• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

300B Summer Build Question

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am deciding on a design and gathering parts for my summer 300B amplifier build. This is a project I have wanted to build for quite some time now. I have looked at a few different designs and I guess I have some questions about the different layouts. This is the main schematic I have been planing on using.

http://homepage.mac.com/parisatnight/.Pictures/CabPics/300B-AMP.jpg

I haven't decided yet if I will use 6J5's or a single 6SN7 since I have a lot of NOS 6SN7's. I am wondering though how this design would compare to something like the Angela Model 91. It uses a 6J7 for the input/driver but I think I would switch it to a C3g or something else, 310A maybe if I can get my hands some. I hear the pentode is a better driver for the 300B but I have never heard one. Would this design (the Model 91 clone) benefit from a choke load. Or would there be another design that would be as good or better than either of the two previous ones?

For reference I am planing on using Electra-Print output transformers (3k) with EH 300B Gold Grids. For the PSU I will be using RCA NOS 5U4G's with Angela or Electra-Print power transformers. I don't know yet if I will us Hammond chokes or Electra-Print chokes. I am planing on using a C-L-C layout with ASC Oil capacitors. I may switch this to a L-C-L-C layout. They will be built as mono-blocks one wooden frames with aluminum or copper tops. These will be driven with a 12B4A line stage or JFET Bride of Zen line stage. In the future these may be joined by a 26 or 10Y choke-loaded line stage.

I am open to any suggestion and feedback, thank you in advance!

Cheers!

James
 
If you search through this forum a bit, you will find lots of opinions on how the 6SN7 is a poor driver choice for the 300b. Moreover, the 12b4 would be a great driver. Why not switch these around. In commercial products, certain amounts of gain, or certain input sensitivities are necessary, but in DIY, you can do what you like. So, make a 6SN7 preamp with lots of gain, then a 12b4->300b (or 6J5->12b4->300B) power amp with a low input sensitivity.

Also, even if you don't want a full DRD, setting up that last cap as an ultrapath cap would be a good idea.
 
^^^^

The design looks pretty good. It could stand to have a grid driver, though. DHTs like the 300B are rather difficult to drive, due to the large Cin, and that the grid starts pulling current even before Vgk actually goes positive. A 6J5 with a plate current of 4.0mA seems iffy there. I would include a power MOSFET source follower grid driver for the 300B, as the disappointing sonics of this type seem to be related to inadequate grid drive. That would also allow direct coupling between the MOSFET and the 300B to remove "blocking" as a cause of sonic degradation. It also makes for easily adjusted fixed bias instead of cathode bias.

I'd also make provision for adding some gNFB as well.

If you decide to go with the 6J5s, remember to ground the metal shells (connected to pin #1).
 
Thanks for the replies. I have been doing some searching both on and off the forum and have come up with a few more ideas. Out of interest though, why is the 76 preferred (by some) at the input over the 6J5/½6SN7? I was looking at the Electra-Print DRD and am doing some reading on it now. If I went this route would it still be a good idea to have a FET source follower? Also can this be done with a tube? Just as a basic cathode follower?

Regarding the ultra-path cap in the Electra-Print and other schematics, do I still have to worry about excessive hum? I have read a few places when I was looking at building a 6J5 Ultra-Path line stage (also Electra-Print) that hum can be a major problem if not properly designed.

I was looking too at the JE Labs 300B DX design, which is where I saw the 76 input, and had a few questions. Instead of ½ can I use both halves of the 6SN7 and change the operating point to better drive the tube, or still is a power follower a better idea? I do however like the PSU on this amp better than the one that was designed to go with the schematic that I first posted. I was looking it though and noticed that it uses a 47uF cap at the input, to me this seems large for a 5U4. I was also wondering, is there a purpose to the 50k resistor to ground other than a bleed off resistor? I am just wondering why this has to be a 50watt resistor.

I am interested too if anyone has tried the Electra-Print / Welborn Labs DRD designs or anything similar. I am especially wondering about the new designs on Electra's site. They are using large electrolytic caps instead of oil caps (in the output).

Would DRD be a better option than choke loading the driver / changing operating points / FET power follower / something else?

Thanks and Cheers

James
 
JPeitzman said:
I was looking at the Electra-Print DRD and am doing some reading on it now. If I went this route would it still be a good idea to have a FET source follower? Also can this be done with a tube? Just as a basic cathode follower?

The DRD is choke coupled, and uses a pretty stiff driver VT. If you go this route (or use xfmr coupling) then a MOSFET driver wouldn't be necessary. It's probably a good design, though I don't like all that iron in the signal path. Also, forget about any gNFB if you do this.

Regarding the ultra-path cap in the Electra-Print and other schematics, do I still have to worry about excessive hum? I have read a few places when I was looking at building a 6J5 Ultra-Path line stage (also Electra-Print) that hum can be a major problem if not properly designed.

I'd say it's not a problem. This includes a power supply design.

I was looking too at the JE Labs 300B DX design, which is where I saw the 76 input, and had a few questions. Instead of ½ can I use both halves of the 6SN7 and change the operating point to better drive the tube, or still is a power follower a better idea?

With this design, you're back to the inadequate drive problem.

I do however like the PSU on this amp better than the one that was designed to go with the schematic that I first posted. I was looking it though and noticed that it uses a 47uF cap at the input, to me this seems large for a 5U4.

It is. I did a project using the 5U4GB for the positive rail. The input capacitance I used was 34uF (two 68uF in series) to get some 900mA of Isurge. 47uF exceeded the max Isurge rating of 1.0A / plate. I forget by how much, but the 5U4GBs'll last longer if you stay within spec.

I was also wondering, is there a purpose to the 50k resistor to ground other than a bleed off resistor? I am just wondering why this has to be a 50watt resistor.

He's loading down the PS, probably for improved voltage regulation. Be less wasteful of power to make a choke input filter.
 
JPeitzman said:
Regarding the ultra-path cap in the Electra-Print and other schematics, do I still have to worry about excessive hum? I have read a few places when I was looking at building a 6J5 Ultra-Path line stage (also Electra-Print) that hum can be a major problem if not properly designed.

Basically you multiply the B+ ripple times the mu (reduced by the OPT winding ratio) to get the hum. With the line stage, mu is 20 and it gets amplified again in the power amp, so it can be an issue. On the output stage, since the 300B's mu is low, it isn't a problem.

JPeitzman said:
I am interested too if anyone has tried the Electra-Print / Welborn Labs DRD designs or anything similar. I am especially wondering about the new designs on Electra's site. They are using large electrolytic caps instead of oil caps (in the output).

Jack seems to not think this makes much difference. I think it does, and these oil caps are cheap anyway. Keep in mind that the idea here is that rather than the cathode bypass cap and the final PS cap both being in the signal path, the design is supposed to remove one of them from the signal path. So, an electrolytic that might be in the PS anyway is not going to be terrible, but is probably bettered by the film cap.

Fwiw, the DRD design is really Jack's. Welborne used it, but Jack did the original. I have DRD amps and like them a lot, but I have not compared them to much else so I can't say they are better. Well, I can say they are better than my old Sumo, or a Gainclone :)


Miles Prower said:
The DRD is choke coupled, and uses a pretty stiff driver VT. If you go this route (or use xfmr coupling) then a MOSFET driver wouldn't be necessary. It's probably a good design, though I don't like all that iron in the signal path. Also, forget about any gNFB if you do this.

For my money, more iron and less gNFB is a good thing. There isn't a right answer here, just different ways of doing things and different preferences.

Search around for user Kofi Annan. He build some DRDs (d3a->300B iirc) and there is information about issues he ran across.
 
Have you considered that the improvement in sound you get by using a DHT like the 300b can be continued right through the amp by using more DHTs as input and drivers?

Keep the circuit simple - pretty much as you have it, but use something like:

Input: 201A/01A or 26 or 10Y

Driver: 46 or 10Y Other choices but less gain, 31 and 71A

If you want the maximum gain and probably the best sound, I suggest 201A into 10Y. If you can afford 10y into 10y that's even better.

In any case, if you use 4 pin UX sockets for input and driver you can play about with some of these DHTs. If you want to experiment, some bench power supplies are great to drive the DC filaments at various voltages. The 46 is a 5 pin UX socket. You can get 4 pin UX sockets that fit in an octal hole (and if you're lucky 5 pin ones), or you can get the big oval ceramic UX sockets, which they also do in octal. If you use those, you could start with 76 (UX5) and play about with DHTs later - at least the sockets would be compatible.

I've built with all the above, and I'd say that the improvement in transparency and detail and treble finesse was well worth it. Andy
 
I think you'll be quite pleased with the sound of the EH Gold 300B output tubes. I have them in my 300B amp and I like the sound a great deal. I've been using them for about the last 7 months since I built the amp and I've used them a lot with no problems at all.

I'd love to get some other 300B's to try but they are quite expensive, maybe I'll leave well enough alone.

I have an old set of Altec Iconic speakers and the amp drives them easily. The speakers I have are 16 ohms and that's no problem with the One Electron output transformers which have a tap for 16 ohms. The setup is not really super loud or anything but it sounds amazingly big. My listening room is approx 26X40ft yet the sound fills the room and the whole house easily my friends are amazed at what 16 watts peak can do!
 
Thanks so much to all that have replied! I figured I would get some action since the 300B is a popular amp but I never expected so much help.

coresta - I am drawn to the 91A type design for some reason, maybe because it is a classic 300B design. From what I have seen / read it seems like pentodes do a good job of driving DHT outputs. It seems though that so do triodes (DHT or not) when properly implemented. Since I have not heard either I am still up in the air on this issue, I may just have to try both and find out. Thanks for the input.

Miles Prower - Thanks for all the input! I am wondering though if I switched the 6J5 (of half a 6SN7) to something else, like a 12B4 (as dsavitsk said) would it drive the 300B better? That and i figured choke loading the driver would help, bigger swings. I know I am trying to weasel my way around the power follower, I am not opposed at all though. I have seen a few designs that use them, some of the Tubelabs amps come to mind, and they seem to be a viable option. Thanks for the reply on the JE Labs power supply, I figured the 50k was doing more than providing a safety factor when the psu was switched off. I had never seen a 50 watt bleed off resistor. I think though I am going to follow your advice and do a L-C-L-C power supply. I had been kicking around this option for a while and it seems like a better option. Thanks again!

dsavitsk - Thanks for all the input as well! I thought you had built DRD amps, I thought I had seem a thread around here where you mentioned them vs. T-Paths maybe... I don't remember for sure. I have a gainclone too actually and was not overly impressed by it. My Knight 6BQ5 PP and RCA 6L6 PP put it to shame and my JLH Class-A just walks all over it in my opinion. I am confident that any of the 300B designs would be better than the GC. I do use mine for movies sometimes, where the higher power gives a little better bass, but I still prefer my Knight for this too. I am hoping after the 300B to build some 807 or KT88 PP's for movies / loud music. Thanks for the info on the ultrapath hum, and I will give Kofi Annan a search. Thanks again!

andyjevans - I had thought about that path too. I was originally worried a bit about hum in all the tubes when using DHT throughout, but this I am sure would be a problem that is easily solved. I have heard that the 10Y sounds very good, it seems to get praise as a choke-loaded or transformer coupled line stage. I would like to listen to a all DHT amp to see how it pairs with IDHT triodes and pentodes. Another design I will have to consider, thanks for the input.

kstagger - Have you by any chance listened to your amp vs. other triode driven 300B designs? Like I mentioned earlier I like the 91A type designs with some tweaks. I have heard though that pentodes have the best sound of all 300B drivers. I have also heard that when compared to some triode designs (DHT or IDHT) that they sound slow, light, and kind of bass shy. Of course these are all opinions and I am sure mine would differ as well. Thanks!

DaveMcLain - Thanks, I was really hoping someone would voice an opinion about the EH Gold 300B. I have considered these, TJ mesh plates, JJ's, and KR's. I can get the EH's at a good price and I figure I can roll them in the future if I really see fit. Just out of interest what design are you using with your 300B's? Yeah my friends are still amazed at my 10watt JLH Class-A. They find it hard to believe that even with movies and loud music this amp just sits there and plays away contently. A couple of them still don't believe me that it is 10 watts.

Thanks again to all that have replied. I started down this simple path, I had a design in hand and parts list all laid out. It seems though that the Zen path of SE simplicity it a lot longer and much more bumpy than it looks. I had considered starting out with a 813 SE just for the fact that they are very easy to drive. I can get about 20 watts in triode and almost 50 as a UL pentode. The biggest problem with them like many other big power tubes (GM70, 211, etc...) they need very high voltage to really "come alive" it seems. Though I guess I could put 400v on a 813 and see what happens. I have always lusted over 300B's though, it seems like all tube heads one day will end up looking at 300B's. I figure it is an easier place to start then the GM70 that I really want as well. I like too that the 300B has so many designs possibilities and all of them seem to sound different. I showed my girlfriend pictures of different 300B's along with the case and layout I would like to do and she said wow, go for it, so I think I have the OK.

Thanks again for all the input! Cheers!

James
 
The amp I built was a kit the CGV-300B. I bought this kit because it seemed to be simple and straightforward to build. It was reasonably inexpensive and I figured that would allow me to get it built and up and running and then I would have a good platform to use for experimentation.

The amp sounds great. I did a few things to mine. I played around with the power supply beefing it up and changed the filter caps to some JJ clamp mount which are now mounted on top of the chassis. I added a standby switch and I changed the coupling caps over to copper PIO types from the Orange Drops that came in the kit. When you look at the big picture this amp doesn't have any expensive parts but it sounds very nice. I think the main reason is because the circuit is incredibly simple from input to output, there's nothing to get in the way of the sound.

I'm very pleased with my amp.
 
Thanks guys! I have been reading over some of the information at Tubelab, mostly because he uses a FET power drive. Thanks too for the Wavelength Audio design, just something more to consider. Seems it should be good since Wavelength makes some great amps. Not sure about the SRPP drive though, I have heard mixed feelings about this topology.

I am wondering, could a version of the JE Labs 300B DX be built with the 76 input choke loaded -> 6J5 (or something else) DRD Ultra-Path like the Electra-Print? Basically the Electra-Print with a 76 (or something) in front of the 6AN4. This seems like it would solve the drive problem and still give a more sensitive input. With a design like this though would the 76 really just sit there and contribute distortion? I still am interested if choke loading the driver would solve the drive problem, it seems like it would be able to swing more power into the 300B, but it wouldn't surprise me if I am mistaken. Could I also do a choke loaded 10 -> 10 DRD... or more problems? I guess I am hung up on 2 input tubes because it makes my design look pretty ( :D :rolleyes: ) but I can just spread things out more. I know that in most cases losing a tube can/could help with over all distortion. I am trying to stick to octal / 4&5 pin drivers too, again looks. Understand though I am not all about looks, that is why I asked for help here. I want my 300B experience to be a good one, and I would like these to be amps that stick with me for some time. Pretty much I am trying to copy this layout with a few mods. When I first saw this amp I was struck by it's looks and fell right in love. Showed it to my better half and she too was in love.

http://www.iol.ie/~waltonaudio/300images/1000Vincenzo.jpg

This amp is based on the schematic that I first posted. BTW if the builder is around at all; you have built a beautiful piece of audio art. Hats off to you :worship: . I am planing though on having the RCA jack on the back and replacing the knob on the front with a power switch. I am also planing in the future to use this chassis design for a 810 SE, GM70 SE, and a few more projects. That is if I can afford to do a GM70 build or 813 SE build. Things may get a little tight but I am going to try to keep room underneath for plate chokes/DRD choke/second PSU choke if needed.

How do people feel about the 417 or similar driving a 300B through a interstage? I know this would be taxing on case space but worth it to cram or redesign the case? To me it seems like the DRD Ultra-Path is one of the better designs because of what is in the signal path...or lack there of. With the 417 it seems like you are trading passives for a big chunk of iron. The DRD seems to do this too but with virtually nothing in the signal path. I know some favor IT's though and I guess it is another design to consider.

Thanks for all the help, Cheers!

James
 
It is worth saying that the "signal path" is more than the single wire from the input, to the grid, from the plate, to the next grid, etc. Current, both AC and DC, must flow in a loop. This is a little oversimplified, but I find it a very useful conceptual way to think about this: http://www.nutshellhifi.com/library/ETF.html

6J5 is inappropriate for a DRD. There are, however, a whole bunch of tubes that have high mu, low rp, and low distortion that can make excellent drivers -- 417 is okay, as is E180F. Better are things like D3a, E810F, 6C45, 437A, C3g, ... in fact, here's a whole bunch. Many, well almost all, of these are pentodes. Many do well as triodes, too.
 
dsavitsk - Thank you very very much for the driver shoot out! For some reason I don't think that I have ever seen that, or maybe just not payed attention to it. That is some very useful indeed, though I am missing the C3g, I know it is not there so I will try to find info. I have always liked the idea of a C3g driven 300B design. It seems though that the 12HG7 is a very tough tube to beat, Mr. Millett uses it for his 812 SE with good results.

Has anyone by any chance tried the Western Electric SE output design? Seems like it could be an interesting output stage.

So it seems with all the input I have compiled a small list of designs I am considering:

1. JE Labs 300B SE Deluxe (76 -> 6J5 -> 300B)
- This would have to have a FET power drive
- This design actually would be at the bottom of my list. It seems this basic design is good for beginners but the driver is inappropriate.
2. Model 91A with C3g or 12HG7 driver
- It seems like this is the best option since the driver can be triode wired or pentode.
3. 10 -> 10 -> 300B
- I like the idea of this design but I feel it would need a lot of iron and be very expensive to implement. It seems like for the cost vs. performance that the C3g implementation would be a better option. Of course I have never heard a 10 so I could be way off.

I said it was a short list. It seems still like the C3g/12HG7 -> 300B is the best and most versatile option. I can do a Electra-Print DRD by triode strapping it, a classic 91A, just a triode driving a 300B, or something else. Still open to any suggestions or information.

Just out of interest, to the people who have built 300B amps, if you could do it all over again what would you do different it anything?

Cheers

James
 
Sorry I meant to edit but I must have waited to long.

Maybe I am mistaken but does the driver tube on a DRD amp have to be high mu? I seems like the only reason for this is because the DRD amps all seem to be one tube before the 300B. It seems like any (within reason) tube would work if you have an input tube. Surprised to see too in the driver test just what the 6SN7 was like. The distortion was 1.6% for the 6SN7 vs. 0.04% for the 12HG7! On top of that the 12HG7 puts out 121v at 5% THD vs. 70 for the 6SN7. I assume that the 76 + 6J5 or ½6SN7 would perform similar. I am still interested though in what happens if you load the driver with a choke? It seems that a 6J5 loaded with a choke would be able to swing a lot more into the 300B than an RC coupled circuit.

Just out of interest is there anyone who has both a 6SN7 300B and a pentode 300B that could compare them?
 
Just out of interest, to the people who have built 300B amps, if you could do it all over again what would you do different it anything?>>

10Y>10Y>300b Best sound I've heard. Frankly don't know what else to use. I've heard all the usual small DHTs so I'm confident of the 10Y - been through multiple shootouts.

If I was more courageous I'd try some higher voltage tubes like 845, 813, 211, SV811........ there are loads.

But to be honest, a good 300b is a pretty fine sound - lush and detailed and quite seductive.

I went off indirectly heated tubes about a year ago. Hearing an all-DHT amp was so jaw-dropping I couldn't do without the drugs!!! I know a two stage amp is attractive on paper, but I haven't heard one so can't really comment. I have heard many high mu triodes in amps like ECC83, 6SL7 and wouldn't use those - too mushy.
 
driver

I have built a couple of the amps you have mentioned with both 2a3 and 300b output. I would not dis the srpp driver simply because it isn't in vogue at present. Properly done, they can sound excellent, and are pretty easy to implement. My present 300b amp is a variation of the "Legacy" 91a. I use either EL84 or SV83 as a driver. I have not tried it triode strapped, as I assume you would give up gain. Run the driver tube so that the plate at full drive never dips as low as the screen voltage (I run about 135 on the screen, 215-225 on plate with EL84 versions, 240 on plate with SV83.) The EL84s run at about 15ma. They sound very good, respond well to tube rolling. I have a 2a3 amp in my office that runs 8-10 hours a day, 5 days a week. It began life as a JE labs design, I later changed the driver stage to the design Gordon Rankin offered in the "Baby Ongaku", which is srpp. Happy building
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.