A first SE design

[simon]

In order to understand this valve malarkey a little better I thought I'd do some calcs on a nice SE amp. I chose 76-45-2A3 for no reason other than it sounded interesting, and I've been listening to 2A3s (and 45s) for a little while. I had no idea whether it would work, but this wasn't the point really, I just wanted to play with some numbers and see what happened.

So the attached circuit is what I ended up with. Can any of our more knowledgeable members have a shufty and see if it looks something like? Cos just because I've done some sums doesn't mean that it's right, or that it will work, or that it will sound any good.

I tried to keep a common HT, set by the 2A3, which was interesting because the DCR of the opt is much smaller than the load resistors for the 76 and 45 so I've ended up down at the bottom of the 76 and 45 curves. For all I know 76s and 45s might not sound very nice down there.

I need some decoupling in there too, but don't know how to work it out. I guess I'd start with 47uF between the 76 and 45 and go from there.

All comments welcome, but be gentle, I've got Dave's L plates firmly attached.

[Nick]

Ok, I think it will work, and you seem to have valid op points. But as you say, I think it might not sound great. The problem is that combining the choice of valve, RC coupling and the B+ you have is causing you the problem. If you had 500v for the first stage, then a 50k load and 5ma on the 76 would be a great place. Likewise, 7k and 35ma on the 45 would put it in a better place.

The first stage will give you a gain of about 8, the second maybe 3, so the overall gain is 24, so to drive the 2a3 to the diode line will take about 30v RMS, so thats a input of 1.3v RMS, so thats ok, but maybe a bit higher than ideal, depends on the preamp used

But in generally, it should work, so give it a go.

The other option to avoid the low op points or high B+ is to choke load the first stage (cheap hammond choke), and interstage couple the second stage, maybe the hammond chokes mentioned elsewhere, if they are good with that much current, otherwise I have a pair of AE 1:1 interstages that you could borrow. I think then it would become a rather interesting circuit.

[simon]

Thanks Nick. I did wonder if choke loaded 76s and ITs might be a nice place to end up.

So, one step at a time, I'll rework the circuit for the better OPs first. The PSU might be interesting though...

I was hoping not to use a preamp, just run the sources into a pot. Should I be aiming for 1V RMS and can I squeeze the extra gain out of these valves?

[Nick]

Well, I think for a work in progress the gain will be ok, it was just somethig to bear in mind. Choke or TX loading will give you a bit more gain, so it will be about 1v for full swing with that. 500v will be easy enough to do for the driver stage.

[Paul Barker]

Have you got any old output transformers kicking about? Use them to choke load the 45 you're part way there.

[andrew Ivimey]

Hello Simon, interesting usage of the concept 'beginner' I am not too familiar with!

Anyway, ditto to all that's said so far, particularly using an(y) old output transformer on the 45 anode, if it ain't got swing it don't do a thing!

The 76 on the input - Are we spoiled by load line choice instead of having only one will be right!? For my Amity front end I used 76 in SE mode, 1k cathode, the amp has 310v on HT so I added a little 10H salvaged Akai choke into 47K anode resistor with a 33mfd electrolytic to earth. O.33 coupling cap to input transformer input of Amity front end. It sounds fine. I know I could up and down anode and cathode loads, but would it sound better??? I don't know.

I have cut out and mounted suitable sockets for several experiments similar to your proposed circuit (including 26!) and if I weren't all Amitied-out that's what I'd be doing now.

Hurry up and build this amp Simon and let us know - you will save me the work

[simon]

The revised circuit is attached. The OPs are now:

76: Va=250V, Ia=5mA, Vgk=13.5V, rL=50k
45: Va=252V, Ia=35mA, Vgk=48.5V, rL=7k
2A3: Va=250V, Ia=60mA, Vgk=43.5V

For the PSU, I need 40mA @ 500V for the 76 and 45, and 60mA @ 310V for the 2A3. So I could drop 190V through a resistor for the 2A3 but is there a more elegant way of doing it?

I have a pair of Hammond 1627SEs I could use for the 45s. Do I use the primaries? My LCR meter is kapput - any idea what sort of L they will have?

But I'm out of my depth now (see Andrew, I really am just a beginner). What does the anode choke do to the operation of the valve? Okay, it has high impedance to AC so does this mean current doesn't change much as the voltage swings? How does this relate to the value of HT I use?

The DCR of the 1627 primaries is 233R, which is pretty small compared to the load resistor so I guess I can ignore it, but ideally should I subtract the DCR from the load resistor?

Y'know, I might just have to build this

[Nick]

I think the simple way of getting the voltages is seperate power supplies, the 500v might be simple using a 240v isolation TX and a voltage doubler circuit if you can't find better matched TX's. You could always use the old two TX's back to back, and get the heater supply from for the 76 from the middle. Or you could try what I think Darren does and share the supplies. Use a 56 instaed of a 76, then use the same supply for the 45 as the 56.

You could play some games with a diode and drive both a choke input supply and a cap input supply from the same TX, that should give you about 500v and 320v, so it would be fun to try, assuming you jad a TX that would supply enough current.

Yes, you can consider the choke as a vary large value resistor to AC, so its like having a large load, without needing high B+. Its not entirly a flat load line, as its in parallel with the following grid resistor, but you can consider it flat for the moment.

Not wanting to suggest you don't build something, but I am tempted to suggest instead of building a 500v power supply, it may be better to jump directly to the choke loads. Two hammond 150H chokes will cost less than the extra hardware for the 500v supply.

But then you don't get to hear the difference.

And beyond that, the 45 was about the most linear driver I found for a 300b, so it might be fun to do that as well.

[simon]

When I placed my last order with AE I "ordered ahead" and have a few mains txs and chokes winging their way to me as I type. There isn't a mains tx big enough for choke input for the 76s and 45s, but I can go cap input, and choke input the 2A3s.

Then I can go to 310V HT on the 45 with the Hammond 1627s?

And I have some 5k opts coming from AE, so I could choke load the 76s as well perhaps?

Mmm, yeah, I had thought about 300Bs

[andrew Ivimey]

you have pm, re chokes,

from
an anonymous also a beginner

[ed]

Hi Simon

I am also looking for a 300/500 split PS at the moment. I had doodled and come up with the naive psud shown below but I'm sure I should be looking at improvements...I would be keen to see what you finally come up with.....ps my Psud final voltage(300) is not final because I havn't worked out the draw for a 12ax7/12au7 aikido yet so I guessed it at 20ma.

[Cressy Snr]

Hi Simon

I think separate power supplies are a good idea. Given the differing voltage requirements of the three sections of your amp, IME it is easier to deal with each one separately and set it up for the requirements of each valve. Doing it this way, you don't get as much of the problem of decoupling each section from the others.

Of course you then have three power supplies to design, but as you are using such good valves, wouldn't you want to pamper them as much as possible? Letting each one sit in its own stable bubble would be a good way to start.

Steve

[Nick]

And I have some 5k opts coming from AE, so I could choke load the 76s as well perhaps?

They may not have enough inductance for the 76, more like 200H is needed, the little Hammonds give 150H

[simon]

Hi Andrew, you also have pm.

[simon]

Hi Simon

I am also looking for a 300/500 split PS at the moment. I had doodled and come up with the naive psud shown below but I'm sure I should be looking at improvements...I would be keen to see what you finally come up with.....ps my Psud final voltage(300) is not final because I havn't worked out the draw for a 12ax7/12au7 aikido yet so I guessed it at 20ma.

Hi Ed,

I guess it'll depend on which iron and caps I have as to what I'll use, rather than finetune in PSUD. I'll start with yours and see where I get to.

I struggled to really understand the current draw in the Aikido too. From memory it was a little higher in reality than I thought it would be.