DCAudioDIY.com

DC Area Audio DIYer's Community
It is currently September 23rd, 2020, 8:57 pm

All times are UTC - 5 hours [ DST ]




Post new topic Reply to topic  [ 54 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next
Author Message
PostPosted: May 11th, 2020, 4:36 pm 
Offline

Joined: February 28th, 2013, 1:19 pm
Posts: 484
Bruce,

I think Grover is talking about the resistor in the tail of the differential driver. Increasing the value of this will reduce the current through each section and decreases the voltage drop across the plate resistors and shifting the 6SN7 opt point further to the right, increasing swing.

David


Top
 Profile  
 
PostPosted: May 11th, 2020, 4:59 pm 
Offline

Joined: June 4th, 2013, 2:39 pm
Posts: 271
FerdinandII wrote:
Isn't that one of the two resistors setting the feedback ratio?

Grover Gardner wrote:
Okay. A dded a note to my previous comment about input sensitivity, something to consider as well.

Speaking of the 470 ohm cathode resistor, one thing you can do is raise that to 680 ohms. This is a common "blessed" tweak to the Williamson that increases the voltage swing of the driver stage and reduces distortion.


No, that's on the cathode of the input stage, also 470 ohm. ;-) Traditional Williamsons used 390 ohms for the driver stage cathode resistor. I assume it's been increased here to provide more swing for the 6550s, but it can be taken higher with good results.


Top
 Profile  
 
PostPosted: May 11th, 2020, 6:04 pm 
Offline

Joined: February 28th, 2013, 1:19 pm
Posts: 484
Grover, Bruce,

I am going to go and fix a few things on the amp, but I want to thank you both for your advice, it is appreciated. I think I may try to do as Grover suggests and start with the original circuit and then modify from there. I was debating this from the start, and was almost ready to punch out the top plate for octal sockets for 6SN7s, but figured that 9pin tubes offer more choices for tube options. Still nice to have 6CG7 and 12FQ7s out there (while still available nos or strong used) to play with before going to more hot rod tubes. Even thought about a "Williamski" with a 6N1P front end and 6N6P driver, but would have to learn a bit more about feedback modifications for that.

Got to get the bias supply fixed first!

David


Top
 Profile  
 
PostPosted: May 11th, 2020, 10:41 pm 
Offline

Joined: February 28th, 2013, 1:19 pm
Posts: 484
Grover,

I just calculated the gain of my driver stage with the ECC99, and due in large part to the higher Gm of the ECC99, I get a gain of 19.2 (25.6 dB), whereas the gain with the 6SN7 stage is 14 (23 dB). So there definitely an impact on gain with my arrangement. The function of a current sink in this circuit is equivalent to a big resistor (and negative supply) in a long-tailed pair, it does not impact gain (from what I read), just improves balance. So, it may be advantageous to retain this in the circuit if I can keep enough headroom (it will need to drop 25 volts or so, stealing it away from the tube). I could reduce the dropping resistor feeding the stage supply capacitor to get back some voltage. Because the current sink is adjustable, I can dial in the current in the stage to optimize the headroom anyway. Same function as a larger cathode resistor in the tail. So...perhaps I use a 12FQ7 and keep the current sink, this will keep the same gain characteristics (no change to the feedback).

David


Top
 Profile  
 
PostPosted: May 11th, 2020, 11:38 pm 
Offline

Joined: June 4th, 2013, 2:39 pm
Posts: 271
Sounds like that would work. If you haven't read Tim's analysis of the Williamson it's here, very useful:

https://dalmura.com.au/static/Williamso ... 20info.pdf

You can always adjust the feedback resistor. I would just be wary of applying over 20dB (instability) and input sensitivity. Kiebert "improved" the Williamson with 12AY7 and 5687 (see Dalmura) which I tried and found intolerable--too much sensitivity and electronic-sounding, generally just plain weird. That's one reason I worry about straying from the original.


Top
 Profile  
 
PostPosted: May 12th, 2020, 6:53 am 
Offline

Joined: February 28th, 2013, 1:19 pm
Posts: 484
Grover,

Actually, I was using that excellent analysis for calculating my ECC99 gain in the driver section.

However, I plugged in the characteristics for a 5687 as the driver, using 180V characteristics to get the Rp and Gm at the operating point I designed for (10mA per side), and it is alot closer match up to the 6SN7 gain, getting a gain of 15 vs 14. I have quite a few of those kicking around, more than 12FQ7s, all NOS. Never really had the opportunity to use really them. I think it would be an good substitution without disturbing too much what I have wired up to date. I have to rearrange a few things on the socket, but except for grid stoppers, I did not wire any components directly to the socket, so should be pretty easy to change a few pins.

Hmmm.

David


Top
 Profile  
 
PostPosted: May 12th, 2020, 7:06 am 
Offline

Joined: June 4th, 2013, 2:39 pm
Posts: 271
The 5687 works. To me it sounds "solid state"-y but it's certainly worth a try. I haven't tried a 6550 Williamson, but it's on my list because I have a pair of Acro TO-330's on hand.


Top
 Profile  
 
PostPosted: May 12th, 2020, 7:55 am 
Offline

Joined: February 28th, 2013, 1:19 pm
Posts: 484
Grover,

One thing I noticed about the Kiebert article is that he used the 5687 at the same current at the 6SN7 it replaced. Plate resistors are 50K. That is really a poor operating point for that tube, it really need more current for linearity. Which is why I currently have 20K plate resistors (for the original planned ECC99). What did you use, if you remember?

David


Top
 Profile  
 
PostPosted: May 12th, 2020, 12:07 pm 
Offline

Joined: March 2nd, 2013, 2:43 pm
Posts: 183
Location: Potomac, MD
David-
Not advised to use a half-wave rectifier on otherwise non-loaded transformer as it dc-biases the core. May be acceptable if current draw is very low because there is a substantial magnetizing current that will reset the core with each cycle, but only if the half-wave component is much lower.

In my experience phase lag and lead diddles in circuits like these never seem to improve the sound. They are done to try to maximize the amount of feedback that can be applied. Specs, Specs. Yes, higher feedback improves damping factor, but at what cost to the overall transparency of the amplifier. When I did the 807 amp that Jim G has, I was able to get the output impedance down to around five ohms--rather poor, but the sound was quite good. Probably only about five dB of feedback to meet all stability objectives. Stability should be checked using a square wave and various capacitor loads, with and without parallel load resistor.


Top
 Profile  
 
PostPosted: May 12th, 2020, 12:59 pm 
Offline

Joined: February 28th, 2013, 1:19 pm
Posts: 484
dberning wrote:
David-
Not advised to use a half-wave rectifier on otherwise non-loaded transformer as it dc-biases the core. May be acceptable if current draw is very low because there is a substantial magnetizing current that will reset the core with each cycle, but only if the half-wave component is much lower.


So a full wave or bridge is a better option. I have a 120V dual winding on the toroid I was planning on using, so could do a full wave pretty easily (I know toroid transformers prefer a bridge, but that is more diodes). The current will be pretty small, a few mA for the bias circuit.

David


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 54 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next

All times are UTC - 5 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 10 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group