I built a 45 amp with a 6j5 (one triode equivalent of the 6sn7) driver two years ago. It really isn't neseserly with more then on of the triods in the 6sn7 to get enough gain to drive the 45. In my system I even can drive it straight from a cellphone. I did this by using an interstage transformer so I could use 250v/7ma on the tube. The same can be achieved by using an anode choke or an ccs. But the easiest is to disconnect the first triode and only use one in each bottle and bias it at 150v/3ma. Also a got operating point.
That's simply not true. Options of driver stages with their pros and cons have been the subject here multiple times in multiple threads (and no, I'm not pointing you to find them because it is a good mental exercise to do that yourself and part of the learning process).
Nowadays it seems normal to open a thread and ask instead of reading and learning first.
Thanks! I want to utilize a current production tube and will stick with the 6SN7.
I have spent considerable time searchIng the subject of drivers for the 45 with few results. The 6SN7 is rarely mentioned and then only vague, subjective terms. Likely the reasoning for driver selection is mundane enough for the experience builders as not to be worth mentioning. Too obvious. Not so obvious for me… Yet. Maybe someday. For now I’m tired of the searches, the videos and the books. Just want to get my hands, dirty a little.
Speaking of things that are obvious to the experienced builder…
Plate voltage. Is the manufacturer maximum based on plate to cathode or plate to ground?
My original understanding was is that it is across the tube, but multiple searches have turned up conflicting answers.
i’m thinking about bumping the plate voltage up a little and need to be certain.
Plate voltage. Is the manufacturer maximum based on plate to cathode or plate to ground?
My original understanding was is that it is across the tube, but multiple searches have turned up conflicting answers.
i’m thinking about bumping the plate voltage up a little and need to be certain.
Last edited:
Makes perfect sense put like that. Thank you.
A lot of bad information out there. Look hard enough and you can find the wrong answer to anything.
Rob7,
Two others and I did extensive research, then wrote an article about paralleling tubes.
"Parallel Tubes Effects", was the Cover Article of the very last issue of "Glass Audio", (Volume 12, Number 5, 2000).
We collected multiple types of measurements on twenty dual triode devices. We collated and analyzed 600 measurements.
Measurements included harmonic distortion to the 7th harmonic.
We did computer analysis of the tube and circuit.
We did a physical construction analysis.
The tube we used for the measurements was Very Linear.
You may have guessed it, it was the 6SN7.
Then, we needed to do a Double Blindfold Listening Test; to try and hear the difference with paralleling, and without paralleling.
The amplifier we chose to test was a two channel single ended amplifier, each channel with a 6N1P triode that drove a 300B triode.
A dual switch and umbilical cord controlled 2 relays, to parallel the 6N1P or not, and to parallel the 300B or not.
Parallel None; Parallel 6N1P; Parallel 300B; Parallel Both 6N1P and 300B.
The double blindfold listening was done in 3 different venues, with 3 different type loudspeaker pairs.
Nobody complained about the sound, paralleled, and not paralleled.
All but one did not hear any difference, the one heard a very small difference.
The article gave many hints and rules for a method on how to properly design an amplifier for parallel tubes. The listening test amplifier followed those rules.
We never received any complaints about the research we did.
I am still waiting for someone to do their own research, including using the hints and rules, and prove that the original research was faulty.
At one of the Vacuum State of the Art Conference (VSAC) in Silverdale, Washington, I presented the essence of the article.
I received No complaints or contrary comments there.
Just my opinions.
Two others and I did extensive research, then wrote an article about paralleling tubes.
"Parallel Tubes Effects", was the Cover Article of the very last issue of "Glass Audio", (Volume 12, Number 5, 2000).
We collected multiple types of measurements on twenty dual triode devices. We collated and analyzed 600 measurements.
Measurements included harmonic distortion to the 7th harmonic.
We did computer analysis of the tube and circuit.
We did a physical construction analysis.
The tube we used for the measurements was Very Linear.
You may have guessed it, it was the 6SN7.
Then, we needed to do a Double Blindfold Listening Test; to try and hear the difference with paralleling, and without paralleling.
The amplifier we chose to test was a two channel single ended amplifier, each channel with a 6N1P triode that drove a 300B triode.
A dual switch and umbilical cord controlled 2 relays, to parallel the 6N1P or not, and to parallel the 300B or not.
Parallel None; Parallel 6N1P; Parallel 300B; Parallel Both 6N1P and 300B.
The double blindfold listening was done in 3 different venues, with 3 different type loudspeaker pairs.
Nobody complained about the sound, paralleled, and not paralleled.
All but one did not hear any difference, the one heard a very small difference.
The article gave many hints and rules for a method on how to properly design an amplifier for parallel tubes. The listening test amplifier followed those rules.
We never received any complaints about the research we did.
I am still waiting for someone to do their own research, including using the hints and rules, and prove that the original research was faulty.
At one of the Vacuum State of the Art Conference (VSAC) in Silverdale, Washington, I presented the essence of the article.
I received No complaints or contrary comments there.
Just my opinions.
Just going back through the thread. Missed this post. The amp did “bite” me pretty hard.
I’ve had a few zaps by 120vac over the years. This was different.
Had been checking power supply voltages with only rectifier tube installed. Then been powered down for 15 minutes or so. I knew there was no bleeder resistor, yet I still got careless…and very lucky. A little embarrassing to admit but a good reminder to others and a lesson I won’t forget. It could’ve been much worse.
Bleeder resistor on the top of the list! Thank you for the reminder.
For a cost of about US$20 an additional DVM across the B+ is probably a good investment. My building partner, with a wife and a mortgage, is moving towards a whole panel of separate cheap DVMs to be attached before firing things up, and connections only touched after double-checking. Five analog plus a digital distortion analyzer, half a dozen scopes, etc. etc. are much, much less important equipment than stuff and procedures (!) keeping Iain staying on this side of the grass. Trust but verify.
All good fortune,
Chris
All good fortune,
Chris
Thanks! Good inspiration for phase two.…Building from scratch. Transformer coupling intrigues me. A lot to learn. What part of Norway? Family there. .
Is this posted online anywhere? Very cool stuff! I would love to read it…even if way over my head!
Hope you do! I’m sure many here would love to read it.
Not quite sure what to take away from the information in your post. Seems paralleling made no subjective difference to the listeners? What about the objective measurements… Did parallel measure better or worse?
Tinkering update.
Given the increasing cost and scarcity of 45 tubes my first priority was to make sure they were “safe” in these amps.
Design disputes aside, the mono blocks are quite pleasant to listen to and the tubes are coasting well below design specs—with one exception:
All filament/heater voltages were 10% or slightly more over spec. Too anal to let that go, especially with the 45s.
I added current sense resistors, one on each leg, to all.
Initially fired amp up with no tubes to measure “new” voltage. Was scratching my head for a moment… They were the same as before (with no tubes)
Rookie mistake. Of course, no effect without the load of the filaments.
Tubes in and voltages measured again. All voltages were within 2%. Success. My tubes should have a long and healthy life in these amps.
Well… there is that first cap in the power supply. That’s next…
Just having some fun!
Given the increasing cost and scarcity of 45 tubes my first priority was to make sure they were “safe” in these amps.
Design disputes aside, the mono blocks are quite pleasant to listen to and the tubes are coasting well below design specs—with one exception:
All filament/heater voltages were 10% or slightly more over spec. Too anal to let that go, especially with the 45s.
I added current sense resistors, one on each leg, to all.
Initially fired amp up with no tubes to measure “new” voltage. Was scratching my head for a moment… They were the same as before (with no tubes)
Rookie mistake. Of course, no effect without the load of the filaments.
Tubes in and voltages measured again. All voltages were within 2%. Success. My tubes should have a long and healthy life in these amps.
Well… there is that first cap in the power supply. That’s next…
Just having some fun!
Last edited:
Rob7,
A co-worker and I thought we could prove that parallel tubes do not perform very well (just do a few measurements, find some bad things like more harmonic distortion, etc.)
That was a mistake.
Just like Michelson and Morley who thought they could prove the existence of Ether throughout the universe (they ended up proving it did Not exist) . . . We expected to quickly prove our expectation of bad performing paralleled tubes.
The proof took a long time, study, and work (we proved that if you use proper techniques, paralleling tubes Do work Very Well).
Both the measurements, and the multiple listening tests proved that paralleling tubes can be very helpful to get more power, and yet sound good.
At one venue, a listener was so mad because he could not hear any difference between parallel and non-parallel,
that he threw the 2 switches and umbilical cord across the room.
Because we could not find a difference during the measurements, we had to add listening proof, math proof, and physical structure analysis proofs.
A co-worker and I thought we could prove that parallel tubes do not perform very well (just do a few measurements, find some bad things like more harmonic distortion, etc.)
That was a mistake.
Just like Michelson and Morley who thought they could prove the existence of Ether throughout the universe (they ended up proving it did Not exist) . . . We expected to quickly prove our expectation of bad performing paralleled tubes.
The proof took a long time, study, and work (we proved that if you use proper techniques, paralleling tubes Do work Very Well).
Both the measurements, and the multiple listening tests proved that paralleling tubes can be very helpful to get more power, and yet sound good.
At one venue, a listener was so mad because he could not hear any difference between parallel and non-parallel,
that he threw the 2 switches and umbilical cord across the room.
Because we could not find a difference during the measurements, we had to add listening proof, math proof, and physical structure analysis proofs.
Last edited: