jcalvarez,
The original mono-plate 2A3 with 10s of filament wires is the most sought after.
The thin filament wires were very fragile, hard to produce, easy to break.
Then along came the 2-plate 2A3.
Only some modern 2A3 production 2A3 are mono-plate again. But there are only a few filament wires; not the multiple 10s of filament wires.
Some modern production 2A3 have 2 plates.
It is interesting you asked about this.
A physical study of a mono-plate 2A3 shows it to be a 'parallel' tube (Left side and Right side).
A problem exists when the spacing of filament to grid to plate on one side, does not match the spacing of the filament to grid to plate on the other side (un-equal parallel parts spacing of the tube).
The dual plate possible problem is even more evident . . . are the 2 structures Exactly equal?
Because these parallel elements are all inside the glass envelope, you can not individually bias them . . . only 2 filament connections, 1 grid connection, 1 plate connection.
And, you can not measure them separately to see if they are matched (no separate connections of what is in a single glass envelope).
The original mono-plate 2A3 with 10s of filament wires is the most sought after.
The thin filament wires were very fragile, hard to produce, easy to break.
Then along came the 2-plate 2A3.
Only some modern 2A3 production 2A3 are mono-plate again. But there are only a few filament wires; not the multiple 10s of filament wires.
Some modern production 2A3 have 2 plates.
It is interesting you asked about this.
A physical study of a mono-plate 2A3 shows it to be a 'parallel' tube (Left side and Right side).
A problem exists when the spacing of filament to grid to plate on one side, does not match the spacing of the filament to grid to plate on the other side (un-equal parallel parts spacing of the tube).
The dual plate possible problem is even more evident . . . are the 2 structures Exactly equal?
Because these parallel elements are all inside the glass envelope, you can not individually bias them . . . only 2 filament connections, 1 grid connection, 1 plate connection.
And, you can not measure them separately to see if they are matched (no separate connections of what is in a single glass envelope).
Thank you 6A3sUMMER! Interesting, I never thought about differences between both sides of the single plate 2A3.
Funny! Good to know we’ve all been there at one point.
I only dropped the voltage to the filaments
The physical construction of the 45 is interesting:
The grid spiral extends beyond the top of the plate, and beyond the bottom of the plate.
(I have National Union and RCA Cunningham 45 tubes).
The only other DHT that I have with the grid spiral extending above and below the plate is . . .
The dual plate Valve Art 2A3.
Your DHT grids may vary.
Fun with tube constructions.
The grid spiral extends beyond the top of the plate, and beyond the bottom of the plate.
(I have National Union and RCA Cunningham 45 tubes).
The only other DHT that I have with the grid spiral extending above and below the plate is . . .
The dual plate Valve Art 2A3.
Your DHT grids may vary.
Fun with tube constructions.
On to the power supply...and more stupid!
First step ditch the Chang 56 uF caps.
While removing a cap the entire tag strip snapped in half 🫤. Again, this is a junk drawer build… All the bits and pieces in this thing are extremely old and maybe brittle. Complicating matters the 14 awg ground bus winds its way through a bottom hole on the lug making replacement challenging. Decided to cut the bus off where it exits the lug and bring individual ground wires back to the new lug from that point forward.
The mounting lug for the new 4 way tagstrip was on the “wrong side” for the tight corner…so I thought I’d be clever and bend it under to the other side. Pleased with my creativity, I soldered up as much as I could before mounting. The I went to mount the strip. Of course, having bent the mounting lug to the other side, it was now much shorter…leaving not enough space for the nut. Back to square one. Took a five place mounting strip, and carefully trimmed one lug off the “correct end”, so the mounting lug would be oriented correctly. Back in place, new ground wires run and second filter cap now a Rubicon 86uF. The first cap, swapped to a 24 uF Solen film, was moved over to the rectifier tube as it was too big for original mounting location. Still oversized for the rectifier tube, but better than 56uF. A compromise. Didn’t want to get too small and increase ripple or lower the already limited B+.
Next step was a new Hammond choke. 7H 150mA. Chosen largely for physical size to fit within the chassis—more than double the dimensions of the old choke. The bottom of the choke sits slightly proud of the mounting tabs. Figured they would bend down a little bit when I tighten the screws. Mounted and wired up. All seemed well until I turned the chassis back over. Instead of bending the choke tabs up to the chassis, the soft aluminum top had dimpled down around the screw heads. Lesson learned shims/washers under the mounting tabs.
Power up went well. Still a slight ear to speaker hum…maybe ever so slightly less and a little less harsh tone than from the unmodified mono block. I might up the second cap to 100uF or maybe even surface mount a motor run cap. There are also some strange grounding decicions. It appears half of the 6SN7 tag strip is not grounded through the bus bar but through the aluminum chassis only. I’m wondering if that was intentional due to hum issues? Going to experiment with with ground connections using jumper cables.
I also tidied up some of the filament wiring in the previous step and forgot to mention.
more stupid rookie tricks to come!
First step ditch the Chang 56 uF caps.
While removing a cap the entire tag strip snapped in half 🫤. Again, this is a junk drawer build… All the bits and pieces in this thing are extremely old and maybe brittle. Complicating matters the 14 awg ground bus winds its way through a bottom hole on the lug making replacement challenging. Decided to cut the bus off where it exits the lug and bring individual ground wires back to the new lug from that point forward.
The mounting lug for the new 4 way tagstrip was on the “wrong side” for the tight corner…so I thought I’d be clever and bend it under to the other side. Pleased with my creativity, I soldered up as much as I could before mounting. The I went to mount the strip. Of course, having bent the mounting lug to the other side, it was now much shorter…leaving not enough space for the nut. Back to square one. Took a five place mounting strip, and carefully trimmed one lug off the “correct end”, so the mounting lug would be oriented correctly. Back in place, new ground wires run and second filter cap now a Rubicon 86uF. The first cap, swapped to a 24 uF Solen film, was moved over to the rectifier tube as it was too big for original mounting location. Still oversized for the rectifier tube, but better than 56uF. A compromise. Didn’t want to get too small and increase ripple or lower the already limited B+.
Next step was a new Hammond choke. 7H 150mA. Chosen largely for physical size to fit within the chassis—more than double the dimensions of the old choke. The bottom of the choke sits slightly proud of the mounting tabs. Figured they would bend down a little bit when I tighten the screws. Mounted and wired up. All seemed well until I turned the chassis back over. Instead of bending the choke tabs up to the chassis, the soft aluminum top had dimpled down around the screw heads. Lesson learned shims/washers under the mounting tabs.
Power up went well. Still a slight ear to speaker hum…maybe ever so slightly less and a little less harsh tone than from the unmodified mono block. I might up the second cap to 100uF or maybe even surface mount a motor run cap. There are also some strange grounding decicions. It appears half of the 6SN7 tag strip is not grounded through the bus bar but through the aluminum chassis only. I’m wondering if that was intentional due to hum issues? Going to experiment with with ground connections using jumper cables.
I also tidied up some of the filament wiring in the previous step and forgot to mention.
more stupid rookie tricks to come!
Last edited:
Double plate 2A3 is not correct in my opinion. The plate is still one but it has different shape (H-Shape) respect to the original single plate. The only double element is the filament (2 V-shaped filaments that form a W). The 6C4C (and the rare 2C4C with 2.5V filament) is the best technical implementation of the concept. Stronger, very durable and high performance. The Russians just made it better.
Repeating power supply update for second monoblock. Meanwhile, trying to get smarter on plate curves and operating points before jumping into changes to the driver section. Lots of charts and graphs. More to come.
The technical shortcomings of the “JC Morrison” design have been discussed.
What about the other side? The designer made some specific choices they obviously perceived as beneficial to the voicing.
Surely the designer, agree or disagree with the results, set out to achieve something.
Whether they fell short or not, what was the specific goal here?
I’m trying to get inside of the designers thought process. Not what the results are or should be or could be. But, what was the designer trying to achieve and how did they think this design would move toward that goal?
What about the other side? The designer made some specific choices they obviously perceived as beneficial to the voicing.
Surely the designer, agree or disagree with the results, set out to achieve something.
Whether they fell short or not, what was the specific goal here?
I’m trying to get inside of the designers thought process. Not what the results are or should be or could be. But, what was the designer trying to achieve and how did they think this design would move toward that goal?
Last edited:
45,
I am not sure what 2A3 tubes you were talking about in Post # 75.
Perhaps I misunderstood your statement "Double plate 2A3 is not correct in my opinion. The plate is still one but it has different shape (H-Shape) respect to the original single plate.
A 2A3 double plate may be made out of only one continuous piece of sheet steel, true.
But, that sheet of steel is first folded over one filament structure and one grid; and secondly folded over another filament structure and another grid.
In my book, that is 2 plates, even if they are made from only one sheet of steel.
I am not sure what 2A3 tubes you were talking about in Post # 75.
Perhaps I misunderstood your statement "Double plate 2A3 is not correct in my opinion. The plate is still one but it has different shape (H-Shape) respect to the original single plate.
A 2A3 double plate may be made out of only one continuous piece of sheet steel, true.
But, that sheet of steel is first folded over one filament structure and one grid; and secondly folded over another filament structure and another grid.
In my book, that is 2 plates, even if they are made from only one sheet of steel.
Researching 6SN7 operating points I came across this on another DIY audio post:
I believe the 6J5 is electronically equivalent to 1/2 of a 6SN7?
.008% THD. Seems this tube can be operated at fairly low voltages with pretty good results? Am I missing something? I am learning and much of this is still foreign to me. Hoping somebody can help break down some of these numbers for me…ie V(p) vs V(PP) and the rest.
Thanks all.
I believe the 6J5 is electronically equivalent to 1/2 of a 6SN7?
.008% THD. Seems this tube can be operated at fairly low voltages with pretty good results? Am I missing something? I am learning and much of this is still foreign to me. Hoping somebody can help break down some of these numbers for me…ie V(p) vs V(PP) and the rest.
Thanks all.
It's worth remembering that valve curves are unreliable down near the edges. They weren't really measured down there, just sketched in with a French curve. You really can't believe them. Secondly, operation with the grid near 0 VDC means that some grid current will flow, more dramatically the nearer it gets to zero. Unless the valve is being driven from a zero impedance source, that very non-linear grid current will appear as a very non-linear grid voltage adding itself to the signal voltage.
Type 6J5, 6(and 12)SN7, 6CG7/6FQ7 are very linear amplifiers, but need to operate in their best range, which is much more volts and current than the drawn example.
All good fortune,
Chris
Type 6J5, 6(and 12)SN7, 6CG7/6FQ7 are very linear amplifiers, but need to operate in their best range, which is much more volts and current than the drawn example.
All good fortune,
Chris