[JoshK]

So the following post, posted in a thread on SE DHT amps sparked a vigorous exchange between myself and Dmason.

My previous post not posted....

This is good stuff... There are lots of opinions out there on how to go about it.  I'll just share some of my perspective for "seasoning" choices. 

James looks to be a really good choice for iron, since iron makes or breaks a amp.  If you really want to go budget Edcor's OPT's are suppose to be quite a bit better than the Hammonds for about the same dough.  I have gotten some iron custom made from Bud Purvine of Onetics (same Bud as in EnAbl) at very reasonable prices and then there is Jack Eliano at ElectraPrint.  Lundahl of Sweden has good iron too.  All good choices.

Some other choices for drivers:
6SL7
1579
6AQ8
6922 cascode
D3A (triode or pentode)
12HG7/12GN7A (triode or pentode)
C3G
5842/417A


Topology choices...keeping it simple
Aikido SET amp (aikido gain + SET output...do search, plenty of examples with really good performance, can buy aikido PCBs)
SimpleSE (http://www.tubelab.com/ can buy PCBs making it a snap)
Thoersten Loech's Legacy amp. (www.thunderstoneaudio.com or something like that)
JELabs amp (more of a traditional SET amp, with various iterations and lots of fans)

Maybe not quite as simple, but still not hard...
Loftin-White (zen simple in parts count, but IMHO, not simple in error-proof-ness...DC coupling can bite you if you mess up.  But follow proven schematic to a T and it is super simple)
DRD amp (maybe not quite as easy, but still pretty simple IMO to follow the proven examples out there)

Anyway, just a few good examples of SET amps out there beyond the plain vanilla that aren't hard to build and/or inexpensive (DRD is arguably not inexpensive but can be).

45's are very good sounding but not very powerful.  1.5 watts won't fill a big room or play loud/complex music on most speakers.  2A3 is an option for 3 watts and there are numerous modern remakes.  6B4G is an octal base version of the 2A3 with 6volt filatments.  300B is 8watts.
But one can get good results with a triode strapped KT88, EL34 for less money.  If it were my money and I was tight on funds I'd rather opt for a KT88 and good iron, than a 300B and so-so iron.

more to follow...

[JoshK]

We got to discussing about eschewing expensive DHT's and building a straight forward and reasonably priced SET amp using tubes from the beam pentode family (KT88, 6550, EL34, KT66, et al).  It got me really interested on a few levels.  1) I wanted to build a SET amp because I am curious, 2) I want to build something a bit simpler before embarking on my Karna monoblocks, 3) I have a lot of circuit ideas I want to try out, and I am sure there are a few other reasons I can think of.

Well I decided to actually pick up some iron and build something.  This is a joint endeavor with Dmason who also wanted to build a different kind of SET amp, with a bit mo powa then flea powerd SETs.  This is to be an effort to build an affordable, moderately powerful, elegantly simple and non-shi-shi SET amp.  Forget audio jewelry and audiophoolery, let's build a serious and practical SET amp using reasonable good and affordable parts, nothing unobtanium or uber priced, and no super high voltage.

I'd like for this to be share here with AC.  I am planning to take lead on the design and testing but I am very willing to open it up for feedback and parrallel testing and flavoring. 

The good doctor and myself have decided on a lot already and discussed a lot of ideas already.  Above was my semi-attempt at catching everyone up.  Maybe Dan can help fill in any important details or help clarify the goals in case I've misstated them. 

So far we have picked the KT88-oid, specifically upon gathering some feedback from a number of sources we have decided to pick the Tung Sol new production 6550.  Apparently its da bomb.  So until the amp is built and we can tube roll to confirm our choice it will be the tube that we design around.

There are a number of good schematics out there for 6550/KT88 SET amps, so this project is about putting the pieces together, trying a couple of them out or making up our own, testing which we like and go from there.  Some really good candidates to take inspiration from are Mikael Abdellah's SE KT88 project which has got high accolades.  There is also George of tubelab.com's SimpleSE amp.  And I really wanted to try out the Aikido + SE output stage amp. I have the parts to build up all three versions to compare, which is why I grew so interested in the project partly.

Hopefully we can document the build process and share parts selection and choices enough for even semi-beginning DIY'ers can build this amp.  Of course, you can just buy George's SimpleSE PCB which apparently comes with detailed documentation and is geared for first time tube builders.  This is likely not to be as organized, as you can probably already judge for yourself if read this far.  But I hope to share schematics, part choices, PSUD screen shots, files, layout diagrams, and hopefully lots of pictures!

There are quite a number of good choices for iron.  I jump to the conclusion and tell you that Dan and I chose the James 6123 for this project.   They seem to be the sweet spot of performance and costs.  If you plan to build this amp and it is a bit of a risk for you, you'd be well advised to pick the also well regarded Edcor transformers that are less expensive and upgrade later when you have reached build success.  That is just good DIY practice, build cheap, then upgrade later when you have a working unit. Of course, I often don't follow my own advice. 

We plan to use tube rectification.  The 5U4GB is an obvious choice, but I plan to use the 5R4GYB/5R4WG.  It is still relatively cheap and easy to obtain and has a great reputation for sounding better than shi-shi audio approved rectifiers.  The really good ones are often not very attractive though, but you can hide it behind the power transformer, we are going for sound here.

We will skip over electrolytic caps and choose motor run caps instead. No reason not to.  HVAC supplier are a great resource for obtaining reasonable cost motor run caps.   We also like chokes in the power supply and reasonable cost options will be shared when finalized. 

I will most likely share details on any/all of the above three choices.  I recommend the PCBs from George and/or JB if you choose those options.  I will be using PCBs (from tubecad.com) for the Aikido and point-to-point for breadboarding George's.  This is because I have the Aikido PCBs on hand and I have the parts for George's.  If I chose to use George in final form, I'll too buy his board.   I like building p2p but PCBs make it easy, simply the layout decisions and minimize risk of hum and frustration.  Plus it is good to support guys like John and George who offer so much to our small community.

Sorry if this is a bit scatter brain but I am trying to throw out all that we have been discussing.

First order of business if firming up our operating points in the output stage.  Of course we can just copy those of designs we are borrowing from, but I have tubecad and should also practice coming up with the points myself.  After that, we will work on the details of the various front ends and then details for the PSU.  I already worked up a basic version of the PSU in PSUDII, by guessing the ballpark B+ we will be working with and pending real measurement of the power transformer (source impedance is important and needs to be measured).   

We ordered power transformers already, which is a bit backwards.  We should have waited but we have enough room to wiggle and we came upon a really good candidate.  surplussales.com has a vintage Chicago transformer that is potted and has a number of other taps, etc.  It just looked too nice for a good price so we chose it. I think it is the PCR360 or something like that.

Well that should get us started....

[JoshK]

Some sources for parts we already picked.

Euphonia Audio has James 2123H $205/pr

New Quest Sound is where I got my tubes.  Jason has good prices and a good selection of modern production tubes.  He sells on ebay and a'gon. 

www.tubecad.com  For aikido pcb's. JB has tubes too. His Aikido kits have a great selection of good parts.
www.tubelab.com  for George's SimpleSE PCB.

Surplus Sales has the Chicago PTx. I have mixed feelings sharing the link to the PTx at this point because we haven't finalized the PSU, so order at your own risk.  We could change. (they also have lots of other parts for amp building, including sockets, wire, meters, etc).  I picked up some Litz wire from them for ground wiring.

[JoshK]

Mikael Abdellah's SE KT88 amp

http://www.tubelab.com/SimpleSE.htm

Aikido+300B in Blog 64

The point here is the it is very similar to what we are after... save for a substitution of a KT88 in the output which is straight forward.

Aikido + PSE KT88 with free heater voltage in blog 65

[jrebman]

Thanks Josh for the links and ideas.  Keeping all options at the moment because it will be a while before I can build anything.

-- Jim

[nodiak]

Hi Josh,
Nice to see this post, and appreciate the documentation of ideas, etc. I have the SimpleSE board and may get to it this winter, this helps quite a bit with inspiration as I've never built an amp before. I have contemplated parts and will put the James 2123H on the list. I have been thinking of having the option of UL tho, to hear what that would do. I've liked what I've heard about the TS 6550 too.
Will definitely keep up with this thread.
Don

[jrebman]

Don,

You should also consider using the preamp cathode bypass switch option, though I myself would probably opt to not use the tube/SS rectification option switch.  I've got one of these boards too, but won't get around to building it for a while yet.

Good to see you back.

-- Jim
 

[JoshK]

Hi Josh,
Nice to see this post, and appreciate the documentation of ideas, etc. I have the SimpleSE board and may get to it this winter, this helps quite a bit with inspiration as I've never built an amp before. I have contemplated parts and will put the James 2123H on the list. I have been thinking of having the option of UL tho, to hear what that would do. I've liked what I've heard about the TS 6550 too.
Will definitely keep up with this thread.
Don

The James 2123H has a screen grid winding for UL.  So it's a matter of adding a switch to flip between UL and triode.  I also like the cathode feedback option George wire's in.

Jim,
I am not sure what you meant by preamp's cathode bypass switch.  Is he adding a switch for bypassing the cathode's resistor with a cap, in the front end?  I didn't look at his schematic that closely, I was mostly looking at his "topology" if you will.

[JoshK]

I've been adding up parts cost.  I am up to around $500 for good quality parts, that includes the tubes, the iron (PTx & OPTx's), the sockets, PCBs, some wiring.  Add to that another $50 for caps for PSU caps, $20-30 for PSU chokes and whatever else for all the resistors and caps in the circuit (probably around $50).  So a rough guess is $650 not including chassis and connectors.   I have some cost effective and attractive options for chassis.  I see the entirety ending up under $800 for all really quality parts and no cutting corners. 

I am keeping a parts spreadsheet with costs and sources that I will share when it is filled out more.

I am not really trying to hit any price target, I am just eschewing extravagent parts and choosing high performing and reasonable costing parts (like James instead of Tamura, TS 6550 instead of WE 300B, etc).  In the end I wouldn't be surprised if this will go toe to toe with just about anything out there in the SET world. 

I have a small box full of Russian military caps that are built like you don't find today.  They also are fairly highly bespoke.  I am willing to send out a loaner kit of different types and values of caps for anyone reporting back.  I can even include a soniccap and an auricap in the experimenters kit, this way a number of people can comment on some of these more critical part choices.

[jrebman]

Josh,

Without being able to see the schematic myself, I'm guessing it is the same thing that Eddie Vaughn uses in the Carina -- which switches a cathode bypass cap in or out of the gain/driver tube circuit.

Just gives you another voicing option (BTW, I like the idea of twin coupling caps in the Aikido -- either to switch between a teflon or a PIO (as an example), to use one position as a bypass for the other, or to have each cap go to separate outputs -- say rolling off one at 80 Hz or so.

I'm liking this Aikido thing more and more .

-- Jim

[nodiak]

Hey Josh, that's good news about the UL abilities of the James 2123H, makes it a big contender for what I'd want.
Jim, I plan on tube rectifier, motor run cap, and probably the cathode feedback switch (as others have used and liked it, I still need to learn it's function).

[JoshK]

SimpleSE schematic

Found it.  I see a switch for cathode feedback, but I don't see a switch for bypassing/nonBP. It could just not be shown.

[jrebman]

Josh, I think we're talking about the same thing, just getting confused with the terminology.

Don, sounds like the configuration I'm going to go with also.

So, now to decide between the James, the Hashimoto, or the Magnequest Robin Hood series.... or then again, maybe something completely different.  I've got time to ponder this though.

-- Jim

[JoshK]

80uf 500VAC motor run capacitors
(no affiliation)

[jrebman]

Josh,

Just curious which Aikido board you are using?

The parts kits look like a very good deal too.

-- Jim

[JoshK]

I am leaning towards the octal Aikido boards because I am partial to octal tubes.  I am thinking 6SL7 into the octal/6H30pi (which Broskie sells).  That gives a gain of 35x, and given that I see biasing the KT88 at around -35 to -45, that should yield an input sensitivity of around 1v which I think it about right. 

I am doing this off the top of my head so I could be forgetting a 2 or something but

1Vrms input = 2.828Vp-p     [1Vrms*2*sqrt(2)=2.828Vpp]
2.828Vpp x 35 (gain) = 99Vpp
Let's say we bias at -39V which is what Mikael used. That is ~ 2x39 = 78Vpp for full output.  Now this is a rough calculation because it isn't really symmetric always.
So 78Vpp/99Vpp = 0.79 => 0.79Vrms input sensitivity.

Someone double check that or I will later when I dig out my copy of Morgan Jones' Valve Amps.

I think 1Vrms input sensitivity is an appropriate goal, since most sources have at least that much output, so extra output would yield headroom in the gain block so the output clips first.  You wouldn't need gain in a preamp, so a TVC would be appropriate.

[Bill Baker]

I am leaning towards the octal Aikido boards because I am partial to octal tubes.  I am thinking 6SL7 into the octal/6H30pi (which Broskie sells).  That gives a gain of 35x, and given that I see biasing the KT88 at around -35 to -45, that should yield an input sensitivity of around 1v which I think it about right. 

 I was thinking the same thing the first time I read this post. In my opinion, I think many would enjoy the 6SL7 or 6SN7 over the 12AX7 variety. The 12AU7 tube is a nice sounding tube in comparison to the 12AX7 but I think the 6SL7 is a better choice. This is not to say that the 12 series arrangement would not sound terrific.

 I considered building a few of these circuits some time ago but was never able to come up with the available time.

 Going with the 6SL7 into the 6H30 is a good choice IMHO.

[JoshK]

Oh, and as to choices for 6SL7, I have a small collection of different types on hand.  The NP TS 6SL7 is suppose to be fantastic, so I picked up a pair to try (~$30/pr).  I also have a few 1579's, which are a Russian military version of the 6SL7 that are apparently the bomb.  They are known to have really low noise/microphonics and great sound.   They aren't expensive *yet*, often around $15/pair or sometimes less.  The 1578's are, which are the Russian military version of the 6SN7.  They go for over $100/pair typically.

The octal 6h30pi is made by Electro-Harmonix, sometimes called the 6SN7X (which is what I will call them in this thread for less confusion).  If you are wanting to build point to point and don't care about mixing octals and novals (9pins) then I'd choose 6SL7 and 6N6P.  The 6N6P is almost the same as the 6H30Pi**, except the 6H30pi is well known and expensive and the 6N6P isn't.  The 6N6P is known to be more consistent from valve to valve and have really tight matching triodes within valve.  Many builders like it better than the 6H30pi and 5687 variants. 

The 6N6P and 6H30 are the russian version of the 5687 family, more or less, but they have the more usual pinout of tubes like 12ax7 and 6922, instead of 5687's pinout.  So if you use the noval PCBs and want to try 6N6P you don't need to get the 5687 PCBs, the regular one is the one to choose.

**annoying inconsistency; 6N6P and 6H30pi are the common markings english speakers use for these tubes.  Except all these tubes have the same cyrillic characters but for some reason we chose to translate some one way and some another.  E.g. 6Н6П is the 6N6P, 6Н30П is the 6H30pi, 6Н1П is the 6N1P, etc.  Annoying.  I like the _N_P lettering personally because it is simpler. 
6Н8П is the basic russian version of 6SN7
6Н9П is the basic russian version of 6SL7

[JoshK]

I am leaning towards the octal Aikido boards because I am partial to octal tubes.  I am thinking 6SL7 into the octal/6H30pi (which Broskie sells).  That gives a gain of 35x, and given that I see biasing the KT88 at around -35 to -45, that should yield an input sensitivity of around 1v which I think it about right. 

 I was thinking the same thing the first time I read this post. In my opinion, I think many would enjoy the 6SL7 or 6SN7 over the 12AX7 variety. The 12AU7 tube is a nice sounding tube in comparison to the 12AX7 but I think the 6SL7 is a better choice. This is not to say that the 12 series arrangement would not sound terrific.

 I considered building a few of these circuits some time ago but was never able to come up with the available time.

 Going with the 6SL7 into the 6H30 is a good choice IMHO.

Thanks for the comments Bill.  Also nice to get professional confirmation!

[JoshK]

P.S. the reason for using the 6H30pi/6N6P/5687 family in the second spot is more than the obvious higher drive ability.  Its true that the quiescent current for biasing these tubes is higher than the 6SN7 family so it gives more current availability for driving the output grid when it is close to grid current (A2).  Even more than that is that 5687 family typically uses a larger value cathode resistor which means it can swing more voltage linearly that many other tubes in the CF config.  Broskie wrote about this in one of his blogs. 

The oft complaint of the CF is that it sounds poor, off or whatever.  It is because the CF doesn't swing much voltage in a linear fashion, but with a high cathode resistor it can swing much more voltage linearly in CF.  The modified White CF that Broskie uses in the Aikido can swing quite a lot of voltage linearly, so there is no problem here.  The 5687 family in a White CF can do the job with aplomb.  So adding this stage isn't going to "color" the sound by its addition the way a SRPP stage would or a standard CF cascaded stage would.  It provides the low output impedance and drive to the output tube and shields the input tube from having to do this.  Win-win.

Then Broskie adds to all of this his noise cancellation, which is what many people think of when thinking of the Aikido.  Brilliant if you ask me.  It cancels out a tons of noise in the joint stage.  But this is in addition to a seriously kick ass gain/drive stage.  Low distortion, low noise, great drive.  Its hard to see how this can be topped, but George's CCS loaded driver has many of the same advantages but in a different package.  It also provides more gain typically.  We may or may not need it.