tubelab.com said:
Yeah Baby! I have some 1kVA toroid supply transformers and it is true, this unit is equal in size!
807's...using eight I thought I was being conservative.
In the end all I will pull out of it will be around 60 watts. Kind of a waste but not really. The power supply is going to cost me big $$. Who would make a transformer to deliver 12 amps of 6.3VAC and 700 milliamps of 400-0-400VAC RMS? Of course I can use multiple transformers to get the end result but if the option to use one unit is there, I would like to price it out. Now I have to fix my Hikock Power supply and fire up some tubes...
Cheers,
Shawn.
Use two of the 800 vct Antek toroids available on Ebay. Item number 270066718873. 800 VCT 500mA and two 6.3 volt 5 amp windings. I have one of these powering a SE amp. I am going to borrow it for prototyping this monster.
My first experiment will be one channel, 4 X 6LW6 sweep tubes in screen drive. I want this one for the baseline. All other experiments will be comared to this one. I am guessing that it will make about 300 watts. This transformer could do 1KW easy at guitar frequencies. That would need 20 807's per channel!
I should be ready to flip the switch on this project tomorrow.
My first experiment will be one channel, 4 X 6LW6 sweep tubes in screen drive. I want this one for the baseline. All other experiments will be comared to this one. I am guessing that it will make about 300 watts. This transformer could do 1KW easy at guitar frequencies. That would need 20 807's per channel!
I should be ready to flip the switch on this project tomorrow.
Toroid Covers
I've been looking at covers on ebay to cosmetically cover & sheild the toroids on the top of the chassis. I can't find any to cover this big SOB! I'd need covers to fit over 7.5 inches wide(190mm) by 3.5 inches high(90mm) to allow room for the leads without pinching them.
Of course I could make my own but I'd much rather skip the case work and get to building.
Any tips where to purchase toroid covers?
Cheerio,
Shawn.
I've been looking at covers on ebay to cosmetically cover & sheild the toroids on the top of the chassis. I can't find any to cover this big SOB! I'd need covers to fit over 7.5 inches wide(190mm) by 3.5 inches high(90mm) to allow room for the leads without pinching them.
Of course I could make my own but I'd much rather skip the case work and get to building.
Any tips where to purchase toroid covers?
Cheerio,
Shawn.
I don't have that part figured out yet. I have mounted power toroids on top of each other before, but I haven't tried it with OPT's yet, or any transformers of this magnitude yet.
I haven't built a tube amp of this magnitude in about 10 years. The previous ones were guitar amps made with power transformers used for OPT's, and distortion was the norm. I have had a long time to dream up ways that I would do it better. I briefly experimented with screen drive a while ago, and decided that I would like to try it again, but had no plans to build anything for lack of an OPT. Then I read this thread, found out about the Plitron, and the snowball started rolling downhill. I will probably try out several different circuits, and maybe some different output tubes until I get what I want, and then the real building will start. In many respects Shawn has a head start, since he already knows what he wants.
At this point it is just an ugly breadboard. I have developed a solderless breadboarding system over the past few years that lets me turn ideas into reality rather quickly. I can usually turn a dream into an amp in a few hours. This one has been slowed down due to some head scratching caused by a bad scope probe and an open clip lead. I didn't have a "module" for 4 big sweep tubes either, so I am making a new one.
I enclosed a picture of the previous 80 watt screen drive amp on the breadboard. I have reconstructed it with a few changes. The driver is up and running making over 300 volts P-P. The output stage is still under construction.
Attachments
I said that I would flip the switch today, and I did. The good news is that nothing blew up. I guess I wasn't thinking too much when I hooked it all up to my bench power supply. My power supply goes up to 555 volts, at 250 mA. Thats 138 watts. Somehow I was expecting it to power a 300 watt amp. Well it really doesn't like it! I pulled out 2 of the output tubes, and wired the 8 ohm load to the 4 ohm tap. I can do some "reduced power" testing this way.
Picture of breadboarded amp:
Picture of breadboarded amp:
Attachments
I ran some basic testing on the low powered version. My power supply breaks into some weird oscillation when I try to draw over 400mA. Turning the drive up to the edge of this effect yields 180 watts. The hottest thing on the bench is the 500 watt load resistor. I have been reminded by people much smarter than me that this is not a "non inductive load", but I am not connecting my new load box up to this arc welder, so some measurements may be off.
I ran a frequency response sweep at 150 watts. LF 3 db point is somewhere below 20 Hz. The power supply freaks out whenever I try to go there! HF 3 db point is 55 KHz. Response at 50 watts is 6Hz to 55 KHz. Ruler flat from 20 - 20. Not bad for a $134 transformer. There appeared to be some saturation effects below 25Hz, but this could be my power supply expressing its distaste for my experiments or DC imbalance. This transformer will probably do DC to daylight with your lower powered design.
I have discovered some DC imbalances in my non adjustable drive circuit, so I am modifying the circuit to make it more adjustable. Some crossover distortion is evident at low power, and hum shows up when the power supply drops out of regulation (over 100 watts). I guess that I am going to steal the Antek toroid from its happy home in a SE amp and make it work for a living. Not to sure when all this will happen, got to go see mom next weekend. Have to make a 2000 mile road trip right after that, but I will hit the Dayton Hamfest.
Picture of the test bench. Scope and audio analyzer across the output. Audio analyzer set to display "watts into 8 ohms". It reads 163.8. Current meter on power supply is buried.
I ran a frequency response sweep at 150 watts. LF 3 db point is somewhere below 20 Hz. The power supply freaks out whenever I try to go there! HF 3 db point is 55 KHz. Response at 50 watts is 6Hz to 55 KHz. Ruler flat from 20 - 20. Not bad for a $134 transformer. There appeared to be some saturation effects below 25Hz, but this could be my power supply expressing its distaste for my experiments or DC imbalance. This transformer will probably do DC to daylight with your lower powered design.
I have discovered some DC imbalances in my non adjustable drive circuit, so I am modifying the circuit to make it more adjustable. Some crossover distortion is evident at low power, and hum shows up when the power supply drops out of regulation (over 100 watts). I guess that I am going to steal the Antek toroid from its happy home in a SE amp and make it work for a living. Not to sure when all this will happen, got to go see mom next weekend. Have to make a 2000 mile road trip right after that, but I will hit the Dayton Hamfest.
Picture of the test bench. Scope and audio analyzer across the output. Audio analyzer set to display "watts into 8 ohms". It reads 163.8. Current meter on power supply is buried.
Attachments
tubelab.com said:
Absolutely. They will have a very easy lifestyle in my 807 set up.
Hmm... I think I should have some T-shirts made with a tube logo and the text: "DC to Daylight"
Morgan Jones came over to my house for a visit! Some of the valves beside my computer swarmmed him like a Hollywood movie star!
Cheers,
Shawn.
Attachments
You didnt get the very last, it was up yesterday, but Im sure its no coincidence.
Maybe this can make crude substitute? Comes to 1550 ohms P-P with 8 ohm load.
Maybe this can make crude substitute? Comes to 1550 ohms P-P with 8 ohm load.
Brazilian output transformer
Hello everybody,
This is my first adventure posting through these wonderful topics. I couldn't help but write to some 807 admirers experts, specially after I accidentally found (and eventually bought) a nice Brazilian made output transformer which suddenly became the foundation for my next project. It is a bold project for me because I have only built small single ended EL84 amps up to now.
I hope that the reasons why I came to this specific thread will clear up now. It is because I noticed you are into projects for lower primary impedances. I have some KT66 tubes at hand, but my OPT specifies 4300 ohms instead of the more usual 6600 ohms primary impedance. Probably for KT88 but I have none at hand. UL taps are 33%, and nominal power is 50W.
I can add more specs for the OPT, if someone tells me which are important. Of course I must say that I am inclined to use 807 in PP configuration (not sure about class AB1 or AB2) instead of the KT66. 6L6GA are also available, but I see no advantage. Also some sweep tubes, which may be even worse (no UL or triode mode feasible because of low screen grid voltage allowed).
In a few words - does the 4300 ohms spec at 50W ring a bell for anybody? Project suggestions most welcome! Thank you folks!
Flavio, from Brazil.
Hello everybody,
This is my first adventure posting through these wonderful topics. I couldn't help but write to some 807 admirers experts, specially after I accidentally found (and eventually bought) a nice Brazilian made output transformer which suddenly became the foundation for my next project. It is a bold project for me because I have only built small single ended EL84 amps up to now.
I hope that the reasons why I came to this specific thread will clear up now. It is because I noticed you are into projects for lower primary impedances. I have some KT66 tubes at hand, but my OPT specifies 4300 ohms instead of the more usual 6600 ohms primary impedance. Probably for KT88 but I have none at hand. UL taps are 33%, and nominal power is 50W.
I can add more specs for the OPT, if someone tells me which are important. Of course I must say that I am inclined to use 807 in PP configuration (not sure about class AB1 or AB2) instead of the KT66. 6L6GA are also available, but I see no advantage. Also some sweep tubes, which may be even worse (no UL or triode mode feasible because of low screen grid voltage allowed).
In a few words - does the 4300 ohms spec at 50W ring a bell for anybody? Project suggestions most welcome! Thank you folks!
Flavio, from Brazil.
Re: Brazilian output transformer
There is a possibility, in that one of the Class AB2 operations of the 807 comes close. Take a look at: STC 807 Application Report. That one does, however, give 75W, so you won't be able to fully crank it if you use the 50W OPT.
Other than that, 807s tend to like higher Rl's.
fbetti said:
There is a possibility, in that one of the Class AB2 operations of the 807 comes close. Take a look at: STC 807 Application Report. That one does, however, give 75W, so you won't be able to fully crank it if you use the 50W OPT.
Other than that, 807s tend to like higher Rl's.
Nice folks,
Thanks to Miles, I could find the compendium on the 807s. I also fear that they will like somewhat higher load impedances. Perhaps one can use them as triodes? Should that bring plate resistance down? And also preferred plate load?
In the meanwhile, a few discoveries of mine. I received my copy of Kevin O'Connors book (POP) that was praise on these columns. There are sample projects for both 1650P and 1650N Hammond outputs - respectively 6600 and (surprise!) 4300 ohms at 60W - almost like the Brazilian "Willkason 4704" rated for 50W (another difference lies in the UL screen taps - 33% instead of Hammond's 40%). Both amp projects use 6550 PP pairs, which can seemingly be replaced unilaterally by KT88s, which are even stronger. Sadly, I haven't any of both at hand.
Also visited the wise pages from Mr. White <http://www.cosmos2000.org/audio/> remebering us that Williamsons could never be Williamsons without the brute force Partridges projected for 2Hz on. Now I wonder how many more watts could a 30W OPT (at 20Hz or 30Hz) deliver, were its lower frequency raised to some 80Hz or 90Hz... Not hi-fi anyway, and probably a call for some sharp LPF... Also difficulties (impossibilities?) about overall NFB... Please do tell me if my thoughts are way wrong.
Back to the tubes choice for my 4300 ohms OPT: might KT66s also work if triode connected? Not a good choice according to Mr. White's words, because they are beam tubes, not true pentodes... Disregarding such issue, I might also consider using my 6BG6Gs (much like 6L6s, though), or even EL34 (I've got a pair from East Germany).
Not many choices left with what I got at hand... One last thought: what if I wired 16 ohms speakers to the 8 ohm secondary? Would that mean double (8600 ohms) reflected primary impedance? In such case, would I achieve only half the power (25W in such case)? Things are getting confusing for me now... Any clues welcome! Regards!
Flavio
Thanks to Miles, I could find the compendium on the 807s. I also fear that they will like somewhat higher load impedances. Perhaps one can use them as triodes? Should that bring plate resistance down? And also preferred plate load?
In the meanwhile, a few discoveries of mine. I received my copy of Kevin O'Connors book (POP) that was praise on these columns. There are sample projects for both 1650P and 1650N Hammond outputs - respectively 6600 and (surprise!) 4300 ohms at 60W - almost like the Brazilian "Willkason 4704" rated for 50W (another difference lies in the UL screen taps - 33% instead of Hammond's 40%). Both amp projects use 6550 PP pairs, which can seemingly be replaced unilaterally by KT88s, which are even stronger. Sadly, I haven't any of both at hand.
Also visited the wise pages from Mr. White <http://www.cosmos2000.org/audio/> remebering us that Williamsons could never be Williamsons without the brute force Partridges projected for 2Hz on. Now I wonder how many more watts could a 30W OPT (at 20Hz or 30Hz) deliver, were its lower frequency raised to some 80Hz or 90Hz... Not hi-fi anyway, and probably a call for some sharp LPF... Also difficulties (impossibilities?) about overall NFB... Please do tell me if my thoughts are way wrong.
Back to the tubes choice for my 4300 ohms OPT: might KT66s also work if triode connected? Not a good choice according to Mr. White's words, because they are beam tubes, not true pentodes... Disregarding such issue, I might also consider using my 6BG6Gs (much like 6L6s, though), or even EL34 (I've got a pair from East Germany).
Not many choices left with what I got at hand... One last thought: what if I wired 16 ohms speakers to the 8 ohm secondary? Would that mean double (8600 ohms) reflected primary impedance? In such case, would I achieve only half the power (25W in such case)? Things are getting confusing for me now... Any clues welcome! Regards!
Flavio
Hmmm... 6080's... Interesting. It recalled me some Soviet made 6AS7G's I've got around. I once took a look at their curves... They look very hard to cutoff - probably a difficult drive situation... I've seen OTL projects around them.
This class of tube suddenly recalled me some 12E1's that I've got at hand. Not a very spread tube I guess... They look hefty with their 35W plate dissipation. One may take a look at them at http://tdsl.duncanamps.com/show.php?des=12E1
Nevertheless, those are still beam tubes, not true pentodes... Is that of any importance at all? Old 6F6's were true pentodes, and I can't find any projects using them. I've read somewhere else that most EL84-like tubes are actually beam tetrodes, despite of the original Mullards being pentodes...
As for the transformer, on second thought I imagined that if I'm going to fool with different impedances (wiring a 16 ohms speaker to the 8 ohms secondary, for instance), nominal power could theoretically be attained, but in practice internal losses (copper, maybe iron too) will raise - in other words, efficiency may suffer - am I in the right track?
Thank you folks, I've got a lot to learn with you!
This class of tube suddenly recalled me some 12E1's that I've got at hand. Not a very spread tube I guess... They look hefty with their 35W plate dissipation. One may take a look at them at http://tdsl.duncanamps.com/show.php?des=12E1
Nevertheless, those are still beam tubes, not true pentodes... Is that of any importance at all? Old 6F6's were true pentodes, and I can't find any projects using them. I've read somewhere else that most EL84-like tubes are actually beam tetrodes, despite of the original Mullards being pentodes...
As for the transformer, on second thought I imagined that if I'm going to fool with different impedances (wiring a 16 ohms speaker to the 8 ohms secondary, for instance), nominal power could theoretically be attained, but in practice internal losses (copper, maybe iron too) will raise - in other words, efficiency may suffer - am I in the right track?
Thank you folks, I've got a lot to learn with you!
Won't anybody give the boy some good advice?Just give him some good advice!
Dear Flavio,
First of all. You can use the 6L6 very well with your 4300 ohm output transformers. Just add a little more negative feedback to compensate for the fact that they will be generating a bit more distortion at that 4300 0hm impendence.
Second: You can very easily connect an 8 ohm loud speaker to your 4 ohm tap to get 8600 ohms plate to plate for the 6L6 ot KT66 tubes and believe me they will love it!
KT66 with 500 volts in U L at 8600 ohms plate to plate will produce the cleanest and sweetest sound you can believe and with very little feedback if any at all.
At my company, we manufacture several amps using 6L6GC and KT66 into 4300 ohm plate to plate load with cathode feedback and , or lots of global negative feedback with wonderful results.
Experiment with various output tubes and you will have fun. You will not hurt anything.
Good luck my friend.
P.S. Your EL34 tubes will match the 4300 ohm xmrs perfectly as well. Dynaco used the magic 4300 number with everything just about.
CunninghamRadio
Dear Flavio,
First of all. You can use the 6L6 very well with your 4300 ohm output transformers. Just add a little more negative feedback to compensate for the fact that they will be generating a bit more distortion at that 4300 0hm impendence.
Second: You can very easily connect an 8 ohm loud speaker to your 4 ohm tap to get 8600 ohms plate to plate for the 6L6 ot KT66 tubes and believe me they will love it!
KT66 with 500 volts in U L at 8600 ohms plate to plate will produce the cleanest and sweetest sound you can believe and with very little feedback if any at all.
At my company, we manufacture several amps using 6L6GC and KT66 into 4300 ohm plate to plate load with cathode feedback and , or lots of global negative feedback with wonderful results.
Experiment with various output tubes and you will have fun. You will not hurt anything.
Good luck my friend.
P.S. Your EL34 tubes will match the 4300 ohm xmrs perfectly as well. Dynaco used the magic 4300 number with everything just about.
CunninghamRadio
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.