|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.140.192.106
RH84 type test amp... for some reason I had the 12AT7 driver stage plates connected together and fed by a single wire from B+.
I decided to wire each triode section to B+ separately with its own wire to see it anything positive would result.
THD @ 1.2W power out dropped from .81% to .70%
THD+N @ 1.2W power out dropped from .88% to .82%
At 1W power output the THD = .60% , which compares favorably to the Mullard 3-3 in terms of power vs. distortion.
Follow Ups:
I was monitoring the AC line to see if changes to the mains voltage (120VAC) would cause changes to THD% at the output of my amp.
There was a dryer running on the same 50A subpanel as my amp.
Sure enough, if the AC line would sag a bit (about 2.5V) the THD% would increase by as much as .07 - .1% @ 1.2W power output.
The noise performance seemed to suffer a bit while the dryer was running as well.
Is it possible the increase in line noise is causing the analyzer to read an increase in THD? Floor noise will affect the harmonics more than the fundamental.
--------------------------
Buy Chinese. Bury freedom.
Hi DT, i have always connected the driver tube plates to seperate B+ dropping resistors. I felt that the circuit was more "balanced" that way and each section could operate more independantly since each half of the at7 drives its own power tube. It is reassuring to know that this arrangement results in lower distortion in the circuit. Thank you for exploring this. Kinda off topic but have you tried the rh circuit with octal driver tubes?
I have always preferred the 6sn7 and 6sl7 and other octal tubes over the 9 pins mini tubes. The 6sn7 seems to be one of the best sounding tubes around. The octal tubes have a sturdier base and aesthetically better appearing than the mini tubes. I have wondered if one were to optimize the RH type circuit for the 6sn7 or 6sl7 what circuit changes would be necessary? Have you tried any of those tubes?
Edits: 09/17/16
I have not tried separate R droppers + cap for the driver stage or measured to see if that would lower THD in the driver stage or at the output, but it would be an interesting experiment.
All I did was connect a separate wire to each plate resistor from a common B+ node vs. one wire from a B+ node to plate resistors that were jumpered together.
Not sure how that really isolated each channel anymore than they already were before I changed the wirings.
Maybe some low noise zener diodes from the B+ node to the driver plate resistors would provide better channel separation, but I have not tried that or know how it measures or sounds.
The report of lowered THD might have been the effect of variations in AC line voltage as opposed to the wiring changes.
Nor have I tried the octal drivers with EL84 finals.
I need to get everything back on a breadboard so I can try more things easier.
Thanks for commenting.
The 12AT7 isn't as linear as many other 9-pin dual triodes. Distortion characteristics can change considerably with a relatively minor change in operating conditions. Not sure if that's what happened in this case, it's just one more possibility in addition to the things mentioned below. Didn't you say earlier that the distortion in this amp was over 1%?
--------------------------
Buy Chinese. Bury freedom.
The THD was over 1.5% @ 1.2W the first time I wired up this circuit around a year ago.
Now THD is at .70% @ 1.2W with the circuit adjustments I have made recently , including the driver stage wirings.
You may want to review some of the earlier threads I had posted which noted other circuit/performance changes.
"The THD was over 1.5% @ 1.2W the first time I wired up this circuit around a year ago"
No, you reported over 1.5% just three weeks ago. That's why I was surprised to see these numbers suddenly looking better. What happened in-between?
--------------------------
Buy Chinese. Bury freedom.
My first attempt with an RH type circuit was critiqued by the designer a long time ago.
All those posts are still in the archives.
More recently, I started looking at the RH type circuit again and tried G2 regulation with gas regulator tubes, but the distortion performance and noise was worse than my earlier/first attempt.
After that, I posted a thread that documented some circuit changes that reduced distortion by about 50%.
Is this the circuit you're using?
Why would both triodes of the 12at7 be wired together?
Shouldn't there be a plate resistor (one for each triode) between the B+ and the plates?
Doesn't each plate get a FB resistor, each from it's own channels el84 plate?
I guess I'm confused.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
My circuit is based on RH84 but not exactly the same.
I posted the schematic in an earlier thread. Here it is again.
Each 12AT7 triode section has a plate resistor which was connected with a short piece of wire, then a single wire going to B+.
I cut the connection wire and used separate wires from each plate resistor to the B+ node.
Each channel has its own feedback resistor.
OK, now I understand what you meant.Shouldn't each driver stage have it's own drop resistor (labeled 1K) and decoupling cap (labeled 50uf)?
On second thought maybe not. The drop resistor is so small there wouldn't be much isolation anyway.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 09/15/16
I should have noted on the schematic that I used a common PS for both channels.
There is the 1K R dropper and 50uF cap that provides driver stage B+ for both channels.
Sorry for the confusion, this circuit is still a work in progress.
My guess is what you are measuring is actually less crosstalk. Crosstalk could manifest it's self as distortion on your analyzer.This is not a case of doubling wires in milliamp circuits making improvements. What you are now doing is providing a separate voltage drop circuit for each stage rather than sharing it.
That has some merit as you have measured.
Edits: 09/15/16
I wondered about crosstalk as well.
Thanks for the comments.
It could also be less noise with your modification as well.
Hard to say without seeing and working with your layout.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: