Pentode AB1 design school???

[mvon21234]

A PROPOSED GROUP PROJECT?

Hello All,

My name is Martin, and I’ve been reading here for a while when I have free time. The level of knowledge on this group is phenomenal, and it’s so different from years ago when you could find one or two buddies to “play” with electronics! I’m not an engineer, but I am a bit more than a “tube jockey” – I could do much more than that, repairing TVs in a shop in the late 60s as a 14 y.o. So I know all about the various voltages, line isolations, and even the smell of burning flesh from having a finger too close to the plate cap of a horizontal output tube.

I’ve built tube amps in the past, most recently the 6Y6 Acrosound print, but with a different B+ supply (FWB, Pi filter) which I knew how to do through acquired practical knowledge. But I don’t really know how to DESIGN an amp, and that’s where I hope this group will become involved.

My proposal is for a community design school which will be breadboarded, and then serve as a springboard for further experimentation. It will not be particularly rushed, as wintertime is the time for basement activities. Now would be the time for unhurried thought about the matter.

My basic thought is for a high powered design that can be built on a budget, if one patiently hunts for power supply iron on Ebay.

Output tubes: 6146 or equivalent, AB1 pentode. (I can hear the groans now! I have read the “critiques” here.) Why? I have a lot of 6159s, and building a DC filament supply is no big deal. For others that would want to play, 12 volt versions sell pretty inexpensively and are easy to find. Edcor sells a 6.6k 100w transformer for about $100, and we can be in the ballpark of 6.6k at about 600v plate, and less than 200v screen on a 6146 – very far inside the safe operating area of these somewhat notorious tubes. Also, we can obtain the plate voltage fairly cheaply and easily with a standard machine tool 120-480v transformer, wired in reverse w/ a FWB. These babies can be found on ebay in the $20-$40 range regularly, and ultrafast diodes are cheap as well. Capacitors could be a problem in this voltage range pricewise, but an experienced builder should be able to properly insulate the top half of series-connected filters for safety issues.

The first of many things that I don’t know how to do is calculate the voltage swings necessary for the proper design of a drive stage – or input stages, for that matter. I keep reading, but it never really sinks in enough for me to go to a piece of paper and do it. For this part of the design, I would plead for the group’s help and patience – and, to turn it into a teaching reference of sorts for all.

What I have lots of, tubewise: 6/12sn7 6/12sl7 6/8fq7 5755 6j7 (why does no one like this tube?) 6ej7 12au7/5814 Again, by choosing alternate voltages, all of these are readily available NOS fairly cheaply. So a basic design using tubes similar to these would lead to an inexpensive amp. With instructions as to how the working parameters were achieved for each tube, all would then feel free to experiment with the likes of a D3a, a pair of 6gk5, or whatever. So it could be useful to design the input and driver stages in both triode and pentode forms – there must be a couple a million of NOS 6gh8 out there yet from TV days!

So, in a way, this could be the Eternal Project – it keeps changing in experimentation form! And, we could save screen drive for later.

But back to our cheap power supplies – for the rest of the amp, although I have a good supply of transformers here (along with an Eico regulated supply for testing use), I’m going to guess we can do an Ebay special with a 120-240v stepup, FWB. For the screen supply, which enjoys total control of the life of a 6146, I’m thinking of the Silas shunt regulator fed from the lower B+ supply.

I’ve thought about stacked supplies, and I don’t think that I’d have any issue, but I would worry about the safety of those behind me, or those that would read this thread and follow along. But if we did stacked supplies, it would be interesting to link the active regulation of the plate and screen supplies to see if we could duplicate the ultralinear effect without a special screen winding for these or horizontal output tubes.


Enough for now, and thank you for your time. Would you share your thoughts?

[airboss]

Kinda like this bunch eh? AX84.com - The Cooperative Tube Guitar Amp Project I like the idea, but don't particularly understand the puzzle of voltage swing from the driver. Pick an operating point on the load line, bias the tube for that OP then double the bias voltage (at full volume) for that OP. Manipulating gain from the driver is a different matter but the tables in the back of the tube manual will offer a lot of explanation.

Get a copy of Radiotron Designers Handbook. The third addition is the one I have, after the third they get really thick. Also, the front several chapters of any tube manual will explain a bunch.

Where this outfit leaves me in fits is the mixing of SS and HS in the same design. They use standard audio techniques that are just foreign to me. I have found in the week or so that I've been here, (more or less the same background as you) just reading every post in (say) the Baby Huey amp thread, or the heretics unity gain preamp thread will give you more than enough heartburn for now.

As far as 6146s in audio use all I can say is you never even saw them used as plate modulators back in the day when AM ruled the airwaves. They didn't put that plate cap on the top for fun, they put it up there because the tube wouldn't work above about 50mhz without it up there. 6550 or KT-88 seems so much more logical. But it is after all just a bottle full of electrons so why not? Take a look at the KT-77 from JJ too.

For basic tube design information look at the technical sections of this website: Aiken Amplification Once again, this is guitar amps. I don't know of anywhere other than here where there is smarts enough to do what you want. It's just getting the brain trust to go along with a project that (to them) is so elementary.

[aardvarkash10]

Its all been done before.... check out the AB2 thread that ChrisH started. It has a HUGE amount of great learning material in it.

As for load lines and B+ - get a big block of graph paper and start sketching. Its the only way.

[tubelab.com]

Its all been done before.... check out the AB2 thread that ChrisH started. It has a HUGE amount of great learning material in it.

The thread mentioned above details the development and design of an amplifier from the initial idea to two complete working amplifiers. The concepts developed in that thread can be applied to many different designs. Chris wanted an amplifier that used 6SN7's to drive 6L6GC's. I had no definite plans for my amp but I had a big box full of 6SN7's and wanted to try a whole bunch of output tubes. I have used my driver boards (2 6SN7's per channel) with output tubes from 45's to big sweep tubes, and yes I tried 6146's.

My breadboard amp has been used for a few experiments with more planned. It is currently back on my bench with 6SL7's in the first stage, 6SN7's in the second stage, and 6L6GC's in the output. More experiments to follow.

The 6146 has been used in audio. Unfortunately they got a bad reputation due to one particular amplifier, the first generation Ampeg SVT. This was a 300 watt bass guitar amp using 6 X 6146's. The amps had a poor reliability and there was a public meltdown of one on a televised Rolling Stones concert in the late 60's. The later vintage SVT's switched to 6550's with much better luck. Is the 6146 a poor choice? Probably not. Ampeg ran the screens far too hot inviting meltdown.

I have been reducing my collection and I recently gave away about 200 6159's so I have crossed them off my list (I still have some 6146's). This is not because of the Ampeg reports (yes I saw one blow up spectarularly in the 70's but it was a 6550 version), it is more a case of the fact that I have more projects than I have years of life left, so some have to go.

Back to the questions in hand. Either you think like me, or read my mind (or my web site), since I have been known to use reverse connected industrial transformers to generate HV. A 480 volt transformer will mkae a B+ of 550 to 600 volts. You already specify a 6600 ohm load. There is a whole bunch of theory involving load lines to determine the drive voltage, load point, and power output. Or, we can just read the 6146 data sheet. With a little interpolation it looks like you will get about 75 watts from a pair in AB1 with about 100 volts P-P of grid drive. Pushing things into AB2 may get about 100 watts with 130 volts or so of drive.

Why no deep theory? In reality, you will want your driver to be capable of delivering much more voltage so that it will remain in a linear range even at full power output. The gain can be approximated at about 50 assuming 2 volts P-P at the input if NO feedback is applied. A pentode amp will need some feedback, so plan on a gain of 200 to 400. How do you get there. That, is up to you. The amp in the above mentioned thread used cascaded LTP's with 6SN7's with a total gain of about 100 which is enough to run 6L6GC's with some feedback, but not enough to light up some big sweep tubes or the 6146's. Sticking a 'SL7 in the first stage provides enough gain. Other options (and one I am exploring) is a LTP using pentodes (the 6EJ7) and an LTP using triodes (6SN7 or 6FQ7) will provide plenty of gain. I haven't decided which stage to put first yet, so I am trying it both ways.

[mvon21234]

The messages to date more or less support my premise that the knowledge is indeed here and scattered about, but maybe I'm too rule oriented in my wishes. I know about load lines, and can draw them. Similarly, I mentioned the 6.6k xformer because I know that it's in the ballpark at the mentioned parameters from the GE manual. And I'm familiar w/ seat of the pants engineering simply by having been knocked there a few times. Maybe I've become too flow chart oriented, in that I was looking for a design rule format, if you will, with one big caveat: The REASONS behind each design rule, such as tube lab simply noting that the gain from the previous stage is made intentionally large to insure linearity. In my world, I'd put a signal generator and scope on a functioning tube amp or two and simply look to see what I needed to achieve. And eventually, after experimentally establishing the results, I would have interpolated parameters in the resistance coupled amplifer chart and found the gain of stages simply looking at schematics. But that still would not have produced the reason!

And, I think in this day and age of riduculous computer/calculator power, we could probably all use real equations for design stages - but again, I would like to know WHY for each step.

Maybe I'm just chasing a dream of a boyhood gone by when I dive into this area, but most books I've read on amplifier design lack the clarity of purpose exhibited by the Radio Amateur's Handbook of the 60s. And that is the sort of step by step design knowledge base I'd ultimately like to achieve in this thread, particularly with hybrid designs.

Thanks again for your indulgence.

Martin

[airboss]

Originally Posted by mvon21234

A PROPOSED GROUP PROJECT?

Output tubes: 6146 or equivalent, AB1 pentode. (I can hear the groans now! I have read the “critiques” here.) Why? I have a lot of 6159s, and building a DC filament supply is no big deal. ?

Originally Posted by tubelab.com

I have been reducing my collection and I recently gave away about 200 6159's

Ok so I get curious you know? I have the RCA Transmitting Tubes Technical Manual TT-5 marqua registrada 10/62 and I open it up to 6159 wondering how anyone in there right mind would give away 200 of anything other than empty St Pauli Girl bottles. So I see this picture of an indirectly heated triode that needs 26.5 volts of heater at .3 amps and am even more puzzled. I'm with Martin, so what? Then I read that this tube is identical to a 6146 except for the heater voltage. Hmmmm. The 6146 is a beam power tube with a screen and suppressor grid in addition to the K and G and P. So I think RCA musta goofed. So tubelab, dude, did you really give away 200 indirectly heated triodes rated at 25 watts ICAS? With a bottle outline 17?

Martin: Just start drawing stuff, post it here and ask for critique, I'll bet before long you got your school. I got a 400-0-400 plus a bunch of other screen taps and stuff big iron rated at 1 amp. That's right 1 amp. Triad to boot. It orta run a few 6159s. I've been saving it for a 807s modulating 807s AM rig on 75 meters. It's a prettier tube anyway. It might be better as 811s driving 811s. Haven't run the numbers. 1 amp might not be enough. I've never done that before either but that usually doesn't stop me. I put a Gates BC1-J on 75 meters once. Trial and error. Some of the errors were spectacular! Like leaving a safety ground on the PA tube cap and hittin' the go button.

Side note. One problem with BIG power amps is speaker efficiency. My speakers take 1 watt to make 104db at 1 meter. Having 100 watts doesn't ring my bell unless I also build a line array with 83db speakers.

[tubelab.com]

Ok so I get curious you know? I have the RCA Transmitting Tubes Technical Manual TT-5 marqua registrada 10/62 and I open it up to 6159 wondering how anyone in there right mind would give away 200 of anything other than empty St Pauli Girl bottles.

When your friend calls you up and says drive 200 miles, help me unload 2 semi trailers, stack what I want in my warehouse, and you can have all of the tubes, I didn't even think, I said yes. It took 3 weekend trips to Orlando to help him unload and sort all the stuff, and I brought back about 1500 pounds of loose unboxed, unsorted tubes each trip. Most were military surplus from the WWII era.

I had to rent 3 rent a space warehouses just to store them all, packed floor to ceiling. Based only on volume I estimate that there was about 100,000 to 150,000 tubes! That was 6 years ago and the rental units were $100 per month each. Now most who have been reading these pages know how I like to blow things up, but 100,000 tubes, it took me almost 5 years just to go through all the boxes.

The first pass was just to throw out all the broken tubes, 10 to 30,000 tubes, all busted. Next, I sold off all of the HV rectifiers, radar tubes, big thyratrons and other stuff that I knew I would never use. I reduced the volume to fit into 2 rental storage units.

About 3 years ago a chain of events started that accelerated the process of reducing our "stuff". My father passed, followed by Sherris step father. The process of going through their lifetimes accumulation of "treasures" and figuring out to do with it all, led us to realize that we have far too much "stuff". Sherri lost her job manging that rental warehouse complex which led to a sharp increase in rent. My job was in serious jeopardy for about two years with weekly layoffs, so we decided both of us had collected enough "stuff" for more projects that we could possibly finish in two (maybe 3) lifetimes, so we each made a pact to eliminate half.

Giving away about 200 POUNDS of 6AL5's and other tubes that I would never use was easy, I actually sold several pickup truck loads of tubes for about $100 each. I have managed to reduce my collection to maybe 10,000 to 20,000 tubes with more to go, but I now have to sort through bushels of mixed 9 pin miniature tubes.

I could have kept the 6159's, but the chances were slim that I would actually use them. I had 2 8 cubic foot boxes full. I gave one to a ham radio friend, and sold the other for $25 at a hamfest. Yes, I sold even more 807's and NIB 1625's too. I have made some wicked sounding guitar amps with 6146's in my younger days, but I am more interested in doing less common circuits these days.

I still have to reduce space again since the A$$hole landlord has raised the rent again, and wonders why his tennants are all leaving. Sherri and I are attempting to combine her stuff with my stuff and fit it all into 1 rental space.

I have several 1500 volt 1/2 amp power supplies for free if anyone wants them, but they weigh about 85 pounds each, so shipping is out of the question. I think I'm going to part with much of my 833A collection too.

Both can be seen in this picture, power supply on the left, glowing 833A on right.


did you really give away 200 indirectly heated triodes rated at 25 watts ICAS? With a bottle outline 17?

Both the 6146 and 6159 are pentodes (or beam tetrodes).

[airboss]

Well I did figure out that the 6159 wasn't a triode. RCA put the wrong tube outline in the book.

We ought to be neighbors you know? I got a lot of stuff I could give you. 6AL5s. Oh yeah. I got some. Detector in all that Military Collins S line stuff. I got this wild idea that a guy could use one for a mixer tube by biasing it at 70% with plate and cathode resistors, tie the plates together, put an input above each K resistor and taking the sum off the plates with a capacitor. I only got about 5 sleeves though. I might blow them all up trying. Then what would I do if somebody gave me a 75S-3B. "Here, I don't need this anymore". It only needs one 6AL5. Arrrgghhhhh.

I wish you hadn't tossed all the 807s though. All mine are pulls. They get
$25-$30 a pair for Jan now. Pedal steel guys like them because of Bob Crooks.

Oh, and your open wired 833A is famous on the net you know. "Wild Man endangers earth with big triode" I'm proud to meet you. No one has ever made attribution for that picture.

[aardvarkash10]

good lord, thats a MILD display of high voltage heroics by George's reckonning. A number of his more spectacular efforts are posted here including a fabulous corona effect effort using a trashed microwave oven and some largish junk tubes.

And by Corona effect, i mean you will spray beer all over your monitor when you see it....

[airboss]

I believe it. We've met in the "Anyone going to Dayton Hamfest" thread. My kind of guy. I got kicked off my High School football team for chasing guys around the locker room with "Tuff Skin" sprayed over a match.

Try it. You'll like it.