|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
76.5.127.109
I'm building a pair of push pull 2A3 amplifiers using RCA bi-plate 2A3s. The design I'm startimg with is Pete Millett's PP 6B4G, except with 2A3 tubes, with individual 2.5V 3A AC heater supplies (Hammond 166M5 - no center tap) and a 750 ohm 10W resistor between each cathode and ground.
As published, this will run each 2A3 at 15W dissipation (100% of RCA's dissipation rating). This seems to be pretty common for cathode biased push pull 2A3 designs.
Is the RCA 2A3 able to be pushed that hard without wearing out quickly? They are expensive, and I'd like to run them at a point that won't chew them up too quickly.
I know how to adjust the bias, but I'm still new to drawing composite curves.
Thanks for your input!
Follow Ups:
Wow! Short but excellent thread. Reminds me the old Sound Practices and VTV days. Special thanks to Dave for the lengthy response and history.
The 2A3 is the tube I settled into years ago and still my overall favorite DHT. I found the 45 to be a little more refined but did not have enough power. The 2A3 has just the right juice.
Cheers,
Jim
JJ 2A3 are single plate from what I've read and also sonically liked.
"At some later date a biplate or dual single plate version was introduced. This appeared to simply be two single plate versions in parallel inside the same bottle. Since adding tubes in parallel doubles the transconductance and keeps the µ the same the resultant rP would be 1/2 that of what a 2A3 should be. Parallel tubes also mean that each individual tube would need 1/2 the gm / double the Rp to meet the 2A3 specification when paralleled. This also means that the filament current and plate dissipation would also double. "
Nothing could be less logical than what you already came out with.
Paralleled valves have big advantages, especially in regard to lower shot (random) noise levels.
How do you have any idea, these valves didn't have the entire curves and spec specially re-arranged to make exactly the same parameters as the single anode versions?
Sheer speculation?
What follows is some folklore surrounding the 2A3. I am not sure of the exact chronology of the biplate versions so I will present them in the order that makes the most sense (vintage 2A3 pictures from http://audiobobinas.com/2a3/)
The original 2A3 had a single plate and an infamous "harp filament". Its introduction carved the 2A3 specification in stone.
At some later date a biplate or dual single plate version was introduced. This appeared to simply be two single plate versions in parallel inside the same bottle. Since adding tubes in parallel doubles the transconductance and keeps the µ the same the resultant rP would be 1/2 that of what a 2A3 should be. Parallel tubes also mean that each individual tube would need 1/2 the gm / double the Rp to meet the 2A3 specification when paralleled. This also means that the filament current and plate dissipation would also double. Cutting each tubes filament current in 1/2 fixes the current issue and the legendary harp filament of the SP 2A3 was changed to a more traditional VV filament structure. This lead many to speculate that this version of the SP 2A3 was simply a pair of 45's in parallel in the same envelope. Since plate dissipation is based on both plate structure and cathode composition it seems plausible that even though the plates could dissipate the power, the 6.25W of filament power could be the limiting factor. It could also be that the specification were simply etched in stone.
The biplate 2A3 with two smaller "300B" shaped plates that we are all familiar with has a different physical structure altogether yet still has the same specifications. Again the greater mass of the plate leads people to assume that it can dissipate more power and JC Morrison used to bark that it could easily be run at 20W all day long.
Flash forward 60 or so years and sovtek came out with a new monoplate 2A3 based on their 300B tooling and the current production 2A3 craze was started. JC Morrison worked for Mike Matthews at New Sensor and had some insight to this tube. Given the 300B plate can dissipate 40W and the 300B filament had the same power as the 2A3 I queried him why the 2A3 was a 15W tube and he said partially due to the etched in stone nature of the 2A3 spec and partially because plate dissipation is much more complex than simple plate area and cathode power and from memory said that 25W was about the maximum the tube could do to avoid current saturation at the dynamic peaks.
After the sovtek a whole slew of other 2A3's of various constructions came on the scene and the only thing seemingly in common is the etched in stone 15W number.
Since I am teetering the line between a nice bedtime story and an urban legend I'll stop here.
dave
...
Back in the day routinely. The general consensus at the time was the vintage tubes could easily handle it and the chinese tubes were only $5 so who cares.
dave
...
Thank you, Dave. That gives me a little more confidence. That's interesting information!
You also provided some missing information of the odd dual single plate 2A3s that I've been wondering about. I have a pair labeled CBS Hytron.
"because plate dissipation is much more complex than simple plate area and cathode power"
Indeed, because the heat must be transferred by radiation the glass bulb area is also important and the transparency of the glass for IR is also of influence.
For a little bit more money any electron tube could be a lot better* by using a better glass bulb.
* longer life by cooler temperture
In my building, I try my best to follow the rule published by RCA and repeated by Morgan Jones that power tube envelopes should be spaced from one another by 1-1/2 times the diameter of the envelope. I see a lot of push pull 2A3 amplifiers that don't apply this rule. I'm not sure why.
Because they had not a good designer / don't care about lifetime of a tubeMost lickly, they simply don't know.
Edits: 10/25/20
Here's a couple data points:
* 2A3 reference operating point is 250v @ 60mA - equal to the maximum spec.
* 300B reference is 350v @ 60mA - 21 watts; max spec is 36 watts (1950 data sheet values)
Either they were rated using very different criteria, or the 2A3 spec is much more conservative. Or possibly both. For what it's worth, almost everybody uses the 2A3 at 15 watts, while almost nobody uses the 300B anywhere near 36 watts.
Thank you, Paul. That's interesting information, in case I want to modify the amp to run 300Bs in the future (with different OPTs).
Years ago I owned, then sold to fund curiosity for other amp designs, a pair of Paramour II amps. I enjoyed them very much, and regretted selling them. The memory of that 2A3 is what is propelling this build, though I understand the sound may be of a different character. Plus, this time I'm using Cornwalls.
easely 15W continous disipation.
Thanks!
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: