Aikido driver into an Interstage in Parafeed

[Greg.]

Congrats Andrew and also to the one who did all the pushing. There's alot to be said for PP you know

Best wishes,

Greg

[Andrew]

Thanks gents, I'll pass on your kindest to Mum and baby who are doing fine....

-- Andrew

[Andrew]

Ok guys, if I go down this route, what valves for my Aikido?

I have a separate 200v-ish supply for my driver stage, I don't have any particular valves in mind for the CF but I have some E180Fs lying around that could be used for pentode input. However, I was originally thinking of using a cascode. But I'm open to suggestions really, I could use the same B+ as the power valve but I've already got most of the PSU built and might as well go down this route for now.

Any suggestions welcomed.

cheers,

-- Andrew

[Andrew]

OK How does this look?

-- Andrew

[James D]

Hi Andrew,

5842s a wee beastie of a valve - lets see if I worked it right Rk=150R and Va=100V - so from the curves about Ia=7mA and Vg=-1.25V. Aikido gain about 20-23 ish with the 5842... So only 28V swing on the output before the input starts to draw grid current and starts to compress. Doesn't seem enough gain to do the job and using the 5842 here seems a waste as you not using its transconductance for anything useful. What was the aim with this one? How much gain do you need? WHat is the output valve? The parafeed bit is great

Assuming a 300B, I reckon a pentode or cascode front end to the aikido is what is needed here to give the gain required but whatever you use should be lowish transconductance to get enough input voltage swing and to take advantage of the nice CF output stage to the aikido.

The E180F is almost too strong for this but could be used, EF86 family would be better, like Dave Dove's Aikido. Or a higher B+ is needed to allow more -Vg and more Ia...

James

[Andrew]

Originally Posted by James D

Hi Andrew,


The E180F is almost too strong for this but could be used, EF86 family would be better, like Dave Dove's Aikido. Or a higher B+ is needed to allow more -Vg and more Ia...

James

Simple answer is that I've got a load of E180Fs, which when triode strapped are as good as 5842s, for about 1/3 the price.

Output could be a 300B or a PT15 (triode stapped, of course).

A higher B+ might be a problem with my PSU. How about I try to work in the E180F?

I don't have a Spice model for one when pentode wired though so I'm without my crutch, so to speak.

Alternatively, I suppose I could do something more mundane, with a EF86, but the cascode might be worth a try! Would that be an ECC88? Bit bored of those little guys as there all over my phono stages.

What I currently have with the 5842 looks a bit like a casode to me (with my niave eyes)? I guess I have that wrong? Must be JB's unique stage?

Sorry too many questions.....and no answers..

cheers,

-- Andrew

[James D]

Hi Andrew,

Ahh so the 5842 isn't its an E180F in triode mode - so same mu but half the transconductance (u=50, Gm=20mA/V, Rp=2.5k) and then the cathode resistors are too small - I think! I couln't find triode mode curves on the web just the typical parameters set in the Philips datasheet. A good start would be to try 330R as the cathode resistors and see where it settles...

To set the E180F as a pentode in the aikido we need half B+ still as Va so 100V. If we look at the plate curves and pick as suitable bias point and Ia - we're looking at Va=100V, Ia= 10mA and Vg=-1.5 volts. I would like to go higher on Vg but we don't have the swing to 3V to do it - too much curve bunching at higher Vg, So we Ia=10mA and 100V to burn in the load resistor gives Rl=10K. Now the gain of the stage is approx. Rl*Gm so 10k*15mA/V so 150 times... whoops too much for the B+...

What to do? Well the Philips data doesn't have transconductance curves but the 6688 data does and it shows the Gm varying nicely with Vsg. (this is typical for a pentode) so we can reduce Gm by running at low Vsg. IN this case we make Vsg 100V and Gm drops to 5mA/V. Now the gain is 10K*5mA/V =50 times. so for our +/-1.5V in we get +/-75 volts out and that fits sort of into the B+ of 200V. This might be a bit low for the 300B but it is all the B+ of 200V really supports. In fact looking at the 5687 curves it really only supports +/- 60V swing... (Note CF V/I charateristics are the same as for common cathode). however you may get away with the +/-75 V

But it is a nice driver for the PT15 as that has Vg=-40 in triode mode so well within the aikido range - but the thing is... the E180F in pentode mode alone makes a good driver for the PT15 - it doesn't need the Aikido CF to help it drive the PT15 as it's already running Zo=10k, Ia=10mA compared to the aikido Zo=100R, Ia=10mA. As we are driving 500K the lower output impedance doesn't gain very much and the current delivery doesn't change so slew conditions remain the same (it would drive into A2 very nicely with the Aikido though - but we're not looking at that with the PT15).

It would be a better driver for the 300B if we had B+=300V and then had +/- 100V available for output swing

Ohh to complete the pentode calcs. We need to set Vsg=100V. A nice 105V VR tube is one option or we can use the Isg and just resistive drop it. Isg=2.5mA (from datasheet or as Ia/4) Rsg= 100/2.5mA =39K then add a decoupling capacitor to the cathode to remove signal variations say Csg=10uF. We also need Rk=1.5/(10+2.5)mA = 120R.

The cascode calcs are similar to the pentode ones - give a shout if you want to run through those or any questions from the above. Oh for the cascode the ECC88 is very boring and horrible :-) so use the 6C45 as it's terrific at low Va see 6C45 at low Va.

ciao

James

[Andrew]

Wow, thanks James, I'm truly flabbergasted, give me a few days for all this to sink in and I'll get back to you.

cheers,

-- Andrew

[Andrew]

Hi James,

Hmm, so it seems that if I want to drive the PT15 (triode) I have four options.

1) E180F Pentode at around 100v/10mA into parafeed interstage.

2) 5842 into interstage using the interstage primary as an inductive load.

3) Trioded E180F into interstage using the interstage primary as an inductive load.

4) E180F - 5687 Aikido (or other)

Its funny I had my doubts about the 5842's ability to really drive a 300B but thought it would be OK with a PT15. Should have gone with my hunch.

For a 300B (where I wanted to go first funnily enough) an Aikido is still the way to go and I think you and Dave D proved this conclusively. Now, to do a 300B sucessfully......I need to...

1) Get more B+ fort my driver...

2) Use a different valve line up (even the 5687 CF stage appears dubious at 200v)

My mains TX has three secondaries, one for the power valve (500v-0v-500v), one for the bias (150v-0v-150v) and one for the driver stage (300v-0v-300v). To get more B+ I could ignore the driver winding and try and draw the extra from the power valve winding. Or, I wonder if I could wire driver winding and bias winding together to get 450V. The windings are defintely separate as the meter beeps prove.

Or I could just jump straight to the PT15s and try one of the first three or four options?

Too many decisions...I suppose I could save the Aikido for a headphone amp or as a last stage on a phono.

On a related note what size top cap should I order for the PT15?

thanks again,

-- Andrew

[James D]

First quick thoughts.

Assuming enough current in the windings, use a hybrid Graetz bridge for the rectifier and get more volts that way...

PT15 topcap is 9mm

Parafeed Aikido to drive the 300B using 300V B+ for the AIkido

Pentode E810F cap coupled to drive the PT15 in triode mode.

Build both and expereince and enjoy their relative strengths



James