Recommended designs

[rockgardenlove]

So I'm looking to build a push pull amp.
Would you guys recommend something like 6V6 / EL84 as power tubes or something more like EL34 / 6L6 / KT88? I'll get something like 15 watts and 40 watts respectably, eh?

I've done lots of searching and found some schematics, but I was wondering if there were a few known goodies that people generally like.

Thanks!

 

[planet10]

El Cheapo
Red Light District
Baby Huey

(all EL84/6V6/6AQ5 family)

dave

 

[Cycline3]

As the power goes up, the price goes up -- and it can really get up there if you build out with the best parts. Many say the sound of the smaller power tubes is better, if you can get by with the lower output power. For me personally, the tube's looks is important too, so I've always went with the larger power tubes, 6L6, EL34, KT66 and 88 for my amps. I won't lie, part of the draw is the look. Either way, you'll get great sound. It also may depend on if you have any existing parts laying around that you might want to recycle too. So, do you have a target budget and do you have any parts already?

 

[rockgardenlove]

I'm aiming for around 400 bucks. I don't care about audiophile components or anything. I have 1/4 watt resistors and caps.

Thanks!

 

[Cycline3]

Well, then I think you are looking at whatever the best deals are you can get, cause you will be cutting it close. Depending on the tubes you choose, you are likely going to get the little EL84/6BQ5 or a 6L6/5881 in the cheap range. I don't think anyone will argue with me that edcor has the best cheap (~$20 us) output transformers (http://www.edcorusa.com/products/class_x.html). Radio Shack has an ALPS volume pot for $3.49 which is a good deal and a decent piece... Parts Express has their own brand of capacitors: "Dayton" that are polypropylene/metal foil at 400V types that are really, really good for just over $1 each. PE also has the hammond chassis and ceramic tube sockets from china for a cheap price. Those are some places to look for deals, but you also should set some goals for the build and it's probably a good idea to at least work out a rough idea of your schematic/build before you start buying.

 

[Eli Duttman]

Magnetics are going to be the "big ticket" items. You can get a pair of the price/performance champion Z565 DynaClone O/P trafos from Triode Electronics for 140 bucks. Well under $100 gets you all the power "iron" needed to build an "El Cheapo".

A heavy walled plain Aluminum baking pan makes a perfectly satisfactory low cost chassis.

Use "reissue" TungSol 6V6s and Mullard made 'T7s for signal tubes. Jim McShane will charge you $124 for that very nice set. A little more buys the 'AL5 B- rectifier.

You are going to end up with a KICK BUTT 12 WPC "integrated" amp.

 

[damilkman81]

What if?

If money were no object and all I have built to this point is two S-5 K-16LS kits. Where to from here. The sound that is produced from the moded kit is fantastic but I crave more. After tweeking as far as I can, "I keant push'm eny ferhter captin". I agree with Cycline3. These 6005 tubes sound great but it would be nice to see a bigger glow and share the experience with the neighbors.



But I have promises to keep, and miles to go before I sleep.... And miles to go before I sleep.

 

[Eli Duttman]

Re: What if?

If money were no object and all I have built to this point is two S-5 K-16LS kits. Where to from here. The sound that is produced from the moded kit is fantastic but I crave more. After tweeking as far as I can, "I keant push'm eny ferhter captin". I agree with Cycline3. These 6005 tubes sound great but it would be nice to see a bigger glow and share the experience with the neighbors.



But I have promises to keep, and miles to go before I sleep.... And miles to go before I sleep.

Money is always an object. FWIW, I'm coming around to the view that the best power/cost setup may be Class "AB2" PP 6L6GCs in a Mullard topology variant. An honest 50 WPC should be available. For 8 Ohm speakers, the O/P trafos would be Edcor's model CXPP60-8-6.6K. KT88 based circuitry will cost more. EL34s can't safely enter a positive g1 current regime.

 

[chrish]

I don't want to steal the direction of this thread, but things look like they might be headed my way. I have stashed away a pair of lovely Tamura 6.6K Push Pull output transformers and a pair of nice aluminium chassis waiting for a pair of PP monoblocks.

I have a quad of 6P3S and another quad of 6P3S E (Russian 6L6 equivalent). Also have a stash of Russian 6SL7 and 6SN7, a pair of Mullard made ECC81, 4 x JAN 12AT7WC GE made (I know you are a fan of the 'AT7 in PP Eli!).

I am just about to finish my university assignments for the year so looking for a project. Don't have to use the Russian valves, but topology is going to determine B+ and therefore power transformers. Eli, care to elaborate on your suggested 6L6 topology? Is it suitable for the Tamura output iron? I am thinking of buying the required power transformers for monoblocks soon. As I have just moved house and now have a mortgage (and the Australian dollar has taken another nose dive), I don't want to buy the wrong ones.

Suggestions?

Thanks,

Chris

 

[Eli Duttman]

Eli, care to elaborate on your suggested 6L6 topology? Is it suitable for the Tamura output iron?

Chris,

The starting point for this scheme is the "classic" Mullard 5-20. A section of a 12AT7/ECC81 is used as the voltage gain device. An ECC99 will be the LTP. I firmly believe that high gm small signal types are indicated in designs with loop NFB. Slew limiting sucks!

Does the Tamura "iron" you have in mind have either UL taps or a tertiary cathode FB capability? IMO, too much global NFB is needed for full pentode mode, when local NFB of some kind is not provided by the O/P trafo.

Assuming that Tamura O/P trafos are suitable, the big topology change is the insertion of IRFBC20 source followers that are DC coupled to the 6L6 control grids and cap. coupled to the LTP anodes. Refer to MOSFET Follies and Tubelab's PowerDrive. Please observe that the amount of coupling cap. created phase shift is unaltered, while the low impedance/high current drive requisite to a positive control grid current regime, like Class "AB2", is present.

 

[chrish]

Thanks for taking the time to respond Eli, and sorry for any thread hijack. If things look like moving away from the original question on the thread I will start another thread.

The output iron is Tamura F684. It has UL taps and "a third winding exclusive for NFB use" 16R.

I thought that I read in Morgan that the original Mullard 5-20 had high sensitivity and poor noise performance. I am guessing the replacement of the EF86 with the ECC81 etc and the sand source followers address these issues.

Very interesting. Have two more assignments due in the next week, then finished till Feb. so will be able to start designing something...

Thanks again,

Chris

 

[Francois G]

I'm aiming for around 400 bucks. I don't care about audiophile components or anything. I have 1/4 watt resistors and caps.

Thanks!

Take a look at http://www.diytube.com. The construction manual is here: http://www.diytube.com/phpBB2/viewtopic.php?t=837

The ST-35 PC board makes it very easy to build an EL84 Dyna ST 35 clone that should set you back less than $400.

 

[Ty_Bower]

The DIYtube ST35 makes a very nice first project. I cobbled together my first one with less than $100 of iron (salvaged power transformer, $20; Hammond 1620 output transformers, $65 for the pair) plus $100 for the board and parts. I stole the tubes out of another amp and the ST35 sounds just beautiful for not a whole lot of dollars.

Having everything on one circuit board (power supply, tube sockets, everything!) makes construction a total no-brainer. Results are virtually guaranteed. Highly recommended.

 

[richwalters]

I thought that I read in Morgan that the original Mullard 5-20 had high sensitivity and poor noise performance. I am guessing the replacement of the EF86 with the ECC81 etc and the sand source followers address these issues.



Chris

Strange, the completed Mullard 5-20 spec quotes a s/n ratio around -90dB down. That's quiet enough. Maybe one was looking at the EF86 spec alone and not the completed amp.
Somewhere is a design of the Mullard 5-20 using the EF86 as a triode and 15dB sens drop but with far reduced noise.. The characteristics then become a ECC82 look-a-like, but with a far better HF response. Obviously the drop in loop gain will req. nfb component changes.

richj

 

[chrish]

I am not an expert, or even an informed commentator, just remembering what I saw in Morgan... I think what Eli is suggesting is going to be different enough from the original design for those comments in Morgan to be irrelevant.

I am reasonably good with a soldering iron and a drill press and following a schematic, but I am no electronics engineer. I would be grateful for some gentle guidance

So, for a 6L6 based design, I am guessing around the 300 volt mark B+? So a transformer something around 370-0-370 200mA (for monoblock) mark? Plus I will need a -ive bias. Excuse the ignorance again, but what kind of current draw is there on the bias? I would imagine quite low...

Thanks for the help so far. If this turns in to another design, I will start a thread

Chris

 

[Eli Duttman]

I am not an expert, or even an informed commentator, just remembering what I saw in Morgan... I think what Eli is suggesting is going to be different enough from the original design for those comments in Morgan to be irrelevant.

I am reasonably good with a soldering iron and a drill press and following a schematic, but I am no electronics engineer. I would be grateful for some gentle guidance

So, for a 6L6 based design, I am guessing around the 300 volt mark B+? So a transformer something around 370-0-370 200mA (for monoblock) mark? Plus I will need a -ive bias. Excuse the ignorance again, but what kind of current draw is there on the bias? I would imagine quite low...

Thanks for the help so far. If this turns in to another design, I will start a thread

Chris

Starting a new thread that links this thread is an excellent idea. Title it AB2 6L6GC or something similar.

B+ for this project will be about 450 V. Look at the 6L6GC data sheet. While you are at it, pay close attention to the plate curves of positive g1 voltages. Notice that when g1 goes substantially positive, quite a bit of current is present. That current is associated with (among other things) the negative rail. B- current requirements here are substantial.

The "idle" current of the 6L6GCs gets controlled by what goes on at the MOSFET gates. Remember, the "beamie" control grids are DC COUPLED to the FET source electrodes. An initial WAG on my part for the 6L6GC idle current is 10 mA.

 

[chrish]

Thanks Eli, new thread started here: 6L6 AB2 Amp