diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Tubes / Valves (http://www.diyaudio.com/forums/tubes-valves/)
-   -   Williamson KT77 Question (http://www.diyaudio.com/forums/tubes-valves/110439-williamson-kt77-question.html)

sorenj07 20th October 2007 11:39 PM

Williamson KT77 Question
 
Is this a reasonable set of voltages for the front end of my KT77 amp? Isn't the first 6SN7 a bit low? Could I switch to two RC's in series for the first two B+ inputs instead of a forked approach?

http://img86.imageshack.us/img86/301...nneltl6.th.jpg

ray_moth 21st October 2007 01:44 AM

It does seem a bit low, doesn't it? I would think this is better suited to a 6DJ8 than a 6SN7. The low voltage across both triodes is caused by the DC coupling and the very low B+ of 261v at the second (concertina splitter) stage. This voltage has to be shared between the plate load, tube and cathode load of the splitter (ideally in the ratio of 0.25:0.5:0.25). I see no need for the B+ to be so low, and you could raise it by reducing the 22k dropping resistor, but you might then have to increase the decoupling capacitor to 100uF or so. The same treatment could then be applied to the first stage, to keep its plate voltage at the required new level for the grid of the second stage.

As to using RC's in series for the first two B+ inputs instead of a forked approach, I imagine you could easily do that, but wait! You have no global feedback, so there is really no need to use DC coupling between the first and second stages. A better arrangement IMHO would be to use AC coupling, which would require you to use a cathode bias resistor and a grid resistor to set the operating point of the second stage but would also allow you to choose the operating point of the first stage so that its current is the same as that of the second stage. You could then dispense altogether with the first decoupling network and connect stages 1 and 2 to the same B+ because there is no need to decouple the first and second stages from one another if their currents are equal.

PB2 21st October 2007 03:30 AM

Improving the Williamson by Talbot M. Wright
 
Might this article from 1961 help?:
http://stiftsbogtrykkeriet.dk/~mcs/W...son/index.html

Pete B.

Cycline3 21st October 2007 02:55 PM

The Mullard Book says using a common capacitor bypassed cathode resistor on your output tubes for transformers less than 8k (which is most EL34 trannies -- 5k to 6.6k) gives the best sound. That's the way I'd go. While on the other hand, I think I would remove the common cathode R on the 6SN7s driving them and move those to each have their own unbypassed ... you don't need the extra gain and the feedback they would provide would help the amp sound overall. To further that though, I don't think I would use this layout at all -- do a search for the EICO HF87 or similar and you can see that you can dispense with one of your stages all together. The HF87 uses a 12AX7 on the input - you could easily use a 6SN7 in it's place - most modern preamps would still easily drive it -- and you aren't using any feedback to lower gain. Which again, I would do. My preamp doesn't have a global feedback loop, but my amp does. It just helps... you could always make it switchable...

Miles Prower 21st October 2007 05:22 PM

Re: Williamson KT77 Question
 
Quote:

Originally posted by sorenj07
Is this a reasonable set of voltages for the front end of my KT77 amp? Isn't the first 6SN7 a bit low? Could I switch to two RC's in series for the first two B+ inputs instead of a forked approach?

I'd leave it alone. Doing a rough loadline of that first stage, the Q-Point looks completely reasonable, with a THD estimate of about 1.7%. That's pretty good. Since this is a low level stage, you don't need much Vo swing either.

jane 21st October 2007 06:03 PM

The voltage over the cathode 22k resistor in the split load is ~80V. You should also have the same voltage drop over the other 22k resistor (it's the same current flowing through all of them) but you don't. The voltage drop over Ra is 261-229=32V and the drop over the resistor in the RC filter is 413-261=152V.
Are you sure your measurements are correct?

Jan E Veiset

sorenj07 21st October 2007 09:44 PM

I'm not positive, but what you say makes sense. My resistors are a bit underpowered - a few of those 22K's are only 1W.

I'd like to simplify the power supply and also use some beefier resitors. Should I leave it forked or not? I haven't calculated voltage drop etc. but I'd also rather leave the topology direct-coupled, because of phase errors with coupling caps...


All times are GMT. The time now is 06:01 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2