Correct measurement of repaired Guitar Amplifiers
Question:
I am doing repairs of Guitar amplifiers but am in debate with a customer on the output power of an amp that has been fixed. It is a Fender Twin Reverb with Master volume with 4 x 6L6WXT new tubes. This is my power measurement setup and proceedure. 1. Connect 1KHZ sine wave generator to input but turned off. 2. Connect output of the amplifier to non inductive high power resistor matched to amplifier output. 3. Connect Osciloscope and True RMS Meter across the resistor. 4. Turn on the amplifier for 1 minute. 5. Turn up the signal generater until you see clipping on the sine wave and back it off a little. Use the RMS value of the output to the multimeter in the following formula. P=V^2/R In this this current case. V = 50V p-p VRMS = 17.675 VDC Aproximatly R = 4 Ohms So this would give P = 17.675 ^2/4 P = 78.1 Watts I would expect about 100W RMS from an amp like this but there is no indication of any fault. Is my formula correct? The frustrating thing with Tube amps is the difficulty of getting schematics or specifications to match the exact amp you are working on. Valves were not in common use when I did my training and started repairs so I am having to learn what seems to be more intuitive repair and testing proceedures in Guitar amps. In all my other power amplifier repairs I have a power output value and a model number that matches the schematic most of the time and that is simple. I am too young to have exprienced Valve repairs but hopefully not too old to learn. Most of the time it is a no brainer but I get the odd amp I have difficulty with like in this case. I have had it put to me by an experienced repairer that I use the forumula above but instead of V being RMS it is Peak then at the end apply *.707 which in fact gives me over 100W but hours on the internet has always lead me back to my above proceedure. I must be doing something right as I rarely get complaints and when I do it is a job with intermitant fault I thought was fixed or the occational faulty output valve. Yes I match and bias valves and replace the complete output set at the same time. Any ideas and suggestions welcome. I don’t however need to know complex power calculations that are mentioned or get into a power debate, I only need to confirm the amplifiers rated output and to check it is performing correctly before returning to the client. Thanks
Response:
– Hide quoted text — Show quoted text – > I am doing repairs of Guitar amplifiers but am in debate with a > customer on the output power of an amp that has been fixed. > It is a Fender Twin Reverb with Master volume with 4 x 6L6WXT new > tubes. > This is my power measurement setup and proceedure. > 1. Connect 1KHZ sine wave generator to input but turned off. > 2. Connect output of the amplifier to non inductive high power resistor > matched to amplifier output. > 3. Connect Osciloscope and True RMS Meter across the resistor. > 4. Turn on the amplifier for 1 minute. > 5. Turn up the signal generater until you see clipping on the sine wave > and back it off a little. > Use the RMS value of the output to the multimeter in the following > formula. > P=V^2/R > In this this current case. > V = 50V p-p > VRMS = 17.675 VDC Aproximatly > R = 4 Ohms > So this would give > P = 17.675 ^2/4 > P = 78.1 Watts > I would expect about 100W RMS from an amp like this but there is no > indication of any fault. > Is my formula correct? > The frustrating thing with Tube amps is the difficulty of getting > schematics or specifications to match the exact amp you are working on. > Valves were not in common use when I did my training and started > repairs so I am having to learn what seems to be more intuitive repair > and testing proceedures in Guitar amps. In all my other power amplifier > repairs I have a power output value and a model number that matches the > schematic most of the time and that is simple. > I am too young to have exprienced Valve repairs but hopefully not too > old to learn. Most of the time it is a no brainer but I get the odd amp > I have difficulty with like in this case. > I have had it put to me by an experienced repairer that I use the > forumula above but instead of V being RMS it is Peak then at the end > apply *.707 which in fact gives me over 100W but hours on the internet > has always lead me back to my above proceedure. I must be doing > something right as I rarely get complaints and when I do it is a job > with intermitant fault I thought was fixed or the occational faulty > output valve. Yes I match and bias valves and replace the complete > output set at the same time. > Any ideas and suggestions welcome. > I don’t however need to know complex power calculations that are > mentioned or get into a power debate, I only need to confirm the > amplifiers rated output and to check it is performing correctly before > returning to the client. > Thanks
A few things.come to mind: 1) 80W is reasonable for clean output on a SF Twin, especially with modern 6L6-like objects. Actually, it’s pretty darned good. 2) Those 100W are "marketing watts" Notice they don’t specify distortion level where the 100W output occurs. Just for fun, crank the amp to a full square wave, THEN use your T-RMS meter (or measure peak watts on the scope and DON’T multiply by 0.7 …this works if it’s a FULL square wave). 3) I’ve noticed that NiCr wound resistors can generate a DC voltage when heated up. This can foul up any meter reading by maybe 5-10%. Try this sometime: Cook the resistor with full output for 5 minutes, turn off the signal and measure voltage across the resistor. A couple hundred mV isn’t unreasonable. –Mike Mike Schway | [Picture your favorite quote here]
Response:
You’re doing it exactly right. Most non-ultralinear Twin Reverbs will put out 70-80 watts RMS across a resistive load. If you correct for line sag with your variac, you can get a little more output before clipping. The RMS output wattage rating on the back of Fender amplifiers is bullshit anyway, as no distortion spec is given. Lord Valve Expert – Hide quoted text — Show quoted text – > I am doing repairs of Guitar amplifiers but am in debate with a > customer on the output power of an amp that has been fixed. > It is a Fender Twin Reverb with Master volume with 4 x 6L6WXT new > tubes. > This is my power measurement setup and proceedure. > 1. Connect 1KHZ sine wave generator to input but turned off. > 2. Connect output of the amplifier to non inductive high power resistor > matched to amplifier output. > 3. Connect Osciloscope and True RMS Meter across the resistor. > 4. Turn on the amplifier for 1 minute. > 5. Turn up the signal generater until you see clipping on the sine wave > and back it off a little. > Use the RMS value of the output to the multimeter in the following > formula. > P=V^2/R > In this this current case. > V = 50V p-p > VRMS = 17.675 VDC Aproximatly > R = 4 Ohms > So this would give > P = 17.675 ^2/4 > P = 78.1 Watts > I would expect about 100W RMS from an amp like this but there is no > indication of any fault. > Is my formula correct? > The frustrating thing with Tube amps is the difficulty of getting > schematics or specifications to match the exact amp you are working on. > Valves were not in common use when I did my training and started > repairs so I am having to learn what seems to be more intuitive repair > and testing proceedures in Guitar amps. In all my other power amplifier > repairs I have a power output value and a model number that matches the > schematic most of the time and that is simple. > I am too young to have exprienced Valve repairs but hopefully not too > old to learn. Most of the time it is a no brainer but I get the odd amp > I have difficulty with like in this case. > I have had it put to me by an experienced repairer that I use the > forumula above but instead of V being RMS it is Peak then at the end > apply *.707 which in fact gives me over 100W but hours on the internet > has always lead me back to my above proceedure. I must be doing > something right as I rarely get complaints and when I do it is a job > with intermitant fault I thought was fixed or the occational faulty > output valve. Yes I match and bias valves and replace the complete > output set at the same time. > Any ideas and suggestions welcome. > I don’t however need to know complex power calculations that are > mentioned or get into a power debate, I only need to confirm the > amplifiers rated output and to check it is performing correctly before > returning to the client. > Thanks
Response:
- Hide quoted text — Show quoted text – > I am doing repairs of Guitar amplifiers but am in debate with a > customer on the output power of an amp that has been fixed. > It is a Fender Twin Reverb with Master volume with 4 x 6L6WXT new > tubes. > This is my power measurement setup and proceedure. > 1. Connect 1KHZ sine wave generator to input but turned off. > 2. Connect output of the amplifier to non inductive high power resistor > matched to amplifier output. > 3. Connect Osciloscope and True RMS Meter across the resistor. > 4. Turn on the amplifier for 1 minute. > 5. Turn up the signal generater until you see clipping on the sine wave > and back it off a little. > Use the RMS value of the output to the multimeter in the following > formula. > P=V^2/R > In this this current case. > V = 50V p-p > VRMS = 17.675 VDC Aproximatly > R = 4 Ohms > So this would give > P = 17.675 ^2/4 > P = 78.1 Watts > I would expect about 100W RMS from an amp like this but there is no > indication of any fault. > Is my formula correct? > The frustrating thing with Tube amps is the difficulty of getting > schematics or specifications to match the exact amp you are working on. > Valves were not in common use when I did my training and started > repairs so I am having to learn what seems to be more intuitive repair > and testing proceedures in Guitar amps. In all my other power amplifier > repairs I have a power output value and a model number that matches the > schematic most of the time and that is simple. > I am too young to have exprienced Valve repairs but hopefully not too > old to learn. Most of the time it is a no brainer but I get the odd amp > I have difficulty with like in this case. > I have had it put to me by an experienced repairer that I use the > forumula above but instead of V being RMS it is Peak then at the end > apply *.707 which in fact gives me over 100W but hours on the internet > has always lead me back to my above proceedure. I must be doing > something right as I rarely get complaints and when I do it is a job > with intermitant fault I thought was fixed or the occational faulty > output valve. Yes I match and bias valves and replace the complete > output set at the same time. > Any ideas and suggestions welcome. > I don’t however need to know complex power calculations that are > mentioned or get into a power debate, I only need to confirm the > amplifiers rated output and to check it is performing correctly before > returning to the client. > Thanks
It depends on a few factors that have to looked for. One of the common reasons for a low output Twin Reverbs is the use of Polished Turds from off shore. I’ve setup win Reverbs with a set of Sylvania 6L6-GC tubes and found a 100 watts. Then, when swapped with a set of off shore jobbers, found the same result you’ve reported. Now, if the B+ is proper for the unit, you have to look at other factors that causes premature clipping. Like the phase inverter circuit gain. Premature clipping in the phase inverter ckt will lead you into thinking the low power output would be cause else where. I have come across phase inverters that clipped before the output section would. Since you report the power rating of 100 watts on the back of the Twin Reverb as 100 watts, this is not a ultra linear output Twin Reverb. Which, states 135 watts as the rated output. With these ultra linear units, the same premature clipping can be experienced, and when corrected, full output yield can be found. It is when you never look for premature clipping from the front of the signal chain, that wrong conclusions are drawn. It is when you work from the front end to the back, that you can find a lot of amazing watts. Having good tubes loaded properly to the output tranny helps. There are many older output trannys that are ofer built and will surprise you the amout of wattage they can supply without protest. The distortion spec has been mentioned in this thread, and I know for a fact that most of the Ed Jahns Fender amps had them in house. For example, the 400-PS by eye on the scope would yield quite a bit over 500 watts to the three test loads required. However, when set to its distortion spec, it would yield the rated 435 watts stated on the back panel. That, provided no Polished Turds were used as power output tubes in the output section. Always look for premature clipping ahead of the stage you are looking at. Always monitor Line Voltage,and Line Current. It’s the only way to know what is REALLY going on with what sits on your test bench. Regards, Rich Koerner, Time Electronics. http://www.timeelect.com Specialists in Live Sound FOH Engineering, Music & Studio Production, Vintage Instruments, and Tube Amplifiers
Response:
> I am doing repairs of Guitar amplifiers but am in debate with a > customer on the output power of an amp that has been fixed. > It is a Fender Twin Reverb with Master volume with 4 x 6L6WXT new > tubes. <snip>
Ugh–get rid of those tubes! There are *far* better tubes to use in a Fender than those cruddy Sovteks. Try SED, Shuguang, Svetlana (all three of which are pretty inexpensive), old Sylvanias, or JAN Philips tubes. You’ll get more clean powre out them in comparison to the WXTs. –E
Response:
– Hide quoted text — Show quoted text ->I am doing repairs of Guitar amplifiers but am in debate with a >customer on the output power of an amp that has been fixed. >It is a Fender Twin Reverb with Master volume with 4 x 6L6WXT new >tubes. >This is my power measurement setup and proceedure. >1. Connect 1KHZ sine wave generator to input but turned off. >2. Connect output of the amplifier to non inductive high power resistor >matched to amplifier output. >3. Connect Osciloscope and True RMS Meter across the resistor. >4. Turn on the amplifier for 1 minute. >5. Turn up the signal generater until you see clipping on the sine wave >and back it off a little. >Use the RMS value of the output to the multimeter in the following >formula. >P=V^2/R >In this this current case. >V = 50V p-p >VRMS = 17.675 VDC Aproximatly >R = 4 Ohms >So this would give >P = 17.675 ^2/4 >P = 78.1 Watts >I would expect about 100W RMS from an amp like this but there is no >indication of any fault. >Is my formula correct? >The frustrating thing with Tube amps is the difficulty of getting >schematics or specifications to match the exact amp you are working on. >Valves were not in common use when I did my training and started >repairs so I am having to learn what seems to be more intuitive repair >and testing proceedures in Guitar amps. In all my other power amplifier >repairs I have a power output value and a model number that matches the >schematic most of the time and that is simple. >I am too young to have exprienced Valve repairs but hopefully not too >old to learn. Most of the time it is a no brainer but I get the odd amp >I have difficulty with like in this case. >I have had it put to me by an experienced repairer that I use the >forumula above but instead of V being RMS it is Peak then at the end >apply *.707 which in fact gives me over 100W but hours on the internet >has always lead me back to my above proceedure. I must be doing >something right as I rarely get complaints and when I do it is a job >with intermitant fault I thought was fixed or the occational faulty >output valve. Yes I match and bias valves and replace the complete >output set at the same time. >Any ideas and suggestions welcome. >I don’t however need to know complex power calculations that are >mentioned or get into a power debate, I only need to confirm the >amplifiers rated output and to check it is performing correctly before >returning to the client. >Thanks
Your formulae seem to be correct… but if you are testing to see if the amp is working properly, check the plate voltage at zero and full power, and check the sine wave at the plates of the tubes when testing. Then calculate the turns ratio of the transformer and see if the output matches. If all is OK, the hell with the power measurement! Power out rating on a tube amp is just a suggestion…
Response:
Thanks for the fast detailed replies. The clipping only occurs at the output all stages up to the output of the driver are clean. I assume plate voltage is an indication of how well the power supply is working under load. I see some people don’t like the sovtek valves but in Australia it is hard to get other ones at reasonable prices but now I know it is such a big issue I will hunt around for others or import directly unless someone knows of a good Australian supplier. I have found sovtek valves will vary and have had instances where one of the same batch would be way down so have thrown it out. It was interesting about the voltage on the resistors when they get hot, I would have never thought of that. Are there any good books or online rescources to lift my game on repairing Valve Amplifiers? I get the job done but at times it takes too long because my diagnosis is slow particulary in the preamp stages which would account for about 20% of repairs. Can someone fill me in on the turns ratio? How do you calculate it and how do you determine a shorted turn in an output transformer other than replacement. I have spare transformers to swap in to test and about half the time I guess the transformer is faulty I am right. I am very good at electronic repairs but sadly lacking a bit in Valve amp theory and practice. I don’t charge out my full rate on Valve amps often as I charge by the fault and repair not time as I consider that is the fair aproach. As I have said last time clients are very happy it is mostly the time it takes me to get the job done. I did not expect fast and quality replies so quick. I tend to avoid newsgroups due to the lack of quality advice but here it looks like we have some experts 
Response:
Just for grins, measure it at somewhere around 250 Hz or 500 Hz. You may find more power. 8^)
Response:
> Thanks for the fast detailed replies. > The clipping only occurs at the output all stages up to the output of > the driver are clean. > I assume plate voltage is an indication of how well the power supply is > working under load.
It is normal in most amps to find a voltage drop of 20 volts between the no signal and and signal conditions. Anything more than that, should be looked into. Power is the product of voltage and current. If one of the two drops or is limited, you will have a loss of power delivery. If you find the power supply to not be the limiting power factor, then you have a case of the power tubes belonging to the Polished Turd family of tubes, where their gain and headroom is not their hallmark. This happens when you live in a world of Polished Turds. Actually, I think it would be a good idea if AGA would provide a Polished Turd LIST on the AGA web site, to warn amp users of these limp performing tubes. If musicians didn’t buy them, they will just go away. It would be a great consumer "service" AGA could provide and save musicians a pile of money playing the "vacuum tube" Roulette game. http://amazinggamblingsystems.com/ – Hide quoted text — Show quoted text -> I see some people don’t like the sovtek valves but in Australia it is > hard to get other ones at reasonable prices but now I know it is such a > big issue I will hunt around for others or import directly unless > someone knows of a good Australian supplier. I have found sovtek valves > will vary and have had instances where one of the same batch would be > way down so have thrown it out. > It was interesting about the voltage on the resistors when they get > hot, I would have never thought of that. > Are there any good books or online rescources to lift my game on > repairing Valve Amplifiers? I get the job done but at times it takes > too long because my diagnosis is slow particulary in the preamp stages > which would account for about 20% of repairs. > Can someone fill me in on the turns ratio? How do you calculate it and > how do you determine a shorted turn in an output transformer other than > replacement. I have spare transformers to swap in to test and about > half the time I guess the transformer is faulty I am right. > I am very good at electronic repairs but sadly lacking a bit in Valve > amp theory and practice. I don’t charge out my full rate on Valve amps > often as I charge by the fault and repair not time as I consider that > is the fair aproach. As I have said last time clients are very happy it > is mostly the time it takes me to get the job done. > I did not expect fast and quality replies so quick. I tend to avoid > newsgroups due to the lack of quality advice but here it looks like we > have some experts
Regards, Rich Koerner, Time Electronics. http://www.timeelect.com Specialists in Live Sound FOH Engineering, Music & Studio Production, Vintage Instruments, and Tube Amplifiers
Response:
Hey Agent! (annotations below)
<snip> > I see some people don’t like the sovtek valves but in Australia it is > hard to get other ones at reasonable prices
<snip> Repeating what others have already said, sounds like the output tubes are ‘neutering’ the amp a bit. I’ll chime in that JJ 6L6 output tubes seem incredibly well designed for older stock Fender amps’ plate loads/voltages etc.; i.e. in your amp, by example, I’d expect ~80 watts out for a quad of Sovtek 5881WXTs’ and 105 watts+ for a quad of JJ 6L6GCs’. FYI on later amps, Fenders’ distortion spec is 5% THD at 1 Khz; if you looked at a scope (while actually measuring THD), you’d see that 5% is a fairly ‘distorted looking’ waveform. <snip> > Can someone fill me in on the turns ratio? How do you calculate it and > how do you determine a shorted turn in an output transformer other than > replacement. I have spare transformers to swap in to test and about > half the time I guess the transformer is faulty I am right.
<snip> AFA a quick and easy way, if you build one of these: http://www.braught.com/TestEquip/QuickTracer/QuickTracer.htm and get used to using it, it’ll immediately show you if you’ve got either a shorted primary *or* secondary turn (when looking at the primary winding of a de-energized output transformer with no load attached.) One no extra tools approach to identify if a short is in the output circuit (can include a shorted bypass capacitor, sometimes found across output transformer connections) is to look at power supply current consumption vs. output power; i.e. when driving an amp with signal and the AC mains current goes up significantly but you’re not getting much out at the speaker connections, that’s a possible indication of a shorted output transformer (when you’ve got a shorted output transformer, the ’signal juice’ stays in the transformer and heats it up.) HTH, -Robert QTS http://www.Braught.com
Response:
>Thanks for the fast detailed replies. >The clipping only occurs at the output all stages up to the output of >the driver are clean. >I assume plate voltage is an indication of how well the power supply is >working under load.
The voltage shouldn’t drop a hell of a lot, maybe 10% is acceptable… but if you can’t get much power out it may be a good place to look… maybe it’s dropping a lot! Don’t forget the screen voltage – very important too… – Hide quoted text — Show quoted text ->I see some people don’t like the sovtek valves but in Australia it is >hard to get other ones at reasonable prices but now I know it is such a >big issue I will hunt around for others or import directly unless >someone knows of a good Australian supplier. I have found sovtek valves >will vary and have had instances where one of the same batch would be >way down so have thrown it out. >It was interesting about the voltage on the resistors when they get >hot, I would have never thought of that. >Are there any good books or online rescources to lift my game on >repairing Valve Amplifiers? I get the job done but at times it takes >too long because my diagnosis is slow particulary in the preamp stages >which would account for about 20% of repairs. >Can someone fill me in on the turns ratio? How do you calculate it
The turns ratio is the sqr.root of the impedance ratio… You can find this data in a tube manual… E.G. Lets say you have a 6BQ5, and an 8 ohm rated amp… then the transformer is matching about 4500 to 8 ohms, or 562 Z ratio, which is a turns ratio of 24 to 1. The signal voltage on the output should be 1/24 the signal voltage on the primary. NOTE that the manufacturer might use a poor matching transformer, this is a test best done with experience on that model amp. Note that some amps are all over the place, especially with P-P 6L6s… so it’s best to make a record of what you find in good amps, and check it in suspect amps. (of the same model) >and >how do you determine a shorted turn in an output transformer other than >replacement. I have spare transformers to swap in to test and about >half the time I guess the transformer is faulty I am right.
Normally a shorted turn or something will cause a GREAT change in output… practically nothing… Don’t forget to look for core shorts – very nasty… – Hide quoted text — Show quoted text ->I am very good at electronic repairs but sadly lacking a bit in Valve >amp theory and practice. I don’t charge out my full rate on Valve amps >often as I charge by the fault and repair not time as I consider that >is the fair aproach. As I have said last time clients are very happy it >is mostly the time it takes me to get the job done. >I did not expect fast and quality replies so quick. I tend to avoid >newsgroups due to the lack of quality advice but here it looks like we >have some experts
Response:
I tried some Electro Harmonics and I got another 10W so in the ball park. Thanks for all the advice. Let me know if EH tubes are considered acceptabe
Response:
Agent86, tied to the tracks, said: > I tried some Electro Harmonics and I got another 10W so in the ball > park. Thanks for all the advice. Let me know if EH tubes are considered > acceptabe
That depends on who you talk to. Would you believe… it’s up to you?
Response:
> I tried some Electro Harmonics and I got another 10W so in the ball > park. Thanks for all the advice. Let me know if EH tubes are considered > acceptabe
They’re fine–another Sovtek tube. Much better than the Sovtek 5881WXT’s in a Fender. I suspect you’d prefer SED 6L6GC’s if you tried them, though. If SED tubes aren’t available in Australia, have you considered ordering them from a vendor in the USA or UK? Even with the cost of shipping, it might still beat your local price. –E
Response:
Related Posts