The following is a collection of emails between Bruce, creator of the Oddwatt amplifiers, and myself regarding the general build of the Oddwatt amplifiers (Particularly the Odd Blocks) found at DIYaudioprojects.com. Bruce had been so kind in assisting me in the build of his design and has written multiples of responses to my questions over the past several days. It began to dawn on me that this was some very good info and all of it was private, address to me alone. I asked Bruce if he would consider allowing me to create a post on the forum and continue to answer all of my questions there so everyone else could see and enter in if they wish. He of course agreed. Below is a cut-and-paste of our email conversation to date with the last question directed to him. Please allow Bruce to answer it before others join in and then we can go from there. Thanks again Bruce for your time, patience, and assistance in helping build the really cool design.
P.S. Bruce has a website (Oddwatt.com) where he sells this amp in a kit. He also has a really cool new budget amp that will be coming out soon, the Poddwatt.
7/11/09 – My first question to Bruce RE: Building the Oddwatt / Oddblock amps
Bruce,
After much discussion with others and contemplation/study of your Oddwatt designs I have settled that I am going to build it. I am going to basically build the KT88 version on a single chassis with a separate PSU like you did with the KT77 amp. I realize they are basically the same with the exception of the 12 vs 6 series tubes and corresponding component values on the pre side of the amp. I had several thoughts about these two designs and was wondering if you would give me some advise. Fist is which do you prefer the 12 or 6 series tubes on the front end of the amp? Second, I found an industrial grade Hammond power trafo unlike anything I have ever seen before....it is massive and produces 250VA with 117AC primaries with a secondary winding that has 0-200-230-460 taps (could be used as a 230-0-230 if the 0, 230, and 460 taps were used with the 230 as the 0 CT and the 0 as one 230 and the 460 as the other 230) It should have plenty of power to run the 2 channels of the KT88 arrangement (I think) My question is could this be used with a few adjustments in the resistors in the PSU to give the desired B+ voltages? I haven't sat down to calculate them out yet. Last question is could I get all of the parts from you and your partner sans the trafos, chassis, and power resistors I will have to change out and tubes? It would save me time in acquisition but if you can't I would certainly understand. I have spent alot of time wrapping my brain around how your design with the LM317 and self-inverting circuit...it really is cool! Did you ever get around to building it with a FET in place of the 317? What kind of 25R pot did you use for the balancing circuit? It seems to me this should be of the highest quality. Please let me know what you think. If you haven't yet tried the FET I may try to figure it out in this one.
Thanks for your time and help.
Sincerely,
Jeff Miller
Lawton, OK USA
P.S. Bruce has a website (Oddwatt.com) where he sells this amp in a kit. He also has a really cool new budget amp that will be coming out soon, the Poddwatt.
7/11/09 – My first question to Bruce RE: Building the Oddwatt / Oddblock amps
Bruce,
After much discussion with others and contemplation/study of your Oddwatt designs I have settled that I am going to build it. I am going to basically build the KT88 version on a single chassis with a separate PSU like you did with the KT77 amp. I realize they are basically the same with the exception of the 12 vs 6 series tubes and corresponding component values on the pre side of the amp. I had several thoughts about these two designs and was wondering if you would give me some advise. Fist is which do you prefer the 12 or 6 series tubes on the front end of the amp? Second, I found an industrial grade Hammond power trafo unlike anything I have ever seen before....it is massive and produces 250VA with 117AC primaries with a secondary winding that has 0-200-230-460 taps (could be used as a 230-0-230 if the 0, 230, and 460 taps were used with the 230 as the 0 CT and the 0 as one 230 and the 460 as the other 230) It should have plenty of power to run the 2 channels of the KT88 arrangement (I think) My question is could this be used with a few adjustments in the resistors in the PSU to give the desired B+ voltages? I haven't sat down to calculate them out yet. Last question is could I get all of the parts from you and your partner sans the trafos, chassis, and power resistors I will have to change out and tubes? It would save me time in acquisition but if you can't I would certainly understand. I have spent alot of time wrapping my brain around how your design with the LM317 and self-inverting circuit...it really is cool! Did you ever get around to building it with a FET in place of the 317? What kind of 25R pot did you use for the balancing circuit? It seems to me this should be of the highest quality. Please let me know what you think. If you haven't yet tried the FET I may try to figure it out in this one.
Thanks for your time and help.
Sincerely,
Jeff Miller
Lawton, OK USA
Bruce’s response:
Hi Jeff, There is no particular reason not to use a single chassis. It can save a few $$ on parts and depending on the rest of your system not cause any significant degradation of the separation. That is probably the strong point of mono-blocks. Another plus with them is that you can place them fairly near the speakers and reduce the length and effects of the speaker wire. The Hammond may be OK. The 230-0-230 is fine. My question is at how much current. You might be right at the limit for the transformer. With the Hammond you would expect about 325 B+ using solid-state rectifiers and capacitor input. If you keep the current at 180 mA per pair it will most likely be OK. The amount available ought to be at least 1.5 times the max-expected drain. If the trannie can deliver 500 mA everything should be fine. The only voltages that would need to be changed are the resistors in the power supply that feed the driver stage. When you decide what power trannie you will use and such I'll do some number crunching for you on the values.
I gave up on the FET CCS. Mostly because I could not get a reliable one that had a lower voltage drop than the LM317s. I am always open for ideas on it though. The 25 ohm pot is from (sad to say) Radio Shack part number 271-0265. There does not seem to be an alternate source for them. we have found 20 ohm ones at 1/2 watt and they would do for EL84s and similar sized tubes, but not for the bigger guys. It is feasible to use a 12 position switch with 1 and 2 ohm resistors and get a close balance. I would do that in a pinch, but my feeling is that part of the excellent performance is due to closely matching the idle currents.
As for getting parts only from us, I'll confer with Rodney. I would not use different output trannies though. There are several tweaks to the circuit that are not obvious that make everything work together and improve the performance from a garden-variety amp to one at a much higher level. The output trannies are hard to beat for $$ and they are so underrated it is incredible. I really cannot guess how good (or bad) other trannies would sound in this circuit. Before tweaking the amps were pretty good. About 1% THD up to half power and around 2% at max. Bandwidth was ok, but it didn't show the above band sensitivity at around 70K. Much of this is related to the output trannies. All trannies have problems like this. So the tweaking brought the distortion down by a factor of 4 and removed the sensitivity at 70K. I am sure others could be tweaked, --- if you have a good signal generator, digital storage scope and distortion analyzer. Few have the equipment. A related issue is that the circuit with the components selected is very tolerant of tube types. I'm not sure that other outputs (and there are many good ones) would work as well. I switch back and forth between Gold Lion KT88s, JJ KT77s, JJ 6L6GCs and Sovtek 6L6WXT. All sound good.
My sense is that if you don't need the power of the KT88s the JJ 6L6GCs sound very similar and are 1/3 the price. You can always upgrade later. The only circuit change is the resistor in the CCS.
Good listening
Bruce
Additional Response:
Hi Jeff, in looking over my earlier answer, I realized that I missed part of your questions. This is regarding the 12 vs 6-volt tubes. The design was based originally on using 6SL7s or 12SL7s. Except for the heater rating they are identical. I wanted to use 12 volts for two reasons, one I was able to get 12 volt SMPS for dirt-cheap prices and use them for the heaters. They make great heater power sources. Second when I discarded the SMPS and wanted DC on the heaters it is easier to get 3 amps at 12 volts than 6 amps at 6 volts from most power transformers and filter circuits. So I stuck with 12 volts. Now additionally the 6/12SL7s are noted for having a very nice sound and pretty high gain. I tried the entire 12A_7 bunch and could not find one that sounded as good with enough gain. So I did a bit of research and came up wit the 5751s that are now in most of my amps. They have nearly the same gain as the 6/12SL7s and sound very nice (the ones I use are Phillips NOS). For lower gain circuits like for 6V6 and EL84 output tubes the ECC802S is the tube of choice. It is electrically like a 12AU7, but sounds a whole lot better. So really it doesn't matter what voltage you choose, there are tubes that will run on it. (both the ECC802S and 5751 will run on either 6 or 12)
Good listening
Bruce
Follow up question by me:
Bruce,
The circuitry is slightly different between the 2 different tubes on the 77 and 88 on the pre-side of the amp (i.e. some of the resistor and cap values are about 1/2 on one vs the other). I was wondering if I were to use a 5751 which of the 2 pre schematics would you use, the front end of the KT77 or the KT88 design? I didn't know if the values made that much difference in either case. The Hammond trafo I have should put out around 1 amp at 250v and 0.75 amps at 375v by my calculations. I actually found this thing on a junk pile and it was given to me. Similar ones I have researched in the industrial lines they make cost about $500! I am going to try it, I will let you know how it turns out. I want to make x2 separate chassis just because I always wanted an amp with a power umbilical and separate supply. This thing is too big to put with the amp I think, rather ugly too. It will be housed for aesthetics.
Thanks again for your help.
Jeff
Hi Jeff, There is no particular reason not to use a single chassis. It can save a few $$ on parts and depending on the rest of your system not cause any significant degradation of the separation. That is probably the strong point of mono-blocks. Another plus with them is that you can place them fairly near the speakers and reduce the length and effects of the speaker wire. The Hammond may be OK. The 230-0-230 is fine. My question is at how much current. You might be right at the limit for the transformer. With the Hammond you would expect about 325 B+ using solid-state rectifiers and capacitor input. If you keep the current at 180 mA per pair it will most likely be OK. The amount available ought to be at least 1.5 times the max-expected drain. If the trannie can deliver 500 mA everything should be fine. The only voltages that would need to be changed are the resistors in the power supply that feed the driver stage. When you decide what power trannie you will use and such I'll do some number crunching for you on the values.
I gave up on the FET CCS. Mostly because I could not get a reliable one that had a lower voltage drop than the LM317s. I am always open for ideas on it though. The 25 ohm pot is from (sad to say) Radio Shack part number 271-0265. There does not seem to be an alternate source for them. we have found 20 ohm ones at 1/2 watt and they would do for EL84s and similar sized tubes, but not for the bigger guys. It is feasible to use a 12 position switch with 1 and 2 ohm resistors and get a close balance. I would do that in a pinch, but my feeling is that part of the excellent performance is due to closely matching the idle currents.
As for getting parts only from us, I'll confer with Rodney. I would not use different output trannies though. There are several tweaks to the circuit that are not obvious that make everything work together and improve the performance from a garden-variety amp to one at a much higher level. The output trannies are hard to beat for $$ and they are so underrated it is incredible. I really cannot guess how good (or bad) other trannies would sound in this circuit. Before tweaking the amps were pretty good. About 1% THD up to half power and around 2% at max. Bandwidth was ok, but it didn't show the above band sensitivity at around 70K. Much of this is related to the output trannies. All trannies have problems like this. So the tweaking brought the distortion down by a factor of 4 and removed the sensitivity at 70K. I am sure others could be tweaked, --- if you have a good signal generator, digital storage scope and distortion analyzer. Few have the equipment. A related issue is that the circuit with the components selected is very tolerant of tube types. I'm not sure that other outputs (and there are many good ones) would work as well. I switch back and forth between Gold Lion KT88s, JJ KT77s, JJ 6L6GCs and Sovtek 6L6WXT. All sound good.
My sense is that if you don't need the power of the KT88s the JJ 6L6GCs sound very similar and are 1/3 the price. You can always upgrade later. The only circuit change is the resistor in the CCS.
Good listening
Bruce
Additional Response:
Hi Jeff, in looking over my earlier answer, I realized that I missed part of your questions. This is regarding the 12 vs 6-volt tubes. The design was based originally on using 6SL7s or 12SL7s. Except for the heater rating they are identical. I wanted to use 12 volts for two reasons, one I was able to get 12 volt SMPS for dirt-cheap prices and use them for the heaters. They make great heater power sources. Second when I discarded the SMPS and wanted DC on the heaters it is easier to get 3 amps at 12 volts than 6 amps at 6 volts from most power transformers and filter circuits. So I stuck with 12 volts. Now additionally the 6/12SL7s are noted for having a very nice sound and pretty high gain. I tried the entire 12A_7 bunch and could not find one that sounded as good with enough gain. So I did a bit of research and came up wit the 5751s that are now in most of my amps. They have nearly the same gain as the 6/12SL7s and sound very nice (the ones I use are Phillips NOS). For lower gain circuits like for 6V6 and EL84 output tubes the ECC802S is the tube of choice. It is electrically like a 12AU7, but sounds a whole lot better. So really it doesn't matter what voltage you choose, there are tubes that will run on it. (both the ECC802S and 5751 will run on either 6 or 12)
Good listening
Bruce
Follow up question by me:
Bruce,
The circuitry is slightly different between the 2 different tubes on the 77 and 88 on the pre-side of the amp (i.e. some of the resistor and cap values are about 1/2 on one vs the other). I was wondering if I were to use a 5751 which of the 2 pre schematics would you use, the front end of the KT77 or the KT88 design? I didn't know if the values made that much difference in either case. The Hammond trafo I have should put out around 1 amp at 250v and 0.75 amps at 375v by my calculations. I actually found this thing on a junk pile and it was given to me. Similar ones I have researched in the industrial lines they make cost about $500! I am going to try it, I will let you know how it turns out. I want to make x2 separate chassis just because I always wanted an amp with a power umbilical and separate supply. This thing is too big to put with the amp I think, rather ugly too. It will be housed for aesthetics.
Thanks again for your help.
Jeff
Follow up response by Bruce:
Hi Jeff, The best front end to use is the one on the Oddblock project. It works equally well with either the KT77 or KT88s. You can use either the 5751 or 12/6SL7s (no changes required) I prefer the looks of the all octals myself. You can skip the B+ relay driver part (with the 6NO30) and just use a toggle switch to simplify the circuit. There are a number of refinements in each of the successive projects in the Oddwatt series so the latest one is the best of all. With the transformer you have (it should work fine) I would build the power supply along the lines of the Oddwatt 225 except for the heater part. Use the one in the Oddblock portion instead. Be sure to use the 2.2uf poly cap (a Solen in the project) between the junction of the resistors from the B+ to the heaters. I have found that the noise level (already quite good) of the amp is much improved with this single part.
Good listening
Bruce
Another Question by me:
Bruce,
I think I am going to go with all Edcor iron after all, including the power trafo. The trafos are cheap. I was wondering.... SMPS.... is this like a power supply for a laptop computer? I have a million of these (brother in law owns a computer shop). If that is what you are using then I am going to stick with your original design completely.
Thanks,
Jeff
Bruce’s response:
Hi Jeff, That is correct. If you look at the first Oddwatt project (with EL84s) you will see that I used one. For the KT88s the requirement is 3 amps at 12v. Some 3 amp SMPS don't like the initial load (it is probably more like 5 amps for a few seconds). I had some that blinked with a fault for the first 10-15 seconds and then the tubes warmed up enough for it to handle the load. If you have lots to choose from pick one that is rated 5 amps. If not try the 3s. Be sure (use an ohmmeter) that the AC mains on the SMPS are not connected in any way to the dc outputs (sparks and arcs if they do).
Good listening
Bruce
Another Question by me:
I have a 12.5 amp SMPS. It’s ok to have more than enough, right? I also have a 5 amp 16v SMPS, wonder if I can bring it down with a resistor? Do these SMPS handle that kind of arrangement (voltage drop w/ resistor)? Also, I was wondering what the dimensions of the Oddblock monoblock chassis were approximately. I am going to build mine with Koa wood and aluminum plating on the top. Tried to find some brass plate but it is really expensive! I ordered the trafos from Edcor today; they have a 2-3 week turnaround at the moment. Lastly, where might I find some 317HV regs?
Always appreciate your help. Just let me know if you need a break from all of the questions. I promise to leave you alone on holidays and Sundays
Sincerely,
Jeff
Hi Jeff, The best front end to use is the one on the Oddblock project. It works equally well with either the KT77 or KT88s. You can use either the 5751 or 12/6SL7s (no changes required) I prefer the looks of the all octals myself. You can skip the B+ relay driver part (with the 6NO30) and just use a toggle switch to simplify the circuit. There are a number of refinements in each of the successive projects in the Oddwatt series so the latest one is the best of all. With the transformer you have (it should work fine) I would build the power supply along the lines of the Oddwatt 225 except for the heater part. Use the one in the Oddblock portion instead. Be sure to use the 2.2uf poly cap (a Solen in the project) between the junction of the resistors from the B+ to the heaters. I have found that the noise level (already quite good) of the amp is much improved with this single part.
Good listening
Bruce
Another Question by me:
Bruce,
I think I am going to go with all Edcor iron after all, including the power trafo. The trafos are cheap. I was wondering.... SMPS.... is this like a power supply for a laptop computer? I have a million of these (brother in law owns a computer shop). If that is what you are using then I am going to stick with your original design completely.
Thanks,
Jeff
Bruce’s response:
Hi Jeff, That is correct. If you look at the first Oddwatt project (with EL84s) you will see that I used one. For the KT88s the requirement is 3 amps at 12v. Some 3 amp SMPS don't like the initial load (it is probably more like 5 amps for a few seconds). I had some that blinked with a fault for the first 10-15 seconds and then the tubes warmed up enough for it to handle the load. If you have lots to choose from pick one that is rated 5 amps. If not try the 3s. Be sure (use an ohmmeter) that the AC mains on the SMPS are not connected in any way to the dc outputs (sparks and arcs if they do).
Good listening
Bruce
Another Question by me:
I have a 12.5 amp SMPS. It’s ok to have more than enough, right? I also have a 5 amp 16v SMPS, wonder if I can bring it down with a resistor? Do these SMPS handle that kind of arrangement (voltage drop w/ resistor)? Also, I was wondering what the dimensions of the Oddblock monoblock chassis were approximately. I am going to build mine with Koa wood and aluminum plating on the top. Tried to find some brass plate but it is really expensive! I ordered the trafos from Edcor today; they have a 2-3 week turnaround at the moment. Lastly, where might I find some 317HV regs?
Always appreciate your help. Just let me know if you need a break from all of the questions. I promise to leave you alone on holidays and Sundays
Sincerely,
Jeff
Bruce’s response:
Hi Jeff, Questions are always OK. A 12.5 should be OK. If need be a single solid state rectifier (5 amps or so at anything over 50v) can be put in series with the output to bring it down a bit if it actually delivers more than 12.0 volts. More current is great. Each diode has a drop of between 0.6 and 0.7 volts. We have found that some tubes other than the ones I usually recommend have thermal problems and the power used by their heaters vary over time and this in turn throws off the ideal current. At 12.0 volts most seem to behave. The JJs and Gold Lions seem fine up to nearly 13.0
The LM317HV are available from Jameco (on the web). Part number 192276 (at 2.25 each) or if unavailable 837927 (at 2.49 ea).
The chassis size for each of the mono blocks is 8 X 12 X 2. 2.5 or 3 inches deep make it a bit easier to build though. A little better air flow as well.
Good listening
Bruce
Another question by me:
Hey,
I was looking through the article you wrote on the Oddwatt again. At the end you mentioned getting a goal of 50-60watts would be attainable. If I wanted to try this would it be a matter of getting larger OPTs with higher input wattage ratings (i.e.50 or 75) and then just adjusting the current up on a pair of KT88s. Or would there be any other substantial modifications needed in the circuit?
Jeff
P.S. Maybe I should start another thread with that question and let everyone read this conversation.
Thanks,
Jeff
Bruce’s response:
Hi Jeff, There are several modifications needed to get the extra power. In theory you can get between 50-60 watts with a push-pull pair of KT88s. I have some preliminary design on that. But, the problem is one of cost and whether it can stay properly balanced. I find the output of the 6L6s quite adequate for most anything and I only rate them at 15 watts. They actually go to 20 before serious distortion sets in. The KT88s I rate at 25 watt and will go to 33. I have used them with several speaker systems and have never found any need for more power. I have several other amps (mostly SS) that go as high as 160 watts per channel and really have never needed that much power. They all now gather dust in the shed.
Good Listening
Bruce
Final Question by me:
Thanks again for the reply. Ok, I just ordered the OPTs and Power trafos from Edcor....only $200! Not too bad! Had a few questions on the specs of the resistors and caps in the amp circuit. the 220uF electrolytic cap connected to the cathode of the 12SL7, what rating in volts and quality do you recommend here? I happen to have a couple 16V Black Gates but not sure of the voltage. Also, the rest of the amplifier circuit, what resistor rating...1/4, 1/2, 1 watt? Was thinking of using all metal films. There are x2 47K resistors in the PSU filter array, should these be 5 watt as well? Was thinking of using a 1000pF silver mica and for the 0.33 PIO, I was going to get the Russian ones...any objections? I was thinking of going with all Roderstein Resista metal films from Percy Audio. I found some high current Jameco 460uH inductors as well at Jameco. Lastly, what gauge of wire to you recommend for the general wiring of the amp?
I think that is enough questions for now. Would you be willing to let me open a new thread at DIY audio and let you answer all of my questions there. If so, I will post this on a new thread Named Oddwatt Construction - general questions and advise.
Let me know, I just thought our conversation may be useful to others and a quick resource to refer future newbies like me to should you get another full of questions.
Thanks again!
Jeff
Hi Jeff, Questions are always OK. A 12.5 should be OK. If need be a single solid state rectifier (5 amps or so at anything over 50v) can be put in series with the output to bring it down a bit if it actually delivers more than 12.0 volts. More current is great. Each diode has a drop of between 0.6 and 0.7 volts. We have found that some tubes other than the ones I usually recommend have thermal problems and the power used by their heaters vary over time and this in turn throws off the ideal current. At 12.0 volts most seem to behave. The JJs and Gold Lions seem fine up to nearly 13.0
The LM317HV are available from Jameco (on the web). Part number 192276 (at 2.25 each) or if unavailable 837927 (at 2.49 ea).
The chassis size for each of the mono blocks is 8 X 12 X 2. 2.5 or 3 inches deep make it a bit easier to build though. A little better air flow as well.
Good listening
Bruce
Another question by me:
Hey,
I was looking through the article you wrote on the Oddwatt again. At the end you mentioned getting a goal of 50-60watts would be attainable. If I wanted to try this would it be a matter of getting larger OPTs with higher input wattage ratings (i.e.50 or 75) and then just adjusting the current up on a pair of KT88s. Or would there be any other substantial modifications needed in the circuit?
Jeff
P.S. Maybe I should start another thread with that question and let everyone read this conversation.
Thanks,
Jeff
Bruce’s response:
Hi Jeff, There are several modifications needed to get the extra power. In theory you can get between 50-60 watts with a push-pull pair of KT88s. I have some preliminary design on that. But, the problem is one of cost and whether it can stay properly balanced. I find the output of the 6L6s quite adequate for most anything and I only rate them at 15 watts. They actually go to 20 before serious distortion sets in. The KT88s I rate at 25 watt and will go to 33. I have used them with several speaker systems and have never found any need for more power. I have several other amps (mostly SS) that go as high as 160 watts per channel and really have never needed that much power. They all now gather dust in the shed.
Good Listening
Bruce
Final Question by me:
Thanks again for the reply. Ok, I just ordered the OPTs and Power trafos from Edcor....only $200! Not too bad! Had a few questions on the specs of the resistors and caps in the amp circuit. the 220uF electrolytic cap connected to the cathode of the 12SL7, what rating in volts and quality do you recommend here? I happen to have a couple 16V Black Gates but not sure of the voltage. Also, the rest of the amplifier circuit, what resistor rating...1/4, 1/2, 1 watt? Was thinking of using all metal films. There are x2 47K resistors in the PSU filter array, should these be 5 watt as well? Was thinking of using a 1000pF silver mica and for the 0.33 PIO, I was going to get the Russian ones...any objections? I was thinking of going with all Roderstein Resista metal films from Percy Audio. I found some high current Jameco 460uH inductors as well at Jameco. Lastly, what gauge of wire to you recommend for the general wiring of the amp?
I think that is enough questions for now. Would you be willing to let me open a new thread at DIY audio and let you answer all of my questions there. If so, I will post this on a new thread Named Oddwatt Construction - general questions and advise.
Let me know, I just thought our conversation may be useful to others and a quick resource to refer future newbies like me to should you get another full of questions.
Thanks again!
Jeff
Bruce’s last response to me:
Hi Jeff, Good questions, and good idea. I'll answer the questions tomorrow. I'm pooped tonight, just got the Poddwatt amp finished and it checks out fine. Tomorrow I will do some listening to it. I believe your idea of a new thread is good and will start one tomorrow as well.
Good listening
Bruce
And now Bruce will continue from here.
Thanks again Bruce!
Sincerely,
Jeff
Hi Jeff, Good questions, and good idea. I'll answer the questions tomorrow. I'm pooped tonight, just got the Poddwatt amp finished and it checks out fine. Tomorrow I will do some listening to it. I believe your idea of a new thread is good and will start one tomorrow as well.
Good listening
Bruce
And now Bruce will continue from here.
Thanks again Bruce!
Sincerely,
Jeff
More QuestionS
Bruce<
Here is a recap of a few questions That I Haven't been answered yet as well as a couple of others.
The specs of the resistors and caps in the amp circuit....the 220uF electrolytic cap connected to the cathode of the 12SL7, what rating in volts and quality do you recommend here? Also, the rest of the amplifier circuit, what resistor rating...1/4, 1/2, 1 watt? Was thinking of using all metal films. There are x2 47K resistors in the PSU filter array, should these be 5 watt as well? Was thinking of using a 1000pF silver mica and for the 0.33 PIO, I was going to get the Russian ones...any objections? I was thinking of going with all Roderstein Resista metal films from Percy Audio. I found some high current Jameco 460uH inductors as well at Jameco but I am unsure if these are good for here. I was wondering exactly what you used here. Lastly, what gauge of wire to you recommend for the general wiring of the amp? I purchased some 28ga 99.9999% silver wire a few months back and have been looking to use it somewhere. Any other specifics you might reccommend I would appreciate as well. I am going to prepare a parts list for this amp soon and post it as well.
Thanks again.
Jeff
Bruce<
Here is a recap of a few questions That I Haven't been answered yet as well as a couple of others.
The specs of the resistors and caps in the amp circuit....the 220uF electrolytic cap connected to the cathode of the 12SL7, what rating in volts and quality do you recommend here? Also, the rest of the amplifier circuit, what resistor rating...1/4, 1/2, 1 watt? Was thinking of using all metal films. There are x2 47K resistors in the PSU filter array, should these be 5 watt as well? Was thinking of using a 1000pF silver mica and for the 0.33 PIO, I was going to get the Russian ones...any objections? I was thinking of going with all Roderstein Resista metal films from Percy Audio. I found some high current Jameco 460uH inductors as well at Jameco but I am unsure if these are good for here. I was wondering exactly what you used here. Lastly, what gauge of wire to you recommend for the general wiring of the amp? I purchased some 28ga 99.9999% silver wire a few months back and have been looking to use it somewhere. Any other specifics you might reccommend I would appreciate as well. I am going to prepare a parts list for this amp soon and post it as well.
Thanks again.
Jeff
Hi, In answer to the earlier questions. The electrolytics in the SRPP cathodes ought to be reasonable quality, I used Elna ones, BGs are fine. Nichecons and Panasonic are as well. The voltage rating is minimal as there is less than 5 volts across them. Most resistors in the amps are 1 watt unless otherwise specified. I built one set with metal film and one with carbon film and really can't hear any difference because of the change. Other parts (namely use of the Auricaps vs K40 PIO) make more of a difference. The 47K resistors in the PS string are 1 watt as well. The current flow is very small. Each SRPP uses only about 1.5 ma. I used silver mica 1000pf cps in the feedback circuit in my amps. A good (WIMA, IA, Auricap, Sprague, etc) poly could go there as well. It needs to be stable and accurate in spec. Voltage rating is nil (1-2 volts across it - the ones I used were 100v). The amps use either Auricaps or K40 series PIO caps for the .33 uf coupling cap. Use ones rated at 400 volts there. There are some unlikely ways that most of the B+ could end up across them for a short period of time. The 460 uh inductors are internal to the AC mains line filter. There are other line filters that will work that have slight variations on values. Pretty much any will do if they are rated for at least 3 amps (most are rated quite a bit more). For general wiring I use 20 guage solid wire. For wires that carry large amounts of current, I use 18 guage. I would not go below 22 guage for any part of the amp. Additionally, be sure to use shielded wire on the inputs. Depending on how you configure the inputs and volume controls, beware of creating ground loops. I like to have the grounds come together at the volume control. It seems to minimize hum and noise pick up. Also be sure not to have a direct connection between the signal ground and the chassis and ac mains grounds.
Good listening
Bruce
Good listening
Bruce
in order...
and with deference to the designer:
1) a 221 ohm resistor is within the range of a 10% range 200 ohm resistor. Well, 1 ohm out, but hey.... Given thermal drift, manufacturer's tolerances etc, I wouldn't be losing any sleep.
2) If you string two caps in series, the voltage capability doubles. The capacitance doesn't change. Hence, your 2 caps in series caps will go to 330uF @ 900VDC.
Enjoy - I sincerely admire the "oddjob" concept!
and with deference to the designer:
1) a 221 ohm resistor is within the range of a 10% range 200 ohm resistor. Well, 1 ohm out, but hey.... Given thermal drift, manufacturer's tolerances etc, I wouldn't be losing any sleep.
2) If you string two caps in series, the voltage capability doubles. The capacitance doesn't change. Hence, your 2 caps in series caps will go to 330uF @ 900VDC.
Enjoy - I sincerely admire the "oddjob" concept!
inteeeereeestiiiiing!!!
I was once told by running caps in series you halved the capacitance and doubled the voltage...it never did make since to me (Learned it on a forum))...thanks for that clarification. I always just took it at face value but wondered why. Can you tell me then how to calculate final capacitance when running two different values in series? I don't want to get on too much of a tangent here but since we are there now....
Jeff
I was once told by running caps in series you halved the capacitance and doubled the voltage...it never did make since to me (Learned it on a forum))...thanks for that clarification. I always just took it at face value but wondered why. Can you tell me then how to calculate final capacitance when running two different values in series? I don't want to get on too much of a tangent here but since we are there now....
Jeff
According to NEETS the final capacitance when running 2 capacitors in series is calculated much the same way you calculate two different (or same) values of resistors running in parallel.....so according to them it would drop in halve. They give a very good explanation of this and why it happens here:
http://www.tpub.com/neets/book2/3e.htm
I will leave it at this.
Jeff
http://www.tpub.com/neets/book2/3e.htm
I will leave it at this.
Jeff
Hi, Thanks for the kind words. The design is deceptively simple. With minimal parts count, and careful selection of components it delivers a far better sound that I would have ever expected. You should hear these guys on large full range speakers and electrostatics (Martin Logans according to one high end store during a demo were awesome). I generally use some rather updated and modded Altec Lansing A7-500s.
Putting the caps in series will halve the value and double the voltage rating.... if you put bleeder resistors in parallel with each one. If not well sparks arcs and smoke will follow as the voltage distribution between them will not be equal. You could use larger value caps in the B+ filter chain. The values chosen were selected to make the B+ very clean. More is not cost effective if you have to buy the parts new. With the SS rectifiers in place and the internal impedance of the power trannie no damage will occur if you do increase the values (don't do this on tube rectifiers). Rifa caps tend to be rather sturdy and well made. With the actual B+ at 450 and depending on your line voltage just a bit more, they will probably be OK to use as is. Certainly in the second stage of the filter. It will be right at the limit and if you use a different power transformer there could be problems. Personally, I prefer to have at least a 50 volt margin on the filters. Even with a built in 30 second delay it is possible to go past 450 volts while the tubes are warming up.
A 200 ohm resistor could be used in the place of the 220. I usually use 5% tolerance resistors in the builds. My experience recently that if you stick to quality sources for the resistors they are often within 1%.
Putting the caps in series will halve the value and double the voltage rating.... if you put bleeder resistors in parallel with each one. If not well sparks arcs and smoke will follow as the voltage distribution between them will not be equal. You could use larger value caps in the B+ filter chain. The values chosen were selected to make the B+ very clean. More is not cost effective if you have to buy the parts new. With the SS rectifiers in place and the internal impedance of the power trannie no damage will occur if you do increase the values (don't do this on tube rectifiers). Rifa caps tend to be rather sturdy and well made. With the actual B+ at 450 and depending on your line voltage just a bit more, they will probably be OK to use as is. Certainly in the second stage of the filter. It will be right at the limit and if you use a different power transformer there could be problems. Personally, I prefer to have at least a 50 volt margin on the filters. Even with a built in 30 second delay it is possible to go past 450 volts while the tubes are warming up.
A 200 ohm resistor could be used in the place of the 220. I usually use 5% tolerance resistors in the builds. My experience recently that if you stick to quality sources for the resistors they are often within 1%.
A bleeder resistor....what, about 1M or higher? You mean one resistor per cap with the resistor tied across the +/- poles of each cap, correct? I also have two humongous Solen 82uF 603v metal/poly caps I could parallel with 18 more uF's to bring them up to 100uF. I have always wanted to use these somewhere. I have heard that polys work pretty good in power supplies but generally are not used here b/c of cost. I won them in a box of goodies on eBay and probably paid about $10 for the pair. Would you use these in the PSU filter caps? I have seen where you put 10uF polys (Solen to be precise) in the S5 mods you have done.
Thanks,
Jeff
Thanks,
Jeff
Hi, The 82uf ones may be fine as is. Polys are usually too large physically and too costly to use in this application. I would give them a try. If the amps are quiet, then fine if not add the extra capacitance. If you have the ability to measure the hum at the speaker terminals with and 8 ohm load and the inputs shorted, it should be under 1 millivolt if all is well. If you don't have the capability, you can try the unscientific ear test. With the inputs shorted there should be no hum or noise with your ear right at the speaker grill. Mine are 93 db/w and they are quiet.
I do specify polys in some portions of the circuits as they outperform electrolytics, particularly where there is high frequency noise involved.
The resistors should be about 47K and 1 watt for paralleling the caps. Hooked up just like you mentioned in the post.
I do specify polys in some portions of the circuits as they outperform electrolytics, particularly where there is high frequency noise involved.
The resistors should be about 47K and 1 watt for paralleling the caps. Hooked up just like you mentioned in the post.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.