my new tube line stage

[7N7]

Originally posted by hesener
thanks everybody for the comments.

...

Chose the 6922 primarily for its linearity (and because CJ chose it, I admit) and with the 4:1 stepdown at the output I can get around 10V unclipped which is a good value, I think, enough headroom. ...

Whoops! Sorry I was tired yesterday and missed the output transformer.

I think that E88CC is not a very linear valve - it's quite good around 85 - 100V. Of course the gm makes it attractive I understand.

You can deservedly rest on your laurels with this fine piece of work, but for the future you might like to look at triode-strapped E282F: mu = 25, gm about 30, ra <1k. It is very linear.

All the best

7N7

[SY]

I think that E88CC is not a very linear valve

Really? My measurements would indicate otherwise, at least for the sort of swing needed for a linestage.

[7N7]

Originally posted by SY


Really? My measurements would indicate otherwise, at least for the sort of swing needed for a linestage.

Yes, your observation about the necessary swing for a line stage is absolutely correct; even so, there are many more-linear types with that sort of mu, albeit with lower gm. 13D3 and 2C51 are two that spring to mind.

7N7

[hesener]

7N7, thank you. Will have a look at the E282, but for the time being I am working on a couple power amplifiers with EL84 in PPP topology (have the PCBs and transformers, so I just have to build it......)
Any opinions on the 6H30?

hesener

[rdf]

It took a few minutes to understand what's happening in that circuit. Fixed bias with LED regulated supply, correct? If you're in the mood for adventure, I'ld bet that front end would work well in grid-leak bias too with much less complexity.

Nice work!

[Shoog]

I'ld bet that front end would work well in grid-leak bias too with much less complexity.

Way to radical for most I fear !!!

Shoog

[7N7]

Originally posted by hesener
7N7, thank you. Will have a look at the E282, but for the time being I am working on a couple power amplifiers with EL84 in PPP topology (have the PCBs and transformers, so I just have to build it......)
Any opinions on the 6H30?

hesener

Hello, you are most welcome.

Sadly I cannot do any building these days, but I remember well how many things I wanted to make and how there was never any time!

If you email me I can let you have a spreadsheet showing some distortion tests I did on E282F - the results were very good - also I have the same for 13D3 (6158)

I still have a pair of 6H30. I used them in my 13E1 amplifier as cathode followers. Steve Bench's curves for them look excellent; before I recently got the pair of Quad IIs I have, I was planning an amplifier that would have used 6H30 as drivers, I think it's a good valve.

Best wishes

7N7

[hesener]

The LEDs serve as zener diodes, to get a stabilized voltage to adjust the (negative) grid bias. LEDs are operated forward, different from zeners that spend their lives in constant avalanche breakdown and create a lot of noise doing it.... This LED section exists only once, for the four tubes.

rdf, thank you for your comment however I could not fully understand what you mean with "grid leak bias"... care to explain?
thanks

[ErikdeBest]

Hi Hesener

I reacently saw the grid leak biasing being applied in another thread, and today started to look at it, how it works exactly. Here is a link...

http://www.tpub.com/content/neets/14...s/14178_45.htm
http://www.tpub.com/content/neets/14...s/14178_46.htm

Look at post 10 of this thread
dc phono It shows a ECC88 with grid leak biasing!

Viel Spass, Erik

[rdf]

Hi hesener. It's a good use of LEDs I've considered as well. I usually though use them directly as cathode loads to eliminate excess capacitors and permit a direct coupled input. However I also like having the cathode tied hard to ground as you've done. Grid-leak preserves the latter. Here's the first Google reference I found: http://www.tpub.com/content/neets/14...s/14178_45.htm .

Essentially the resistor from grid to ground is sized much larger than data sheets recommend - more than 2 meg isn't unusual - and the grid coupling cap retained. Since the grid resistor is so much larger the cap be much smaller, usually equating to higher quality as well. What happens is the impedance from grid to ground is made too large for the flow of random electrons striking the grid to drain off. They collect and a negative charge builds on the grid until it stabilizes at a point where electrons start draining again. Since you no longer have exact control over the grid-cathode bias voltage the plate voltage must be tuned by altering the plate current and possibly the supply voltage. This might make well outside the scope of your project.