... Get*SET*Go.....B+ FILTER improvements.....to the 6B4G Finals Tubes

dowto1000

Junior Member
Get SET Go amp, POWER SUPPLY - Part ONE : Reducing Series Resistances to the Final's tube, PART 1 of 3.


The very first useful thing I was taught, in 1982, about 2A3 type directly heated triode amps was, that the chokes in the power supply to the output stage had to be " 20 Ohms or less, IF I can find them ". This was sage advice in 1982, pertaining to push-pull tube amps, from my very first audio amplifier design Mentor, Robert W. Fulton ( 1925-1988 ).

http://fultonmusicalindustries.com/

http://fultonmusicalindustries.com/biography.html


Since 1982, I have followed and experimented with RWF's guidelines, in both the push-pull and single ended amp genres, and now, in 2017, I have refined his advice. In 2017, I PREFER hearing chokes that are only 8 Ohms or less in DCR, for directly heated triode tubes, to filter the B+ to the amp's Final tube… in any SET amp.

This past weekend at RMAF-2017, I spent three full days and I listened to SET amps, that had two 4 ohm DCR custom-designed Magnequest chokes in it, and - it was truly FUN to experience. There was a " freedom" to the sound, that one never can or will hear, otherwise !!

Hammond makes a 159ZA choke, for under $40.00, that has a DCR of only 6 Ohms, and it can be bought off the shelf, easily.

I have taken the "Get SET Go" amp, owned by a local audio friend in the Kansas City area, and by Thermionics on this Forum, and on 10-10-2017, theoretically applied ( simulated ) Hammond 159ZA chokes ( two ) to the to the "GSG" amp's design.

My goal was to permanently "fix" the Get SET Go's extremely low performing, totally misdesigned ( IMHO ) , stock power supply, by reducing SERIES RESISTANCES of it's power supply, as a start, as the very first step…. one ( of several ) that is needed. .

What I suggest herein, will transform the GSG amp, and ANY SET amp, as it plays music, into an actual speaker load !!

Step 1, calculating from a schematic, the stock GSG amp's total series resistances in the B+ filter:

Download the Get SET Go's Owner's Manual, here :

http://www.diytube.com/getsetgo/getsetgo_manual.pdf


Notice, the choke portion of the B+ filter ( L/C ) drops 6 volts across it ( 333-327 ) and the resistive element portion of the B+ filter to the Finals ( R/C ) drops 8 volts across it . ( (327-319 ). Using Ohm's Law, we can easily determine that the stock GSG stereo amp draws a TOTAL current of ( 90 Ohms ( R17a and R17b in parallel ) divided by 8 Volts ) 0.088888 Amperes….or 88.9 milliamperes.

Since the choke drops 6 Volts across it, at 0.088888 A., the choke ( using Ohm's law ) has a computed DC Resistance ( 6 VDC divided by 0.088888 Amperes ) of 67.50 Ohms.

The total series resistances of the "driven" parts of the STOCK GSG's B+ filter to the Finals are this: the choke ( 67.50 Ohms ) and the two resistors in parallel ( 90 Ohms ), or, combined, a total series resistance of 157.50 Ohms .

The design goal of a DHT . tube amp is to have adequate filtering, with as LOW a total series resistance as possible, in the B+ filter to the Finals ( Output ) tube.

The fact that the "Get SET Go" uses a 90 Ohms resistance, as part of a B+ filter to the Finals, tells ME, right off the bat, the designer is UNAWARE of what is being discussed here. One seldom ever sees, ( and likely should NEVER see ) an R/C filter in ANY filter to the Finals. The goal is to make the series resistance as LOW as possible, and not EVER to add 90 Ohms !!!

CHANGING the B+ filter topology:

GSG Stock is this , ( after the rectifier tube, to the Finals ):

C1/L1/C2/R1/C3, to the Finals …total series resistance … 157.50 Ohms

GSG Suggested is this ( after the rectifier tube, to the Finals )

L1/C1/L2/C2, to the Finals,…total series resistance……12 Ohms !!!



157.50 Ohms, VERSUS 12 Ohms .


This simple topology change brilliantly reduced unwanted series resistances to the Finals by a cool factor of 13.125 times !! We have gone from one of the worst, to one of the very best designs, in terms of lowest possible series resistances to the Finals..


Next question : Have the B+ voltages changed too much, after making TOPOLOGY changes ??

We would need to simulate the two B+ supplies, ( stock and the newly suggested one), to be sure that the amps' operating voltages are not substantially different or inadequate in any way.

STOCK SIMULATIONS : For many years, I have used PSUD2 to simulate stock and modified power supplies. In the case of the Get SET GO amp, unfortunately, PSUD2 does NOT have an entry for a rectifier tube, using the stock amp's 6AX5G rectifier tube.

Substituting arbitrarily a 6AX4G into PSUD2, as the rectifier tube, the GSG stock amp Finals' B+ simulates to 365.10 VDC at 28 mVAC. ( This is where the stock amp shows 319 VDC in the Owner's Manual. ) The Voltages and Ripple in PSUD2, for the INPUT section, simulates in the stock amp to 353.80 VDC at 0.169 mVAC. ( versus 300 VDC in the Owner's manual ).

Using the VERY DIFFERENT L1/C1/L2/C2 topology, ( with values of L1, L2 being 400 mHY at 6 Ohms, C1= 50 uF, C2= 60 uF ), we get the following SIMULATED B+ voltages, by ONLY changing the Final's Filter :

352.56 VDC and 758 mVAC to the Finals, and 341.23 VDC and 1.0059 mVAC to the Input section.

In each case, the voltage difference, stock versus modified, in PSUD2 is ONLY about 3.5 percent ( lower for the modified ).

I am SATISFIED that the proposed change is " close enough " so as not to materially effect the amp's operation.. The two 6 Ohm 159ZA chokes are " shoe ins " !

In a next a different - write up, we will discuss the GSG SET amp's power supply's capacitor choices, both in terms of types and values, suggesting the very latest in capacitor technology, which is both affordable and super-high in audio performance. To be continued. Hope I didn't bore you.

Input to this first part / write-up, is appreciated.

Dowto1000.

 
Interesting. The chokes would be fairly easy from a circuit perspective to add in. The only problem is that you'd have to completely redesign the amp layout as those chokes are huge compared to the stock build. Most likely you'd have to bury the chokes under the top plate, flanking the power transformer (currently where the OPTs are sitting) and then put the 6B4G tubes and output transformers on the other end of the amplifier a la Bogen's DS265. The other downside is then the input tube would be smack-dab in the middle of the chassis so it would be a long-ish run to the back side of the amplifier (you'd have to pick which side was "back" of course) - assuming you stick with the G*S*G board, of course.

 
Got 'ya. I wasn't considering your particular layout at all, sorry.

'Was only considering the existing GSG circuit and topology, and how to improve its most obvious design errors, as effectively as possible.

It would not be worthwhile to re-package that design, IMHO.

Since you are listening to it, and liking it, consider a new build, simply build a better design - this time, doing (1) a direct coupled two stage triode amp and (2) make use of a JJ 2A3-40 as the Finals tube !!

The reason for these two suggestions are as follows : because (a) cap coupling is the worst possible way to connect two audio stages, the exact opposite to a direct-couple and ( b) a JJ 2A3-40 with a beefy 2.5 Ampere filament ( think current delivery from the cathode, versus a 1 Ampere rated cathode ( 6B4G ) ). is the best overall choice in that range of tubes.

The JJ 2A3-40 avoids a bi plate design, has a mono plate structure, has symmetrical stringing of the filament ( compared to a " V" or a "W" stringing of the filament ), ALL these factors add up to a better-sounding output tube choice. A 6B4G, 6A3, 6A5G is not a 2A3, and any biplate 2A3 is never equal to a JJ 2A3-40 !!

I could assist you GREATLY with the design, parts and build, since I am DIY building a pair of mono 2A3 DC AMPS for myself, over next two months. Picked up the two 14 gauge steel mono chassis, with nicely welded corners, just yesterday afternoon. It can become your final destination in amps for efficient speakers !! It will be mine !!

Dowto1000
 
It'll be a while before I can commit to any new amplifier designs as property tax season is upon us. However, I will keep this in mind for a long-term solution (e.g. monoblock 2A3-40 design with your enhancements).

Thanks!
 
The biggest thing we need to remember about the Get*SET*Go design is that it was built to a price-point: an easy way for someone who is handy with a soldering iron and doesn't have a lot of money to spend to get into directly-heated single-ended-triode amplification. Could it have been built in such a way to have higher fidelity per Jeff's suggestions above? Of course, but that would have driven the cost up significantly (remember that at the time the G*S*G amp was offered, you could get a pair of 6B4G tubes for the price of a couple of lattes and a scone or two).

I am interested in (someday) taking all that has been suggested and maybe building a "Super Get*Set*Go" using the modified power supply, biasing, and coupling that Jeff has laid out in his three posts, but that will probably require a significant investment in an appropriately spec'd power transformer, and better output transformers than the (otherwise quite good) Edcor GSXE-series that I'm using.

-D
 
Thermionics :

The circuit of GSG is not worth your time and money to re-create, point to point in my opinion.

The GSG's use of a cap couple, in my opinion, makes it like so many of the other " not good enough " SET amps , even if you were to do ALL the changes I suggested.

I can NOT honestly recommend the GSG's point to point effort / expense. Build a better design !!!

The 6B4G, with its 6.3 VAC at 1 Ampere filament needs to be avoided.

It has a whimpy filament, cathode, compared to what a 2A3 tube, with a 2.5 Ampere cathode, has.

That 2A3's 2.5 Ampere cathode, when executed RIGHT, can deliver energy to to the speaker under music reproduction, that IS superior sounding than a 1 A. 6B4G tube. Also, you use AC heating, on a 2A3, which always, in my opinion, sounds better than DC heating of these Finals tubes.

You already HAVE a plenty good enough pair of Output transformers, the EDCORS, to do a 2A3 amp. You need to use a JJ 2A3 40, as the best possible 2A3 tube choice in my opinion, money versus design and long term performance - wise.

I can help you with the design, confidentially, privately, much like the four DC 2A3 amps we are doing this last quarter of the year. If DIYing, do the best possible design, make an amp which will LAST in your hi fi hobby.

Dowto1000
 
Last edited:
It'll be at least 6 months before I have time for another amp project (minimum). Work is far too busy, and I don't want to piss off my wife by spending a lot of money right now. But I appreciate the offer and may get back to you on this if conditions change.
 
Sounds logical to me. I will post photos of the mono JJ 2A3 40 build, coming up, for people to see.

Have fun, I am !!! Can't wait to do these mono amps.


Dowto1000
 
Last edited:
Thermionics :

The circuit of GSG is not worth your time and money to re-create, point to point in my opinion.

The GSG's use of a cap couple, in my opinion, makes it like so many of the other " not good enough " SET amps , even if you were to do ALL the changes I suggested.

I can NOT honestly recommend the GSG's point to point effort / expense. Build a better design !!!

The 6B4G, with its 6.3 VAC at 1 Ampere filament needs to be avoided.

It has a whimpy filament, cathode, compared to what a 2A3 tube, with a 2.5 Ampere cathode, has.

That 2A3's 2.5 Ampere cathode, when executed RIGHT, can deliver energy to to the speaker under music reproduction, that IS superior sounding than a 1 A. 6B4G tube. Also, you use AC heating, on a 2A3, which always, in my opinion, sounds better than DC heating of these Finals tubes.
I have a G*S*G board and as I’m planning to complete it in the near future I was reading about it. I stumbled upon your posts and was intrigued by them.

You stated that GSG circuit is not worth doing point-to-point. What about doing it with the PCB that I have? I shall appreciate hearing more about your preferred circuit that does not need a coupling cap between the voltage amp and out put tube. Could you provide a schematic? How do you deal with drift over time in the front tube while maintaining bias on the final?

You recommend the JJ2A3-40 and that seems to be a very good tube, but it is not a real 2A3. Just look at the specs for dissipation capability. I’m also a bit puzzled by your statement about the 2.5 amp cathode on the 2A3 being able to “deliver energy to the speaker”. Could you explain your statement, please. The 2A3 filament consumes 6.25 watts, the 6B4G filament consumes 6.3 watts.
 
Dowto (Jeff) got kicked off this forum long ago. He went from here to the Klipsch forum, where he was kicked off even more quickly. There are good reasons for why this continues to happen to Jeff, and you've already scratched the surface a bit with some of your questions.
 
Hi Paulbottlehead,

Thanks for the background. Sorry downto1000 is gone; I don’t mind learning from opinionated people as long as they can explain their opinions persuasively.
 
Back
Top