6GK6 Ultralinear

jaymanaa · 12th December 2007, 12:31 AM

Hi everyone, it's been awhile since I posted here, but at last it's winter and time for some tube stuff. I've been working on a 6GK6 PP UL driven by a 6AN8, with a Blumlein Garter circuit for the cathode bias. I have one channel up and running (it's dual mono), but I think there is room for improvement. The amp seems way to easy to drive and has a bit of a ragged edge on the upper frequencies. The bass is good, but it seems overly bright. I thought some of you might take a look at the schematic and give your opinions. As allways, Thanks for the help, Jay

hailteflon · 12th December 2007, 08:58 AM

Jay, Interesting circuit. I am not qualified to give any opinions on this circuit, but I do have some questions.

(1) Don't you have three types of feedback: from the OT to the screen, unfiltered cathode, and also the tap from the cathode resistor divider to the grid of the other tube?

(2) Is the 6GK6 the tube that these component values were intended for?

(3) Possibly the 6BQ5 transformer is not right for this setup.

It seems to me that such a circuit would require careful choice of component values.

What specific circuit does the word "ultra-linear" come from?

I looks like it could be ultra-linear if everything is accounted for.

I'm basically curious to see what you say. No criticism intended. Mark

Tom Bavis · 12th December 2007, 12:54 PM

I don't see a cathode resistor for the first stage... what value? Split for feedback or using it all? The screen bypass should be returned to the cathode rather than B+. I would definitely bypass the output cathodes - you'll lose a lot of power and raise the output impedance without.

6BQ5 and 6GK6 should be identical except for pinout,

hailteflon · 12th December 2007, 01:59 PM

Tom has a good point. Pin 9 of the 6AN8 has no reference to B-.

Also, the 27K + 33pf RC filter in the feedback loop has a -3db frequency of 1.8 MHz, if I calculated correctly. Seems like you would want more roll off.

jaymanaa · 12th December 2007, 03:31 PM

I'm sorry guys, I do have a resistor from pin 9 to ground (value is 1K). I'll look at the other points you made today and see if I can figure out some answers. This is my first PP amp, and one of my first attempts at drawing a schematic. I really appreciate the responses, because it seems like this is the only way I can learn. Reading the books doesn't seem to get it through my thick skull. Basically this circuit follows the Heathkit AA151 amplifer (which the OPTs came out of) very closely except for all the gain stages and tone controls before the 6AN8. I also added the .068 cap and grid leak resistor as a filter to enable the use of larger value coupling caps from an old post of Eli Duttman that I dug up. The Blumlein Garter circuit for the output cathodes was found with the help of Bandersnitch. So far I have learned more from this amp than any before. Thanks for the replies, Jay

jaymanaa · 12th December 2007, 05:32 PM

BTW, this is the thread that inspired me. I found it while recapping a AA151 for a friend.

http://www.diyaudio.com/forums/showthread/t-70726.html

jaymanaa · 12th December 2007, 08:45 PM

Quote:

Originally posted by Tom Bavis
I don't see a cathode resistor for the first stage... what value? Split for feedback or using it all? The screen bypass should be returned to the cathode rather than B+. I would definitely bypass the output cathodes - you'll lose a lot of power and raise the output impedance without.

6BQ5 and 6GK6 should be identical except for pinout,

Ok, to clarify, I do have a cathode resistor on pin 9 of the 6AN8, the value is 1K. The screen bypass is actually connected to ground at this time (I really mucked up my schematic drawing), but I will try moving it to the cathode. Could you explain the reason for this (I'm just trying to understand things). I was just starting to experiment with bypassing the cathode resistors on the 6GK6s. Is there a rule of thumb on the value, or a way to calculate bypass caps in general (this is something I've been wanting to learn as well).

All in all, the amp sounds pretty bad. There is no hum or anything like that, it just sounds slightly distorted throughout the frequency range. I feel like the 6AN8 is overdriving the 6GK6s, but I'm not sure how to decrease it. I guess I could increase the R of the 150k resistor between the AF amp and the phase inverter (pin 6 - 2), or would that mean I would have to change the values of the cap and resistor to ground from pin two?

As far as the term "ultra linear" I was just refering to the OPT having screen taps. Thanks for your time guys, Jay

Miles Prower · 12th December 2007, 08:55 PM

Quote:

Originally posted by jaymanaa
The amp seems way to easy to drive and has a bit of a ragged edge on the upper frequencies. The bass is good, but it seems overly bright. I thought some of you might take a look at the schematic and give your opinions. As allways, Thanks for the help, Jay

These sonic defects seem to be related to inadequate gNFB. "overly bright" bass is an underdamping problem, most likely related to the lack of adequate cathode resistor bypassing in the finals. This will increase the effective r(p) of the finals, and pretty much undo what the local NFB is trying to accomplish.

Edginess at the upper mids and highs is due to letting pentode nastiness through. I've done two such designs (using 807s and 6BQ6GTBs). If you're gonna do a pentode amp, you're gonna need adequate gNFB to tame that edginess. Of course, add too much and that will lead to a "subdued", solid statey sound. That's not much better.

hailteflon · 12th December 2007, 08:57 PM

It seems, at first glance, that the situation with pin 2 of the 6AN8 is that the grid needs a resistor to ground.

Keeping in mind that I have much to learn, it looks like the 300V through the undefined plate resistor at pin 6 is going to put the grid at the wrong bias. It needs to be negative.

This would surely distort everything.

Tom Bavis · 12th December 2007, 09:18 PM

I think a pentode stage will have lower distortion with the screen bypassed to the cathode. Every other similar circuit (i.e. Dynaco) does this. It's always good to check the voltages in a direct-coupled circuit like this, as tubes vary, and you want close to 1/4 B+ on the phase splitter cathode, and 3/4 on the plate. This might be little higher than Heath's voltages, if your B+ is higher. If necessary adjust the value of the screen resistor. But it's less critical in this amp than in the Dynacos, since you need less drive to the output tubes.

For a cathode bypass cap, it should be much lower in reactance (at your lowest frequency of interest) than the cathode resistors. 39 Ohms (1/10 the cathode resistor) at 20 Hz is 200 uF - I'd consider that a minimum - 470 uF would be good. It's MUCH less important with a common cathode resistor - since the cathode currents are changing in opposite directions, the changes cancel out, until you drive it into class AB.

12th December 2007, 03:31 PM	#5
jaymanaa diyAudio Member Join Date: Jan 2006 Location: Kansas	I'm sorry guys, I do have a resistor from pin 9 to ground (value is 1K). I'll look at the other points you made today and see if I can figure out some answers. This is my first PP amp, and one of my first attempts at drawing a schematic. I really appreciate the responses, because it seems like this is the only way I can learn. Reading the books doesn't seem to get it through my thick skull. Basically this circuit follows the Heathkit AA151 amplifer (which the OPTs came out of) very closely except for all the gain stages and tone controls before the 6AN8. I also added the .068 cap and grid leak resistor as a filter to enable the use of larger value coupling caps from an old post of Eli Duttman that I dug up. The Blumlein Garter circuit for the output cathodes was found with the help of Bandersnitch. So far I have learned more from this amp than any before. Thanks for the replies, Jay __________________ "Make everything as simple as possible, but no simpler" Einstein

12th December 2007, 05:32 PM	#6
jaymanaa diyAudio Member Join Date: Jan 2006 Location: Kansas	BTW, this is the thread that inspired me. I found it while recapping a AA151 for a friend. http://www.diyaudio.com/forums/showthread/t-70726.html __________________ "Make everything as simple as possible, but no simpler" Einstein

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Rewiring a socket to take 6GK6 instead of 6BQ5	diamondsouled	Tubes / Valves	3	26th June 2008 03:16 PM
Ultralinear pse	mark_titano	Tubes / Valves	1	5th November 2007 10:54 PM
6V6 Ultralinear	gingertube	Tubes / Valves	12	4th May 2007 07:56 AM
6gk6 tubes	Tony	Tubes / Valves	2	8th July 2004 04:19 AM
PP-1c as Ultralinear	Charles BL	Tubes / Valves	1	28th January 2003 03:52 AM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read

12th December 2007, 08:58 AM	#2
hailteflon diyAudio Member Join Date: Sep 2006	Jay, Interesting circuit. I am not qualified to give any opinions on this circuit, but I do have some questions. (1) Don't you have three types of feedback: from the OT to the screen, unfiltered cathode, and also the tap from the cathode resistor divider to the grid of the other tube? (2) Is the 6GK6 the tube that these component values were intended for? (3) Possibly the 6BQ5 transformer is not right for this setup. It seems to me that such a circuit would require careful choice of component values. What specific circuit does the word "ultra-linear" come from? I looks like it could be ultra-linear if everything is accounted for. I'm basically curious to see what you say. No criticism intended. Mark

12th December 2007, 12:54 PM	#3
Tom Bavis diyAudio Member Join Date: Jun 2002 Location: Macedon NY	I don't see a cathode resistor for the first stage... what value? Split for feedback or using it all? The screen bypass should be returned to the cathode rather than B+. I would definitely bypass the output cathodes - you'll lose a lot of power and raise the output impedance without. 6BQ5 and 6GK6 should be identical except for pinout,

12th December 2007, 01:59 PM	#4
hailteflon diyAudio Member Join Date: Sep 2006	Tom has a good point. Pin 9 of the 6AN8 has no reference to B-. Also, the 27K + 33pf RC filter in the feedback loop has a -3db frequency of 1.8 MHz, if I calculated correctly. Seems like you would want more roll off.

12th December 2007, 08:57 PM	#9
hailteflon diyAudio Member Join Date: Sep 2006	It seems, at first glance, that the situation with pin 2 of the 6AN8 is that the grid needs a resistor to ground. Keeping in mind that I have much to learn, it looks like the 300V through the undefined plate resistor at pin 6 is going to put the grid at the wrong bias. It needs to be negative. This would surely distort everything.

12th December 2007, 09:18 PM	#10
Tom Bavis diyAudio Member Join Date: Jun 2002 Location: Macedon NY	I think a pentode stage will have lower distortion with the screen bypassed to the cathode. Every other similar circuit (i.e. Dynaco) does this. It's always good to check the voltages in a direct-coupled circuit like this, as tubes vary, and you want close to 1/4 B+ on the phase splitter cathode, and 3/4 on the plate. This might be little higher than Heath's voltages, if your B+ is higher. If necessary adjust the value of the screen resistor. But it's less critical in this amp than in the Dynacos, since you need less drive to the output tubes. For a cathode bypass cap, it should be much lower in reactance (at your lowest frequency of interest) than the cathode resistors. 39 Ohms (1/10 the cathode resistor) at 20 Hz is 200 uF - I'd consider that a minimum - 470 uF would be good. It's MUCH less important with a common cathode resistor - since the cathode currents are changing in opposite directions, the changes cancel out, until you drive it into class AB.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)