New amp: 13DE7 SE hybrid

[Tony]

what i mean is that the 100k and the 300k should be interchanged, the 50volts taken from the junction of 300k and 100k....

any reason why the isolation output traffo is returned to B+ instead of ground?

lastly, i see in you drawing that there is no grid leak resistor to ground for the second triode, did you build your amp like that? just curious.....

[Fenris]

Yeah, the 100K and 300K got transposed on the schematic, they're correct in real life. Also, in post 4 I noted that I accidentally left the grid leak resistor out of the schematic. If you look at the picture of the underside of the actual amp, there's a small resistor on the right side of the tube sockets that runs to the ground bus bar - that's the grid leak resistor, it's a 300K IIRC.

I connected the parafeed to the B+ because that way the amp is a constant current draw amp. It makes life easier for the power supply, and should reduce noise. It will work hooked up either way, but I think that it's a bit more efficient since there is no current in addition to the CCS.

[Tony]

Originally Posted by Fenris

Yeah, the 100K and 300K got transposed on the schematic, they're correct in real life. Also, in post 4 I noted that I accidentally left the grid leak resistor out of the schematic. If you look at the picture of the underside of the actual amp, there's a small resistor on the right side of the tube sockets that runs to the ground bus bar - that's the grid leak resistor, it's a 300K IIRC.

I connected the parafeed to the B+ because that way the amp is a constant current draw amp. It makes life easier for the power supply, and should reduce noise. It will work hooked up either way, but I think that it's a bit more efficient since there is no current in addition to the CCS.

thank you for your response, now i understand your design better.....

[Fenris]

Shoog - I did a little critical listening tonight. Speakers were bookshelves with Peerless woofers, Usher tweets as designed by Wayne J. at the former speakerbuilder.net:

The PeeCreek

Source was mp3s from a Galaxy Nexus. I listened to a variety of material and the one thing that stood out was the mediocre quality of the cell phone. The amp is good enough that it isn't the weak point in the chain right now.

I tried with my computer speakers as well, Tang Band 3" Bamboo drivers in a tapered quarter wave pipe. They're very lively, with a definite tone to them (probably due to their solid wood construction and limited frequency range. That tone came right through, telling me that the amp is fairly neutral and not a weak point. Eventually, this amp will be used with the larger bookshelves as an amp for my computer.

[payloadde]

Originally Posted by Fenris

Next up, how to design the output stage. The reflected impedance with an 8 ohm load would be about 480 ohms, too low for most any tube, at least the ones I could afford.

Just an afterthought:
Wouldn't a tube cathode follower be feasable instead of the MosFet ?
Assuming gm of - say -15 mA/V should give Z=1/gm=66 ohms for the CF.
120mA is also well w/in the range of many sweep tubes.

I've done just that recently, albeit with a 240V/24V Xfmr.

[Fenris]

A tube cathode follower would work just as well, but because there is no voltage amplification, I doubt there would be a sonic difference. I think Tubelab would agree that a mosfet follower doesnt color the sound. I just scaled up his Powerdrive to the next level. There would be several diadvantages however. First, tubes cant swing as far to the rails, so that may limit power. Second, a big tube requires a big heater, which is just wasted power if not needed. Third, the Mosfets were a buck each, less than the cost of a good socket. This amp was inexpensive and I didnt want to spend alot on parts I didnt have on hand.

[Junm]

Hi, i've tried your design concept, i dont have dissimilar triode at hand so i tried with my available resources too....see schematics...but i have some problems....at low volume...it's sound okay but when i go higher there's some annoying sound just like pops,scratches etc.but no hum....and one thing i've noticed is the 3 green bias LEDs on 6N6p are blinking/dimming but the 5.8v read out stays...what might have cause this thing...I've increased the current for about 7ma but no effect on the sound and the dimming.

The led on 6n1p is stable...at low volume LEDS has stable light. would it be clipping or whatever..please help me out...what wrong on the schematics

[Tony]

you can try increasing the current on the 3 seriesed green leds...a resistor from the B+ of say 47k 2 watts ought to increase the led current to about 10mA...

maximum voltage swing happens at 1/2 Vcc of B+ to the blocking cap, 20ufd..

[Junm]

Originally Posted by Tony

you can try increasing the current on the 3 seriesed green leds...a resistor from the B+ of say 47k 2 watts ought to increase the led current to about 10mA...

maximum voltage swing happens at 1/2 Vcc of B+ to the blocking cap, 20ufd..

Hi sir tony thanks for dropping by...i've tried increasing the bias current to 7ma by adjusting the source resistance of the CCS...still no effect but increased the plate voltage of 6n6p at point B to 167v...by lowering the ccs current i can achieved 1/2 B+ at point B...and also to point C (output)

I will try the feed resistor you propose and see what the effect...

One thing with 6n6p at 5.8 v input and gain of around 20 or 22 (=116v) will it be enough for a full swing on the trafo?

[Tony]

the load seen by your output stage is about 1400k, how low does your B+ drop when you increase the volume?

assuming that your B+ stays 255volts, the ac out is about 85 volts so that output power is about 5 watts.....less if your B+ sag.....lesser still as you deviate from 1/2 B+.....

having 2 stages of tubes, gain is plenty.....

gain of the 6N6 is limited by the mu of the tube...the 5.8v is the bias needed to get the target plate current....it is not an input voltage...

6n1p has a mu of 33 and the 6H6/6N6 has a mu of 22, so theoretically you have a voltage gain of 726....actual voltage gain is much lower....