks a ton!
John
The BOM should be fine,
Its from Tubelab..🙂 all the above would be taken into account.
Nothing wrong with up rated parts, however don't confuse voltage rating with wattage rating they don't always follow.
Eg higher wattage isn't always higher working voltage don't get confused.
Wattage is the current through the resistor X the voltage across it<<voltage drop. Which is the heat generated in the device when in circuit.
Regards
M. Gregg
Its from Tubelab..🙂 all the above would be taken into account.
Nothing wrong with up rated parts, however don't confuse voltage rating with wattage rating they don't always follow.
Eg higher wattage isn't always higher working voltage don't get confused.
Wattage is the current through the resistor X the voltage across it<<voltage drop. Which is the heat generated in the device when in circuit.
Regards
M. Gregg
Last edited:
There is a tendency in this forum of exceeding manufacturers 'absolute maximum' ratings particularly on semiconductors but if you build something and it burns your house down, the insurance investigation won't be good.
For your resistors you have max voltage and dissipation and for small value resistors, current. For a 1/2 Watt resistor you'll reach 1/2 Watt at 300 Volts with 180K or less. You also have the insulation resistance of the coatings which is usually greater than the voltage across the resistor. Grid stoppers wouldn't be an issue but the plate and screen circuits would be. BTW it's not a good idea to run at the max value on any component.
Bottom line is make sure your components are 'legal' for the voltage, current and dissipation and remember to derate based on ambient temps.
G²
Note that RN-60 are rated 1/4W at 70 degrees C, and are essentially the military equivalent of a 1/2W resistor rated at 25 degrees C. I generally treat them as half watt resistors and in use derate by 50% or more if feasible. (CMF-60 resistor equivalents of the RN-60 may dissipate as much as 1W at 70 degrees C, make sure you operate them well within the maximum manufacturer ratings for good reliability.)
There are two important considerations with respect to the voltage rating of a resistor.
One has to do with the dielectric strength of the coating on the body of the resistor. This is very important in PTP construction since there is a possibility of two components touching, or a component touching the chassis. A resistor in the power supply may only have a few volts across it, but may have 450 volt to ground. If the body of the resistor were in close proximity to the chassis, the insulation coating on the resistor would be under electric stress and could fail over time, especially with thermal cycling.
This is less of an issue with PC board construction, and can be minimized with proper assembly techniques. The assembly instructions explains that you should not place the resistors in direct contact with the PC board. This allows better airflow around the body of the part, and reduces the electrical stress on the insulation coating since there is air surrounding the resistors, and they are not free to move under normal operating conditions. It's still not a good to test the coating by sticking your finger on a resistor in a live amp to see if it's too hot.....Don't ask me how I learned this!
The second condition is the operating voltage applied between the two terminals of the resistor. This is a separate issue from the resistors dissipation rating. It is possible to apply too much voltage to a high valued resistor without violating its dissipation rating. The ceramic form that the resistive film is deposited on has a breakdown rating. The voltage rating is usually very conservative, but should be respected. The rating is there to prevent the resistor from arcing over if it fails from over dissipation.
Yes, I managed to explode a resistor. It happened in Pete's big red board when I got a little carried away with the power supply knob and fed about 600 volts to the amp. There was a 1 watt resistor feeding a 450 volt cap. The cap shorted placing the entire power supply across the resistor. It exploded, leaving only 2 wires and a black spot, and passed enough energy to scatter the shorted cap all over the room.
The 1/4 watt resistors in the SSE see a low voltage across them. The resistors connected to the CCS IC's do see the full power supply voltage to ground during warm up, but the PC board has ben designed such that there is no adjacent ground plating. During the design phase of the SSE, I tested my board to 550 volts and the only issue I ever had with resistors were the screen stoppers on the output tubes when I ran EL34's into hard clipping by using the SSE as a guitar amp. These resistors were changed to 2 watts before the design was completed.
The SSE has proved to be a reliable design and there is no reason to use larger resistors. In fact reliability may be compromised if physically larger resistors are used due to mechanical stresses placed on the wire leads.
Coupling caps tend to come in 250 volt, 400 volt, and 600/630 volt sizes. The coupling caps may see the full B+ voltage under warm up conditions, and the B+ can be over 500 volts with solid state diodes before the tubes warm up. You could use 450 volt caps with a tube rectifier if you can find them....I have done it and seen no issues, but they are running close to spec, and maybe over spec if the rectifier warms up faster than the 12AT7.
One has to do with the dielectric strength of the coating on the body of the resistor. This is very important in PTP construction since there is a possibility of two components touching, or a component touching the chassis. A resistor in the power supply may only have a few volts across it, but may have 450 volt to ground. If the body of the resistor were in close proximity to the chassis, the insulation coating on the resistor would be under electric stress and could fail over time, especially with thermal cycling.
This is less of an issue with PC board construction, and can be minimized with proper assembly techniques. The assembly instructions explains that you should not place the resistors in direct contact with the PC board. This allows better airflow around the body of the part, and reduces the electrical stress on the insulation coating since there is air surrounding the resistors, and they are not free to move under normal operating conditions. It's still not a good to test the coating by sticking your finger on a resistor in a live amp to see if it's too hot.....Don't ask me how I learned this!
The second condition is the operating voltage applied between the two terminals of the resistor. This is a separate issue from the resistors dissipation rating. It is possible to apply too much voltage to a high valued resistor without violating its dissipation rating. The ceramic form that the resistive film is deposited on has a breakdown rating. The voltage rating is usually very conservative, but should be respected. The rating is there to prevent the resistor from arcing over if it fails from over dissipation.
Yes, I managed to explode a resistor. It happened in Pete's big red board when I got a little carried away with the power supply knob and fed about 600 volts to the amp. There was a 1 watt resistor feeding a 450 volt cap. The cap shorted placing the entire power supply across the resistor. It exploded, leaving only 2 wires and a black spot, and passed enough energy to scatter the shorted cap all over the room.
The 1/4 watt resistors in the SSE see a low voltage across them. The resistors connected to the CCS IC's do see the full power supply voltage to ground during warm up, but the PC board has ben designed such that there is no adjacent ground plating. During the design phase of the SSE, I tested my board to 550 volts and the only issue I ever had with resistors were the screen stoppers on the output tubes when I ran EL34's into hard clipping by using the SSE as a guitar amp. These resistors were changed to 2 watts before the design was completed.
The SSE has proved to be a reliable design and there is no reason to use larger resistors. In fact reliability may be compromised if physically larger resistors are used due to mechanical stresses placed on the wire leads.
Coupling caps tend to come in 250 volt, 400 volt, and 600/630 volt sizes. The coupling caps may see the full B+ voltage under warm up conditions, and the B+ can be over 500 volts with solid state diodes before the tubes warm up. You could use 450 volt caps with a tube rectifier if you can find them....I have done it and seen no issues, but they are running close to spec, and maybe over spec if the rectifier warms up faster than the 12AT7.
Ah! This is great! This answers most of my questions. Thanks George, Kevin (point taken re: the RN60), Gregg, Stratus46, Frank, Chrish, Clifforrest.
Just to be clear, I am in no way criticizing the circuit. It has been very stable and sounds very good. it's just that if I am going to the trouble of rebuilding, I want to be sure I do it right. Also, I am naturally curious and want to learn more.
I did some measuring last night (with test clips and switching off the amp in between) and the only 1/4 watt resistors that see much of any voltage at all are the two that are directly connected to the CCS IC. One is a 330R resistor that sets the current of the CCS. The other is a 1k resistor that lies between the output of the CCS and the inter-stage coupling capacitor which is there to prevent any oscillation of the CCS IC.
Even on initial start up with the SS rectifier (not wanting to risk a short in the 5ar4 switching on and off) I got a max reading of 253-258 volts to ground. This would be getting close to the max working voltage of the RN60. I have no idea how to test the voltage drop. I tried to measure one leg, then the other, in reference to ground, but there was no difference. I (perhaps stupidly) measured across the resistor (neither one goes directly to ground) but got a reading of 0.00 mV.
Is there a way to measure voltage drop with a DMM?
I do plan to rebuild my SSE point to point just for fun, learning, and for future flexibility. My chassis is fairly big and I will not route any component too close to anything electrical or chassis-wise.
(This is off topic but since George was so kind to chime in: since I am building p to p, I thought I might try an extra stage of PS filtering for the driver tube. I have seen this in schematics for many designs including the Dyna ST70 I have at home.
I would use the original CLC filter for the power tubes and add an extra RC stage for the 12at7. (So the PS filter would be CLCRC). Would this reap any benefits, or just be an unnecessary complication?)
Thanks again!
John-
Just to be clear, I am in no way criticizing the circuit. It has been very stable and sounds very good. it's just that if I am going to the trouble of rebuilding, I want to be sure I do it right. Also, I am naturally curious and want to learn more.
I did some measuring last night (with test clips and switching off the amp in between) and the only 1/4 watt resistors that see much of any voltage at all are the two that are directly connected to the CCS IC. One is a 330R resistor that sets the current of the CCS. The other is a 1k resistor that lies between the output of the CCS and the inter-stage coupling capacitor which is there to prevent any oscillation of the CCS IC.
Even on initial start up with the SS rectifier (not wanting to risk a short in the 5ar4 switching on and off) I got a max reading of 253-258 volts to ground. This would be getting close to the max working voltage of the RN60. I have no idea how to test the voltage drop. I tried to measure one leg, then the other, in reference to ground, but there was no difference. I (perhaps stupidly) measured across the resistor (neither one goes directly to ground) but got a reading of 0.00 mV.
Is there a way to measure voltage drop with a DMM?
I do plan to rebuild my SSE point to point just for fun, learning, and for future flexibility. My chassis is fairly big and I will not route any component too close to anything electrical or chassis-wise.
(This is off topic but since George was so kind to chime in: since I am building p to p, I thought I might try an extra stage of PS filtering for the driver tube. I have seen this in schematics for many designs including the Dyna ST70 I have at home.
I would use the original CLC filter for the power tubes and add an extra RC stage for the 12at7. (So the PS filter would be CLCRC). Would this reap any benefits, or just be an unnecessary complication?)
Thanks again!
John-
So you intentionally don't thank the person who gave you the first direct answer, yet thank people who gave you dangerous advice?
You are welcome.
You are welcome.
Oh! Sorry H E Pennypacker! The list of replies was a full page back by the time I replied. I just rattled off from memory! I didn't mean to skip anyone!
Ideally I would like to hear a broad range of answers based of different sets of information that take different criteria into consideration. Best way to learn, I feel. And your input was important. In fact, your reply echoed my concern in the first place.
So, belated but sincere: Thank you H E Pennypacker.
Ideally I would like to hear a broad range of answers based of different sets of information that take different criteria into consideration. Best way to learn, I feel. And your input was important. In fact, your reply echoed my concern in the first place.
So, belated but sincere: Thank you H E Pennypacker.
Its easier,
to just say thank's everyone...its all done in the best possible taste 😛..😀
Regards
M. Gregg
to just say thank's everyone...its all done in the best possible taste 😛..😀
Regards
M. Gregg
" There is a tendency in this forum of exceeding manufacturers absolute maximum ratings..."
I don't think so, as a matter of fact, I tend to do the opposite. Right now my 12B4 amp is getting dual Mains transformers just because I don't like running anything anywhere near maximums. Each channel has 1.2A filament draw, sourced by a 3A secondary. My single 12AU7 at 150mA filament, sourced by a dual bobbin 20V 1.5A transformer going into an LM317 with a big honking heatsink. The 317 runs at 1.2A.
I like to go 33% to 66% of whats available. My opts run at 6W ...I'll be Lucky to get 2.5W.
Poking around inside commercial stuff, I just can't stand components pushed to maximums just because they are cheap, ever looking at the bottom line.
________________________________________________________Rick......
I don't think so, as a matter of fact, I tend to do the opposite. Right now my 12B4 amp is getting dual Mains transformers just because I don't like running anything anywhere near maximums. Each channel has 1.2A filament draw, sourced by a 3A secondary. My single 12AU7 at 150mA filament, sourced by a dual bobbin 20V 1.5A transformer going into an LM317 with a big honking heatsink. The 317 runs at 1.2A.
I like to go 33% to 66% of whats available. My opts run at 6W ...I'll be Lucky to get 2.5W.
Poking around inside commercial stuff, I just can't stand components pushed to maximums just because they are cheap, ever looking at the bottom line.
________________________________________________________Rick......
The 'voltage drop' across the resistor is the difference in voltage between the point in the circuit before the resistor to the point in the circuit after the resistor. So, to measure the difference in this voltage, you just probe the two points (i.e., the two legs of the resistor) with your DMM in DC voltage setting. That simple. In fact, by using a 1 ohm (or 10 ohm) resistor at strategic points in a circuit such as cathode to ground on the driver tube in a fixed bias amp, you can use the voltage drop across that resistor to easily calculate current using ohms law.
Regarding 'upgrading' a SSE, don't even bother replacing the resistors for 'magic' parts. You are likely to damage the printed circuit board and won't hear any difference. As George states on his site, from memory, the best gains in the design are to be had by using a decent output transformer, plus a choke and motor run capacitor for better power supply filtering. If you really can't help yourself, maybe consider using a reasonable quality coupling capacitor, but remember 'boutique' parts are simply a method of separating the ignorant from their cash 😉
- Status
- Not open for further replies.