Hello all,
I've ordered a tubelab se board and am in the process of coming up with a tube selection. For the output tubes I will use Shuguang 2A3C. For the input tubes I would like to use something other than the 5842's suggested on the site. Is this possible? I would like to use something like the 12AX7. What modifications would need to be made to accommodate this or similar input tube? Thanks,
Matt
I've ordered a tubelab se board and am in the process of coming up with a tube selection. For the output tubes I will use Shuguang 2A3C. For the input tubes I would like to use something other than the 5842's suggested on the site. Is this possible? I would like to use something like the 12AX7. What modifications would need to be made to accommodate this or similar input tube? Thanks,
Matt
I have been trying to understand why the 5842 is uniquely qualified for this amp as well.
Is it because there are very limited tubes that can drive a 300b as a single stage? Even with a "power drive" circuit?
I have seen a few 300b schematics using D3a and others with more than one driver tube, of course.
Maybe the modification requires abandoning a 2 tube per channel design, although if i were planning to experiment I would try the D3a.
From my very limited knowledge, the 5842 is very linear and has a very low plate impedance. I understood that the latter is the most important thing.
If you follow the various design that you find around they try one way or another to have a driver stage with the lowest impedance. So SPPR, parallel tubes, interstage transformers and so on.
What I wonder is why nobody produce this good tube any more.
Best Regards,
Davide
If you follow the various design that you find around they try one way or another to have a driver stage with the lowest impedance. So SPPR, parallel tubes, interstage transformers and so on.
What I wonder is why nobody produce this good tube any more.
Best Regards,
Davide
The D3a has a similar plate impedance, greater mu, but greater transconductance as well. I hear positive things about it. Has anyone tried to sub it in? I suspect George has thought about this and has a reason he hasn't considered it.
What I read about the 5842 tube is that it can tend to be microphonic. I think it is particularly true of the WE 417.
The 5842 that I've found are around $40 for a pair (on ebay). Not quite what I had in mind in price. But, it'll work.
Thank you for the replies!
Matt
Matt,
I'm actually doing the work for you... I'm designing an amp using a 300B, 5k OPT, and optimizing for a reasonable level of power at low distortion.
I think part of the problem is that the 300B does not have much amplification by itself, hence, needs about 160 Vpp to be driven to the point of grid current. The actual drive voltage depends on a lot of things, but for the operating point I've chosen, 160~180 Vpp is my design target. Unless you have a preamp that can deliver 10 Vpp or more, your options are limited. I would prefer to have full power at 1 V RMS input (2.8 Vpp). So I need a gain of 180/2.8 = 64 V/V in the stage driving the 300B. Very few triodes can do this in a single stage. EF86 (or Russian 3S3P-EV), ECC81/12AT7, 5842, 6SL7 come to mind. There might be a few others. So that's the box you're painted into...
I'm playing with various other options including a 2-stage driver. You'll find some info in this thread: http://www.diyaudio.com/forums/tubelab/164324-limiting-grid-current-powerdrive.html. Eventually, I should probably spin off a separate thread in the Tubes forum.
~Tom
I'm actually doing the work for you... I'm designing an amp using a 300B, 5k OPT, and optimizing for a reasonable level of power at low distortion.
I think part of the problem is that the 300B does not have much amplification by itself, hence, needs about 160 Vpp to be driven to the point of grid current. The actual drive voltage depends on a lot of things, but for the operating point I've chosen, 160~180 Vpp is my design target. Unless you have a preamp that can deliver 10 Vpp or more, your options are limited. I would prefer to have full power at 1 V RMS input (2.8 Vpp). So I need a gain of 180/2.8 = 64 V/V in the stage driving the 300B. Very few triodes can do this in a single stage. EF86 (or Russian 3S3P-EV), ECC81/12AT7, 5842, 6SL7 come to mind. There might be a few others. So that's the box you're painted into...
I'm playing with various other options including a 2-stage driver. You'll find some info in this thread: http://www.diyaudio.com/forums/tubelab/164324-limiting-grid-current-powerdrive.html. Eventually, I should probably spin off a separate thread in the Tubes forum.
~Tom
The selection criteria for the 5842 in the Tubelab SE was in reality none of these things. There are some basic electrical criteria that must be satisfied, minimum gain, low noise, low Rp, and good linearity. This narrows the list a bit.
As mentioned on my web site somewhere the Tubelab SE was never intended to be built by anyone other than me. I learned electronics in the 60's on tubes. Since then I built guitar, HiFi, and RF amps for myself and friends using tubes and solid state components. Some were quite powerful, 1200 watt tube RF amp, 300 and 1200 watt SS audio amps all built in the early 70's. Throughout those early years I was power hungry so all of the tube audio stuff was push pull. I thought the low power SE stuff was just a fad, until I heard some. That started a 5 or 6 year case of SE fever. The fever was so intense that I built nothing else, just SE tube amps for several years.
I built (or breadboarded) several of the popular SE designs of the day including the common 6SN7 - 300B - 5AR4 design using cathode bias. They sound really nice with a single female vocalist played at a level that does not touch clipping. I wanted something different. It had to sound as good as, or better than the amps that I had built, or heard, but unlike many SE amps of the time it also had to ROCK. Now, I realize that the typical SE amp builder wouldn't crank Metallica through their SE amp, but improved transient and overload recovery without degrading other traits is good for everybody.
By the time I started on what would become the Tubelab SE I had amassed a collection of over 100,000 tubes. During the design cycle I made a list of every possible driver tube that I had at least 50 of and tested them. I made measurements on tubes in individual gain stages, and tested them in a breadboarded amp using a 45 output tube.
I was in search of my own "ultimate SE amp" and during the course of about a year I discovered the CCS load, and the PowerDrive. These made a single triode driver for a 45 or a 300B possible. I built several amps during this time using several different tubes. The 5842 with a CCS load rose to the top of the list. At the time I designed the Tubelab SE the 5842 could be purchased for $4 to $6 each!
I built my first Tubelab SE about 6 years ago and was happy with it. I showed it off to several friends and poof it was gone. Someone made me an offer that I couldn't refuse. That happened a few more times. Then I met a well to do MD who showed me his "ultimate system". He had a 300B amp that he had bought in an audio shop in Hong Kong for some ridiculous sum of money. I told him that I built tube amps, and I had a little SE amp that he should hear. He wanted another amp for his bedroom so I loaned him my Lexan amp. A few weeks later I went to retrieve my amp and I found it wired into his main system in place of the megabuck amp. He was quite insistent about keeping it. I explained that it was made with a possibly unreliable home made PC board (it still works). Enough money flowed to make the purchase of professionally fabricated PC boards possible. I wound up building him 3 Tubelab SE's and sold a few more, but I no longer have time to make complete amps. I did have about 30 extra PC boards, so.......
At the time the Tubelab SE was designed the D3A was not widely known in the US. It is still not commonly available. I bought a pair from Germany about 2 years ago (not cheap either) and it is an excellent tube, but I think that they are as hard to get as the 5842 if not more so. The 5842 has a rather unique pinout, so using a different tube would require hacking the PC board, or making an adapter.
I have thought about re-designing the Tubelab SE, but it doesn't make sense right now. Board sales are very slow (3 orders for April) and the Simple SE is by far the most popular. I have enough Tubelab SE PC boards to last a while. It also helps that I found some more 5842's in my warehouse.
The amplifier circuitry in the Tubelab SE is still capable of the same stunning sound that captured my attention several years ago. If anything needs updating it would be the power supply, especially for operation at voltages over 400 volts. The Tubelab SE board was designed to operate from an external supply, and some users are already doing this. Well, so am I. My 845SE used an external supply, and I also have another Tubelab SE amp (still unfinished) that uses an external regulated power supply (currently built on perf board).
The power supply design is still a work in progress right now, and with my current work schedule I can't even say when it will be done. I am trying to design something "universal" for use with several different tube amp designs, both SE and P-P.
Hey George...
That's a good story how the Tubelab SE came to be. Thanks for sharing.
Just last night I had a friend over to listen to some music. I was all set to play some records with the "main" system. Then he spied the table where I have my DIY amps sitting. He was intrigued by just the 4 tubes sticking out of my SSE, and insisted on hearing it. Now, I'm on the hook to build him one!
It may be "simple", but it sounds very good!
That's a good story how the Tubelab SE came to be. Thanks for sharing.
Just last night I had a friend over to listen to some music. I was all set to play some records with the "main" system. Then he spied the table where I have my DIY amps sitting. He was intrigued by just the 4 tubes sticking out of my SSE, and insisted on hearing it. Now, I'm on the hook to build him one!
It may be "simple", but it sounds very good!
Attachments
Yeah, the D3a is not cheap and most of the eBay sellers are outside the US and so the shipping charges approach the cost of the tube.
Keep an eye out for a seller selling sleeves of 5 Raytheon 5842s. He tests them all and he had enough at the time that he could even sell a sleeve with 5 tubes that match fairly close. The consistency of these Raytheon military grade tubes is not the best. I was able to contact him directly and buy a couple of sleeves. This was about a year ago and he mode it sound like he was swimming in these things.
Keep an eye out for a seller selling sleeves of 5 Raytheon 5842s. He tests them all and he had enough at the time that he could even sell a sleeve with 5 tubes that match fairly close. The consistency of these Raytheon military grade tubes is not the best. I was able to contact him directly and buy a couple of sleeves. This was about a year ago and he mode it sound like he was swimming in these things.
Uh Oh, you are in for a really long night. I have some circuits breadboarded, and several more ideas that remain untested. I have this mega powered P-P amp breadboarded, but missing a power supply. I have a tweaked up red board that makes 100 WPC, again, missing a power supply. I have a dual Simple P-P board that needs a screen regulator, and of course both the Simple SE and the Tubelab SE could benefit from an external regulated supply. Now the complicated part is making one device to fulfill all of these requirements. Yeah, and I leave town on Wednesday for 2 weeks.
I think my pair cost me about $25 each with shipping. I got them at SY's insistence to build an early version of his phono stage. They are very quiet tubes.
I had a box with about 50 used Raytheons in it that I used to make my original amps. They are highly variable tubes. These tubes are the reason for the trimmer pot in the cathode circuit. The WE versions are not much better. While digging through the warehouse last weekend, I found another box with about 20 5842's.
Not to get off the track, but the popularity of that tube can be traced to some "NY audio mafia" phono stage work from the early 1980s. Prior to that it was used in its original design application which is/was a VHF type tube as a gain stage or in a high IF section... some people liked the way they sounded run for AF... Western Electric developed the tube - an interesting read on the design - and Raytheon went on to use it for various military applications. The "urban legend" has it that they are/were used in the front end of a certain anti-missle system fairly recently, maybe even still... where they seem to vary the most is in self-noise which becomes evident when you try to use them in a low noise circuit, you'd better have a big batch on hand to select from. 
Btw, they work a very very long time... very reliable tubes.
_-_-bear
Btw, they work a very very long time... very reliable tubes.
_-_-bear
I'm guessing that's for the lifespan of their cathode, filament or other things inside. I had one of Raytheon 5842 in my TSE dead after about 100 hours of use. The bottom of glass cracked and the silver coat in the tube burnt out.
Hopefully it's an isolated case.
Yes but it didn't feel different by much before and after. The other Raytheon 5842 had no problem though which led to my conclusion of structural defect of glass.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.