|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
83.187.70.153
Hi all,
I need your help to finish my PP amplifier design. So if you have comments to upgrade it I appreciate.
Thanks in advance.
Follow Ups:
Hi all,
As every people advise me to build an NFB amp so why not build one amp with tons of NFB (something like in Peter Baxandall papers says 40 dB).
Do you have such schematic?
It’ll sounds good?
PS: isn’t a free project. Amp has a chassis with 8 holes (4x EL34, 4x 9 pins sockets).
> > As every people advise me to build an NFB amp so why not build one amp with tons of NFB (something like in Peter Baxandall papers says 40 dB). < <
LARGE amounts of NFB that include the O/P trafo in the loop make the amp highly vulnerable to both O/P trafo core core saturation and slew limiting. :> ( Think of the GARGANTUAN error correction signals involved at both frequency extremes.
What you want to use is cicuitry that's reasonably linear open loop. Then, a few dB. of loop NFB will improve damping factor and linearize your questionable Chinese O/P trafos. NFB is a tool that can easily be abused.
Eli D.
Is Baxandall chart only to SS?
hey-Hey!!!,
Remove the entire 12AU7 stage. The first stage is OK, but I'd suggest some changes. Up the current by lowering the plate loads to ~30k. You can use any valve with that pinout. 6CG7, 6BK7B, 6H30, 6GU7 to start with.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Hi,
Just to clarify isn’t a free project. Amp has a chassis with 8 holes (4x EL34, 4x 9 pins sockets).
Now things make more sense.Unless SUPERIOR quality O/P trafos are used, stability can be a MAJOR issue in the Williamson circuit. Please keep in mind that 6SN7 sections and GLORIOUS Partridge "iron" were used in the original.
All triode and no loop NFB can work well. However, I think you may be selling yourself short by excluding loop NFB and UL "finals". Your Chinese O/P trafos just might benefit from a little linearization. The Mullard topology (schematic attached) has fewer issues with phase shifts than Williamson topology has. So, the risk of instability is reduced. There is way too much gain in Mullard's complement. You will do quite well with 1 section of a 12BH7 in the voltage gain position and a 12AT7 in the LTP position. Wire 1 channel to use system 1 of the 12BH7 and the 2nd channel to use system 2. All the sockets are occupied and the total heater draw of the amp will be unchanged. When the voltage gain tubes show wear, simply exchange them between the 2 channels.
Oh yeah, you don't have to use Mullard's cathode bias scheme. "Fixed" bias WORKS in that style of circuit.
Eli D.
"You will do quite well with 1 section of a 12BH7 in the voltage gain position and a 12AT7 in the LTP position."
If i do that could i remove cap between first and second stage?
> > If i do that could i remove cap between first and second stage? < <
Look at the 5-20 schematic again. The voltage gain tube IS directly coupled to the phase splitter. ;> )
What we are talking about is retaining the overall topology Mullard devised, but changing the small signal types. A 12BH7 section is adequate in the voltage gain position, when the I/P signal is 2 VRMS, which is what a CDP delivers. The 12AT7 is (IMO) SUPERIOR to the 12AX7 as a LTP phase splitter. The 'T7 has low Rp and high gm, along with a high mu. Those characteristics are DESIRABLE in that position. The distortion spectrum of the 'T7 is an added benefit.
Hop over to TDSL and study the data sheets for the 12BH7 and 12AT7. You want to set the 'T7 sections up with 200-220 V. on the plate and Ib = 3 mA.
Eli D.
How to improve it more?
You could start by improving the resolution of the schematic - it's impossible to read most of it.
> > What we are talking about is retaining the overall topology Mullard devised, but changing the small signal types. < <
and now?
I'm having a hard time reading some of the parts values in the schematic. Age taking its toll, I suppose. Please use AA EMail to send me an image that's scaled larger.
What are the bias arrangements in the OEM Chinese circuitry? How many adjustments, if any, are provided?
I don't know if you can put a CCS in the LTP's tail, when the grids are positive, with respect to ground. The cathodes must be at a yet higher DC potential, in order for proper biasing to occur.
Point blank, the 12BH7 is more linear than the 12AU7. You can use the parts values you calculated for the 12AU7.
Have you given thought to providing for toggling between triode and UL modes? Just switch the 100 Ohm g2 resistors between the EL34 anodes and the UL taps.
Pay close attention to the phasing of the O/P trafo. The 3 tube stages invert the I/P signal. So, you must wire the O/P trafo to invert too. When NFB is applied to the cathode, as is the case here, the signal must be in phase with the I/P signal.
Eli D.
You have email.
You have PM.
You need resistors from the grids of the 12AU7s to ground - it looks like maybe they were there at one point and then accidently removed.
What is the input tube - I can't quite read the schematic. ECC???. FWIW, I much prefer octals over 9 pin tubes. I would replace the 12AU7 with a 6SN7. Kowever, as others have mentioned, it may not be needed at all.
ECC88.
Why so much gain? Mike Samra's remark about the 12AT7 is (IMO) spot on. The differential splitter/driver used in "El Cheapo" (schematic attached) has ALL the gain needed to drive no loop NFB PP triode wired EL34s, when a CDP is the source. Just ground the non-inverting triode's grid.
The 'T7 is EXTREMELY well suited to PP circuitry. The NET harmonic distortion spectrum is a PLEASING "waterfall", with 2nd > 3rd > 4th.
Eli D.
The first help I can give you is to get rid of that 6922 up front in the first stage..You have el34s and you don't need that hard sounding highly microphonic tube there when there are tons better choices.A high transconductance there wouldn't be necessary because I'm sure you intend on using the amp with a preamp correct? I would use a 12ay7 there or even a 12at7 but your circuit is pretty conventional so it wont be a problem.
If it was my amp,I would be using an ef86 up front even if it is a pentode,its very good and my eicos work extremely well with it.
I would suggest a smoothing cap from the junction of R12 and R10 to ground. That stage is the most sensitive and therefore needs the cleanest B+ it can get.
You could also replace R45 and R46 with two 5 ohm resistors in series and run the "ground" end of R33 and R34 to the junction between the two 5 ohm resistors of the other tube. This will force the tubes to equally share the load and thus be in balance.
roN
Hi,Did you notice that C112 is already at the junction of the R10 and R12 to ground?
My take is different - that C112's 100 uF is much too high a value of capacitance for this use-point. It will sound stodgy and slow. I would like to try just the minimal capacitance to do the job.
I'd like to see, at the very MOST, a 20 uF value film at that point, not 100 uF, In fact, I would start with only a 4 uF film cap there first, bypassed with another .01 uF film cap, Dynami Caps in both cases. I am listening to that exact value / type combination, as I type this post, and it sounds quite good to me. :-)
Jeff Medwin
I was looking at C112 as being at the juncture of R10 and R4 but actually you're correct. Wouldn't hurt to put one between R10 and R12 but..........
I agree with you about 100 being too much. 22-25 is plenty I'd think.
Thanks for pointing out my oversight.
roN
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: