Home Theater Forum and Systems banner
1 - 20 of 21 Posts

·
Registered
Joined
·
1,752 Posts
Discussion Starter · #1 ·
Here is an amp I built from scratch about 10 years ago. It is based on a Loftin-White design first published
in the Radio News January 1929.

 

·
Registered
Joined
·
1,752 Posts
Discussion Starter · #2 ·
The actual circuit was done by my friend John Day on a scratch pad sitting around the dinner table during one of my audio get togethers.

 

·
Registered
Joined
·
1,752 Posts
Discussion Starter · #3 ·
I took his sketch and turned it into an actual schematic.

 

·
Registered
Joined
·
1,752 Posts
Discussion Starter · #4 ·
I cut the chassis top with some tools in my garage.

 

·
Registered
Joined
·
1,752 Posts
Discussion Starter · #5 ·
Layout is a lot of trial and error moving things around before I solder in the first piece. It all kind of goes together in my head.

 

·
Registered
Joined
·
1,752 Posts
Discussion Starter · #6 ·
The mess of wires and parts turned into this. It isn't perfect but it was the first tube amp I ever built.

 

·
Registered
Joined
·
1,752 Posts
Discussion Starter · #7 ·
The only part I didn't do myself was having the slots machined into the bottom plate.

 

·
Registered
Joined
·
1,752 Posts
Discussion Starter · #8 ·
All in all not a bad first tube amp. Now I must admit this isn't the first piece of electronics I ever built. I spent over 25 years in high tech electronics; from low earth orbit satellites to engine control systems and a whole lot in between.

 

·
Registered
Joined
·
1,752 Posts
Discussion Starter · #11 ·
Wow, first-rate work, really classy looking. Have you ever run any measurements on it?
No I have not. I'm not a big numbers guy. If I like the way it sounds, that's good enough for me. I expect this is about 3 ½ watts per channel and quiet enough for high efficiency speakers.
 

·
Registered
Joined
·
2 Posts
Hello,

This is my first post on this Forum. I have been designing and building tube amps since 1982. I want to address this to any "newbie" who might want to spend the time and effort to build such a beautiful amp as we see in this thread.

I think the chassis and execution of this amp is simply "breath-taking beautiful". If I were a "newbie" I would aspire to copy this circuit and execution - as best as I could.

Also, it is my personal opinion, for high sensitivity speakers, a two stage direct coupled amplifier, such as we have here, is the very best topology possible. Properly done, to my ears, it potentially offers the highest possible overall audio performance of any tube-type amplifier I have ever heard.

Despite this amp's decent audio circuit topology. it has design errorrs in several places, such that a "newbie" will get short-changed if they were to copy the schematic verbatim.

I would like people to avoid that. So, I will do a critique. I hope to not offend the original builder, as it is my intent to offer positve, easy, cost effective solutions to design aspects that troubled "me" in earlier parts of this thread. YMMV, but I doubt it !!

POWER SUPPLY :

The most important area of design of a 2A3 amp is its power supply. This amp's supply is inadequate to do a 2A3 tube justice !!! The original owner has "never heard fully" the circuit and what it can do, nor will a "newbie" if you copy it.

The very first things I was told by my audio mentor ( a musician and hi fi manufacturer ) in 1982, was "That you design the power supply first ". The second thing he then told me was this ..... " Chokes have to be 20 Ohms or less in DCR ( D.C. Resistance, measured with an Ohm meter ) IF you can find such chokes."

In 2014, we can find them !!

OK,... got it?? 20 Ohms or less.

The first choke in this amp's schematic is a Hammond 193J, which had a DCR of 83 Ohms. Verbotten. This 2A3 amp "loses" audio possibilties for high performance right THERE , sixty three Ohms too high, with this single part. Sorry about that!

To add insult to injury, the second choke, one for each channel, is a Hammond 159M, has a rated DCR of 256 Ohms, about twelve times too high. This combined high DCR is literally "choking the music delivery" to the 2A3 and to the speaker's voice coil, ...... it is hundreds of Ohms too high.

When this amp was designed, I do not think power supply modeling programs such as PSUD2 were very popular. The supply does not exactly "model" for me in a predictable way, in PSUD 2, so I had to fudge the input voltage of the power transformer some ( 550 VAC instead of 500 VAC, ten percent higher), and I am NOT fully confident and happy with my simulation ( recall the term GIGO ? ) but here is the best I was able to do :

Text Line Font Parallel Design



I have subjected this supply at 4.1 seconds to about a 10% increase in current, called a current-step. As you can see, the result is a wildly ringing supply that does not settle down fully until 800 mS ( milli seconds ) has gone by ( 4.9 minus 4.1 seconds ). This ringing, some audiophile EEs say, is one of the worst sounding and looking results you can get in PSUD2, and it should be avoided. Had the original designer done such an analysis, it might have given him some pause as a suggested design.

There is a second MAJOR ( IMHO ) flaw in this design. Very simple. The peak-current-hungry 2A3 finals stage is directly connected to the highly sensitive ( mu of 70 ) 6SL7 input/driver stage. The B+ fluctuations in the finals stage are directly feeding into the input/driver stage, smearing the amplification of the sensitive input/driver stage with fluctuations from the finals, a giant NO NO. This is very sub-standard and inadequate power supply design folks.

Because the two audio stages are not decoupled from each other, the original design strived towards obtaining low ripple, by not providing the PEAK currents and "on time" energy delivery to the finals - that only a low DCR supply can provide. The amp over-used the chokes, too many Henries and far too high in DCR. The original supply above simulates its B+ ripple at 19.78 mVAC .....feeding, unfortunately, BOTH audio stages.

After years of build practice, we find the sensitive input/driver 6SL7 stage needs to have one or two mVAC maximum of B+ ripple for ultra high definition and fullest musical expression. The 2A3 finals stage is the opposite, a LOT less fussy, anything under 700 mVAC of B+ ripple at the plate of the 2A3 is very acceptable for high sensitivity speakers. The finals ripple is stepped-down by the turns ratio of the output trannie, and we have had 850 mVAC supplies successfully run ALE speakers which are $$$$ 113 dB / 1 Watt/ 1 meter.

MY SUGGESTION :

Replace each of the three Hammond high DCR chokes ( having 338 Ohms of combined series resistances to each channel ) with inexpensive ( $25 to $35 each ) EDCOR XC75 -1.5HY-250 MA. model chokes. These EDCORS have a DCR of only 17 Ohms each !!!! This means a total series resistance, due to filter chokes only, of 34 Ohms versus 338 Ohms ( 82 plus 256 Ohms ). What this will do for bass and perceived dynamic contrasting will astound one upon listening to it.

Recall, I suggested we want ripple to the 2A3 plate to be 700 mVAC or less. In my suggested supply, it is simulated at 543 mVAC which is plenty good enough, believe me !! Here is what the suggested changes look like :

Text Line Font Parallel Design



Notice the much smoother response to a current step, and it fully settles in just 250 mS ( milli seconds ) instead of the original supply's 800 mS to full settle. ( 4.35 minus 4.10 seconds ). Also please note, the horizontal time scale for the second simulation is only 400 mS, whereas the first simulation required a 900 mS. horizontal scale to fully settle. The revised supply can play quickly occuring transient information ( AKA "music" ) back much better, more predictably and faithfully, unlike the wildly-ringing supply first simulated.

QUESTION, what about the mVAC ( ripple) to the input/driver 6SL7 stage ??? Very simple, decouple the two supplies with a R/C filter, as has been done for decades by ( Robin & Lipman=1947, then later - Asamu Asano, Nobu Shishido, Dennis Fraker, Craig Uthus, Don Garber, etc. ) The R or resistor value will be in the tens of thousands of Ohms. and the C or capacitor value will be 50 uF, possibly one of the new ultra low ESR Wima DC-Link film capacitors, which sound wonderful. It is easy to obtain 1 to 2 mVAC at the input/driver stage's B+. which is a factor of ten better - for the ultimate in audible performance, compared to the original design's supply.

If people need help with such a design, feel free to contact me privately and I will do the best I can to design / help, on strictly a amateur basis.

ONE OTHER THING : Audio circuit.

(1) The 2A3 as shown, is dissipating the typical 15 Watts continuous across its plate ( plate to cathode, 250 VDC times 0.06 A. ) . This thermally stresses the tube, wears it out in just a few thousand hours, and it sounds thermally stressed. Best practice, in terms of sonics and reliability combined, is to run a 2A3 at about 250 VDC at 42 mA. give or take one mA. It will sound better and last tens of thousands of hours while constantly putting a smile on your face. We have high performing AVVT 2A3 mesh plates ( 1999 ) that have 50,000 hours on them, and they still sound wonderful.

The 2A3 will specifically need to have either two paralleled Mills MRA-12 power resistors of 10 K Ohms or 11 K Ohms value, to obtain a 5,000 Ohm or a 5,500 Ohm Rk value. The stock amp's two 7.5 K Ohm paralleled resistors, equaling 3,750 Ohms is too low, runs the 2A3 too hard.

-----------------------------

I hope I have not offended the original builder, and helped some people here. Get out that soldering iron !!! Time to boogie with ALTECS and JBLs some !!

Jeff Medwin aka .......... drlowmu
 

·
Banned
Joined
·
4,838 Posts
Jeff (drlowmu),

Welcome to Home Theater Shack, and thank you for the informative post. I am not a DIY tube amp builder, but if I was, and found myself about to undertake a project such as this, your suggested design mods would be well received and appreciated. Thanks again for the contribution!
 

·
Registered
Joined
·
1,752 Posts
Discussion Starter · #15 ·
Thanks for the input Jeff. As I mentioned before, I built this over a decade ago from a schematic I was given. I am not a designer. This was actually the very first piece of tube gear I ever built. I was happy that it didn't smoke when I turned it on.

I may take your suggestion into account if and when I ever use this amp again. I am no longer into high sensitivity full range speakers, so don't use this amp anymore.

Thanks again
 

·
Registered
Joined
·
1,752 Posts
Discussion Starter · #17 ·
I gotta ask, what are you into now? Are the Jubilees for sale? :D
Jubs are still my mains in the home theater. That system is pretty much done for now. I'm selling off the dining room system (Martin Logan Prodigy's) and focusing on my main living room system and the sun room vintage rig. I moved the Roksan Radius 5 turntable to the vintage room to enjoy more vinyl time there. I have been building up my music server for my main system. I have about a terabyte of Flac, Wav and DSD files now. I'm really having fun with it. Still into big horns and tube power.
 

·
Registered
Joined
·
2 Posts
Thanks for the input Jeff. As I mentioned before, I built this over a decade ago from a schematic I was given. I am not a designer. This was actually the very first piece of tube gear I ever built. I was happy that it didn't smoke when I turned it on.

I may take your suggestion into account if and when I ever use this amp again. I am no longer into high sensitivity full range speakers, so don't use this amp anymore.

Thanks again
HELLO !

What concerns me now, is that you have posted full schematics up here and that some newbie may want to copy same. That would be unfortunate.

So let me say this, if anyone wants to build a two stage DC amp, have them contact me and I will do my best to give them a higher performing circuit to build. That is really all I wanted to say. I am always open to assist you or others, in the future, if there is any interest.

As I said in my original first post, nothing I know of will out-play the topology of a two stage direct coupled amp on sensitive speakers, but my assumption is that the amp is designed and built to a high overall standard, in ALL areas, layout, wiring, circuit wise, etc, etc.

Best to you.

Jeff Medwin
 
1 - 20 of 21 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top