And using a higher watt resistor resulting in a little lower temperature will decrease thermal noise some.
Voltage spikes are not necessary. If you put a resistor rated for 250V in a circuit position where it normally has 300V(ish) across it (e.g. Mullard 5/20 first stage anode resistor) then it will eventually fail even if well within its power rating and not subject to voltage spikes.DualTriode said:
As we keep telling you, this is not true. You can lean as much as you like, but the resistors will continue to fail.
PS it may be that for very low value resistors there is a current limit, which again has nothing to do with power dissipation.
it’s not voltage spikes
Hello DF96 and All.
We have made some progress, it’s not voltage spikes.
DF96 I knock on your door because you are smart guy and I am interested in you insights.
I looked up and found as close as I can find to a 5-20 circuit in an old Mullard publication.
Compromised URL removed by Moderation
I am looking at the 20 watt amplifier on page 32 Fig 1. The EF86 anode resistor is R6, 100K ohms and ½ watt. Doing a quick finger exercise on the calculator gives 0.9 watts with 300 volts across the resistor, nearly twice the specified ½ watt resistor. This must not be your specific example. Table 1 gives 85 volts at the EF86 anode and 160 volts on the other side of R6 or 75 volts across the resistor. More finger math gives 0.056 watts.
Will you share the conditions that caused the resistor in you Mullard amplifier to fail?
Thanks DT
Hello DF96 and All.
We have made some progress, it’s not voltage spikes.
DF96 I knock on your door because you are smart guy and I am interested in you insights.
I looked up and found as close as I can find to a 5-20 circuit in an old Mullard publication.
Compromised URL removed by ModerationI am looking at the 20 watt amplifier on page 32 Fig 1. The EF86 anode resistor is R6, 100K ohms and ½ watt. Doing a quick finger exercise on the calculator gives 0.9 watts with 300 volts across the resistor, nearly twice the specified ½ watt resistor. This must not be your specific example. Table 1 gives 85 volts at the EF86 anode and 160 volts on the other side of R6 or 75 volts across the resistor. More finger math gives 0.056 watts.
Will you share the conditions that caused the resistor in you Mullard amplifier to fail?
Thanks DT
Last edited by a moderator:
schematic
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Sorry, I had a senior moment! My '5-20' is a modified one with half an ECC81/12AT7 instead of EF86. My 'R9' is much smaller and my 'R6' is much bigger so the anode resistor has to drop much more voltage. When I used a 0.6W resistor both channels failed within a couple of months, even though the dissipation was probably OK. I think I had about 280-300V on a resistor intended for 250V.
Just stumbled across an interesting Metal Film resistor series. HHV by Yageo. Only in 100K and up, but their 1/4 watt is rated at 1600v, 1/2 watt at 3500v and 1 watt at 5000v.
http://www.yageo.com/documents/recent/High Voltage & High Ohmic Type_HHV.pdf
Digikey has them and singles are like $0.63 CND
http://www.yageo.com/documents/recent/High Voltage & High Ohmic Type_HHV.pdf
Digikey has them and singles are like $0.63 CND
I'm paying $.17 for Vishay and Multicomp 2 and 3 watt metal film 5% resistors from Newark.com. The come in made in India or Thailand, which I view as an advantage compared with another country with outrageous QA failure stories. They are usually 500 v rated, and are about the length of an old carbon comp 1/2 watt. They hiss less than the carbon comp plate resistors I replaced, pretty obviously without a blind A-B setup. The only disadvantage, stock on these seems to stop at > 1 megohm.
KatieandDad's suggestion of spray on coating strikes me as a great way to overheat a resistor. Just buy them long enough to use the 1000 v/inch in air rule, arcovers won't occur. smoking is not the only source of organic dust, humans shed dead skin daily which gets eaten by dust mites and comes out as ****. One arcover, that stuff becomes a carbon track. I don't smoke, and I cleaned lots of dust off the components of my 1961 built ST70 amp in 2010. The builder, a minister leaving on a missionary trip, didn't smoke either.
Last edited:
I have been away for a week.
If you want to read more about resistors than you ever wanted to know and how they fail read this.
Five questions about resistors | EDN
The critical concept behind not exceeding the Maximum Working voltage is Critical Resistance.
Critical Resistance = (Max Volts* Max Volts) / Power Rating
2 examples, Vishay Dale CMF resistors;
CMF55, Max Working Voltage := 250, Max power Rating := 0.5 watt
CMF60, Max Working Voltage := 500, Max power Rating := 1.0 watt
(I pick these resistors because I went off on Ebay and bought an orange crate full of 130+ values)
Just because, I pick ½ of the manufacture’s Max Power listed on the data sheet as Max Power. (less heat and greater reliability)
CMF 55; 250V * 250V/ 0.25W = 250K ohms (Critical Resistance)
CMF 60; 500V * 500V/ 0.5W = 500K ohms (Critical Resistance)
Using a resistor with a value greater than the Critical Resistance puts the Max Voltage is in play, a lesser value resistor places Max Power as the parameter in charge.
Real world I am going to the orange crate and selecting the larger CMF 60 resistor. Check that the power dissipation is less than 0.5 watts and the 500 volts Max Volts will never be a problem.
DT
If you want to read more about resistors than you ever wanted to know and how they fail read this.
Five questions about resistors | EDN
The critical concept behind not exceeding the Maximum Working voltage is Critical Resistance.
Critical Resistance = (Max Volts* Max Volts) / Power Rating
2 examples, Vishay Dale CMF resistors;
CMF55, Max Working Voltage := 250, Max power Rating := 0.5 watt
CMF60, Max Working Voltage := 500, Max power Rating := 1.0 watt
(I pick these resistors because I went off on Ebay and bought an orange crate full of 130+ values)
Just because, I pick ½ of the manufacture’s Max Power listed on the data sheet as Max Power. (less heat and greater reliability)
CMF 55; 250V * 250V/ 0.25W = 250K ohms (Critical Resistance)
CMF 60; 500V * 500V/ 0.5W = 500K ohms (Critical Resistance)
Using a resistor with a value greater than the Critical Resistance puts the Max Voltage is in play, a lesser value resistor places Max Power as the parameter in charge.
Real world I am going to the orange crate and selecting the larger CMF 60 resistor. Check that the power dissipation is less than 0.5 watts and the 500 volts Max Volts will never be a problem.
DT
Except at start-up when the full supply voltage could appear across a resistor before various capacitors becomes charged, even though the steady state dissipation may be minimal.
Speaking of resistors:
Hello,
Speaking of resistors:
Remember that resistors need to have current flow to see a voltage across them. No current means that there is no delta V.
On startup the cathodes are cold and there is no current flowing. There will be a delta V to earth, however there will be no delta V across the anode resistor or for that matter there will be no delta V across the cathode resistor either. I do not see any STOL (Short Time Overload) across either the anode or cathode resistor during startup.
The power supply is a different matter. Looking at the power supply capacitors first, they do have a connection to earth and do see the full peak power supply voltage across them to ground and need to be selected for this occurrence. The power supply resistors also will see peak current and peak voltage across them during start up. To some degree start up inrush needs to be considered for the power supply resistor selection.
Also remember that the resistor Max Working Voltage and Max Power are continuous ratings and that start up is short term.
Personally I do not like to see the lights dim when I turn this stuff on. I include a soft start feature. (GE NTC Inrush Current Limiter).
Short version:
Even on start up the CML-60 500 volt Max Working Volt rating is not a problem, and I do not put ½ watt resistors in the power supply.
DT
Hello,
Speaking of resistors:
Remember that resistors need to have current flow to see a voltage across them. No current means that there is no delta V.
On startup the cathodes are cold and there is no current flowing. There will be a delta V to earth, however there will be no delta V across the anode resistor or for that matter there will be no delta V across the cathode resistor either. I do not see any STOL (Short Time Overload) across either the anode or cathode resistor during startup.
The power supply is a different matter. Looking at the power supply capacitors first, they do have a connection to earth and do see the full peak power supply voltage across them to ground and need to be selected for this occurrence. The power supply resistors also will see peak current and peak voltage across them during start up. To some degree start up inrush needs to be considered for the power supply resistor selection.
Also remember that the resistor Max Working Voltage and Max Power are continuous ratings and that start up is short term.
Personally I do not like to see the lights dim when I turn this stuff on. I include a soft start feature. (GE NTC Inrush Current Limiter).
Short version:
Even on start up the CML-60 500 volt Max Working Volt rating is not a problem, and I do not put ½ watt resistors in the power supply.
DT
There's an easy solution to the max working voltage problem... Put multiple resistors in series. If chosen properly (i.e. to be relatively equal in resistance), they'll split the voltage nicely.
I generally derate power resistors by at least a factor 3~4, otherwise they get too hot for (my) comfort. A power resistor dissipating its rated power at 25 ºC with unrestricted airflow will reach 250 ºC. Even at 1/3 the rated power, they still reach 100 ºC in actual applications.
I don't see a point of using power resistors in applications that don't require it, however. My grid stoppers are all 0.25 W types, for example.
~Tom
I generally derate power resistors by at least a factor 3~4, otherwise they get too hot for (my) comfort. A power resistor dissipating its rated power at 25 ºC with unrestricted airflow will reach 250 ºC. Even at 1/3 the rated power, they still reach 100 ºC in actual applications.
I don't see a point of using power resistors in applications that don't require it, however. My grid stoppers are all 0.25 W types, for example.
~Tom
I never use anything smaller than 1/2 watt resistors in my builds, mainly due to a distrust of tiny parts.
Sent from my SCH-R530U using Tapatalk 2
Sent from my SCH-R530U using Tapatalk 2
Hello, Ok , tempco etc ... oook BUT : why does dynamic increase with wattage ?? i 've tried on different tube amps switching from 1/4w to 2 watts on grids, anode , cathode: result is ALWAYS a big sound 😕
yes : i modified one mono amp, and left the brother unmodified (AB series in both) ... switching from one to other is evident 😕
Something wrong with the resistors then. Maybe the smaller ones were getting too hot and shifting value?
No, they were genuine ones, on a professional amp (ALCATEL amp for the former national french broadcasting agency) . SAme as volume pots : Did you notice the increasing size of "good sounding" potentiometers ? The kozmo and others are a lot bigger than Alps ? the effect is THE same as increasing size of an antenna 😉
NB : compare resistors size from the MacIntosh C22 comemoration and those of an original one .... Size matters : bigger is ... better !
NB : compare resistors size from the MacIntosh C22 comemoration and those of an original one .... Size matters : bigger is ... better !
Last edited:
Sigh.coresta said:
You need to learn some circuit theory. Most resistors in most valve circuits would do little harm even if they were a bit non-linear.
- Status
- Not open for further replies.