DCAudioDIY.com

DC Area Audio DIYer's Community
It is currently March 28th, 2024, 7:21 am

All times are UTC - 5 hours [ DST ]




Post new topic Reply to topic  [ 61 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6, 7  Next
Author Message
PostPosted: December 10th, 2020, 12:31 pm 
Offline
User avatar

Joined: July 24th, 2015, 4:17 pm
Posts: 1701
Location: Parkville, Maryland
Roscoe Primrose wrote:
David McGown wrote:
Roscoe,

OK, here goes the Cliff Notes version... It's really a very simple amp, standard pentode operation of the driver, gain around 10 if I remember correctly. The output tube is just like a standard pentode output stage, except the control grid sees no signal. If you think about it, with no input signal at all, it operates exactly like any other self biased SE pentode operating stage. Plate at about 400VDC, screen grid at 250VDC, control grid at 0VDC, cathode at about 16VDC. The difference comes in where the signal is injected into the output stage... Take a look at a typical pentode characteristics curve:

The screen grid really does isolate the plate from the rest of the circuit. In a triode, if you increase plate voltage and keep everything else the same, plate current goes up. In a P/T, changing only the plate voltage does essentially nothing (unless you go to ridiculous extremes) but if you change screen grid voltage, you do get a significant change in current. That's what I'm doing here, modulating the screen voltage as a way to control output current. Doing things this way eliminates the huge power consumption/heat generation common on most DC tube designs, as the cathode can be at the same reference above ground it would be with typical control grid modulation. The 5.1K resistor probably isn't necessary, but I wasn't comfortable just connecting the grid to ground (why do I type gride every time and have to go back and remove the 'e'?) on the first try...

This design does make some significant demands on the driver though, keep in mind that you have to swing a lot of voltage AND supply current to the screen. You're not going to pull this off with a 12AX7 for a driver;)

I should have the parts to complete the stereo pair Monday, then I can give the design a fair test.

Roscoe

Interesting that you mentioned a 12AX7 -- because as a "driver" I don't get why some designers like Jadis would use it for that purpose. So --there is the 12BZ7 which is basically a 12AX7 on steroids.

BTW -- the metal-case 6V6s sound better than most of the glass 6V6s. Possibly because of the shielding the metal-case provides and the possible stray-emissions control.

_________________
Walt


Top
 Profile  
 
PostPosted: December 10th, 2020, 1:03 pm 
Offline
Site Admin
User avatar

Joined: February 28th, 2013, 10:38 am
Posts: 1682
ratbagp wrote:
Maybe I am wrong but using the formula
Ck = 1 / (2 * pi * f * Rk) 

I get
160 uf for 200 ohms
127 uf for 250 ohms

Is there some other formula I should be using?

ray


That's a gross over-simplification. You need to be worried about stage gain, which varies as the external cathode impedance changes, but it's not the only factor. Remember, gain can be calculated as effective plate impedance/effective cathode impedance. For -3dB, gain needs to be .707 times the gain with 0ohms AC to ground from the cathode. BUT, effective cathode impedance is a combination of the external network impedance (the R&C in the cathode circuit) AND internal cathode impedance (1/gm (in mhos, not umhos)). So, there's no one right value for a given frequency & cathode bias resistor value....

Roscoe

_________________
I can explain it to you, but I can’t understand it for you.


Top
 Profile  
 
PostPosted: December 10th, 2020, 5:56 pm 
Offline

Joined: March 2nd, 2013, 2:43 pm
Posts: 216
Location: Potomac, MD
I got turned on to screen drive when I was building amps in college—a looong time ago. I was looking for getting good performance with lower idle current as I was into Earth Day and power waste etc. I noticed the 6L6GC transfer characteristics shown in the RCA tube manual. Both my RC20 and RC28 tube manuals show these for G1 held at 0 volts and varying G2 volts. I would assume that any of the other RC editions between my book editions would have these curves.

I was taken by the linearity compared to typical G1 transfer characteristics, and indeed, after building a 6L6 amp with G2 drive I was quite pleased. But then I tried some other things that really turned out to be exciting. I changed from 6L6 to 6DQ6 TV sweep tubes, and upped the plate voltage to 800 V. The screed drive was done with direct-coupled high-voltage bipolar transistors, emitter-follower configuration. This amp really rocked; 90 watts out and idle dissipation less than five watts. Unlike Roscoe’s SE, a P-P amp. Transformer loading 6600 k P-P. My record for power with G2 drive is 240 W with a pair of 6LF6s.

I have done a couple of screen-drive amps based on audio tubes, but the TV sweep tubes work better because the screen sensitivity is much higher. For audio tubes, the screen drive should really be at least 600 V peak-to-peak, where as 300 V p-p is good for sweep tubes.
You do indeed need direct coupling for screen drive because the screen needs dc current.

As to cathode-bypass caps, best to ground the cathode and bias the grid accordingly. But if you use a cap, make it big. Roscoe is correct in that you must consider the cathode impedance when determining the LF roll-off. The bigger the cap, The less the parasitic impedance issues; ESR etc. ESL issues come in at high frequencies, but generally not an issue for audio. If you want to hear the mess an undersized bypass cap makes, do the following: Hook up a high-gain audio amp and speaker, and cap couple it to the bypassed cathode in your amp under test (AUT). Run a music signal through the AUT with a load resistor connected to the speaker output. Listen to the gooey mess that is on the cathode of the tube being bypassed in the AUT. Try a much larger cap for the bypass, and listen as the gooey mess is suppressed. The gooey mess WILL impinge on the audio signal and color the sound.


Top
 Profile  
 
PostPosted: December 10th, 2020, 7:58 pm 
Offline
User avatar

Joined: July 24th, 2015, 4:17 pm
Posts: 1701
Location: Parkville, Maryland
dberning wrote:
I got turned on to screen drive when I was building amps in college—a looong time ago. I was looking for getting good performance with lower idle current as I was into Earth Day and power waste etc. I noticed the 6L6GC transfer characteristics shown in the RCA tube manual. Both my RC20 and RC28 tube manuals show these for G1 held at 0 volts and varying G2 volts. I would assume that any of the other RC editions between my book editions would have these curves.

I was taken by the linearity compared to typical G1 transfer characteristics, and indeed, after building a 6L6 amp with G2 drive I was quite pleased. But then I tried some other things that really turned out to be exciting. I changed from 6L6 to 6DQ6 TV sweep tubes, and upped the plate voltage to 800 V. The screed drive was done with direct-coupled high-voltage bipolar transistors, emitter-follower configuration. This amp really rocked; 90 watts out and idle dissipation less than five watts. Unlike Roscoe’s SE, a P-P amp. Transformer loading 6600 k P-P. My record for power with G2 drive is 240 W with a pair of 6LF6s.

I have done a couple of screen-drive amps based on audio tubes, but the TV sweep tubes work better because the screen sensitivity is much higher. For audio tubes, the screen drive should really be at least 600 V peak-to-peak, where as 300 V p-p is good for sweep tubes.
You do indeed need direct coupling for screen drive because the screen needs dc current.

As to cathode-bypass caps, best to ground the cathode and bias the grid accordingly. But if you use a cap, make it big. Roscoe is correct in that you must consider the cathode impedance when determining the LF roll-off. The bigger the cap, The less the parasitic impedance issues; ESR etc. ESL issues come in at high frequencies, but generally not an issue for audio. If you want to hear the mess an undersized bypass cap makes, do the following: Hook up a high-gain audio amp and speaker, and cap couple it to the bypassed cathode in your amp under test (AUT). Run a music signal through the AUT with a load resistor connected to the speaker output. Listen to the gooey mess that is on the cathode of the tube being bypassed in the AUT. Try a much larger cap for the bypass, and listen as the gooey mess is suppressed. The gooey mess WILL impinge on the audio signal and color the sound.

There it is -- grounded cathode. I've tried to advise people about the advantages of a grounded cathode but was spearned as a moron. Yet the increase in dynamics is unmistakable. I did grounded cathodes on a re-done ST-70 decades ago with unexpected amazing results. My Carys have cathode-bias on the 845s and the bias resistors are about 1,125-ohms with 100-ufd of polypropylene cathode bypass caps that has the response down to 2-Hz.. Dennis Had nailed it. There is no gooey mess -- quite the opposite. Musicality and speed. It doesn't get any better!

I've had the same experience with other amps. -- and it is not my imagination. The common thread running through the different amps. with smaller value film caps has proven to provide musicality and speed. Large electrolytics are counter intuitive to the goal of good sound reproduction. One only has to sit in my living room to hear the benefits.

_________________
Walt


Top
 Profile  
 
PostPosted: December 11th, 2020, 12:29 pm 
Offline
User avatar

Joined: December 14th, 2013, 2:19 pm
Posts: 948
Roscoe Primrose wrote:
So, while the rest of you were arguing about guns & who has the best geriatric friendly exercise program, I built an amp. Yes, AN amp, can't find the parts to build a matching channel using the same parts... Mouser here I come ;)

Image

Dead quiet, and gain is pretty low, but definitely worth building a second one, will give another update when I get the other one built...

Roscoe


Curious why you chose to drive the screen from the plate of the voltage amp, rather than use a follower (SS or tube) direct coupled to the grid with lots of standing current and a very low output impedance? Was it simplicity, elimination of a third stage, or other reason?

Slick

Attachment:
Source-follower.PNG
Source-follower.PNG [ 498.87 KiB | Viewed 6788 times ]


Top
 Profile  
 
PostPosted: December 11th, 2020, 2:29 pm 
Offline
Site Admin
User avatar

Joined: February 28th, 2013, 10:38 am
Posts: 1682
Stuart Polansky wrote:
Curious why you chose to drive the screen from the plate of the voltage amp, rather than use a follower (SS or tube) direct coupled to the grid with lots of standing current and a very low output impedance? Was it simplicity, elimination of a third stage, or other reason?

Slick


Simplicity. Also, the choice of driver takes that into account, with about 45mA of standing current....

Roscoe

_________________
I can explain it to you, but I can’t understand it for you.


Top
 Profile  
 
PostPosted: December 12th, 2020, 4:53 pm 
Offline

Joined: January 14th, 2015, 11:15 pm
Posts: 499
It looks pretty easy to build.

Image

Yellow lines are heaters and connection from the 6V6 plate to 6550 screen.
Green lines are cathode / ground associated.
Red lines are everything else.

I usually don't use the turret strip method but I will in this case just for fun. According to my spreadsheet, I have some strips in small box 2 inside large box 16 inside our Movecube.

If I have missed out on anything, let me know.

ray


Top
 Profile  
 
PostPosted: December 12th, 2020, 5:00 pm 
Offline

Joined: February 28th, 2013, 1:19 pm
Posts: 914
May need to increase the spacing between the tubes, looks pretty close for a 6550.

David


Top
 Profile  
 
PostPosted: December 12th, 2020, 5:04 pm 
Offline

Joined: January 14th, 2015, 11:15 pm
Posts: 499
Found my own error - input from RCA to grid of 6V6.

Image

ray


Top
 Profile  
 
PostPosted: December 12th, 2020, 5:12 pm 
Offline

Joined: January 14th, 2015, 11:15 pm
Posts: 499
I use PCBExpress just to get the general placement outline. I would expect to have a few inches between the tubes.

I will also be trying out a new construction method using an Ikea Trolley since I will need to move the amp around a lot. It will use a three layer trolley from IKEA called a Rashult which is a smaller version of a Raskog.

https://www.ikea.com/au/en/p/rashult-tr ... -50445990/

The power supply will go in the bottom layer. The heavy output transformers in the middle layer and the rest of the amp at the top. The trays can be installed upside down.

ray


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 61 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6, 7  Next

All times are UTC - 5 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 42 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