Remember These?

[Cressy Snr]

Having a nice listening session this afternoon after analyzing the amp on the scope and sig gen all morning. It performs really well with the current feedback resistor of 4K7. Tried a few different resistors, higher and lower than the 4K7 and all were worse, in terms of sound. Raising the value (less feedback) buggered up the bass, whilst lowering the value made the sound overly straight-laced and slightly too bright. So...looks like I figured out the right value for use on my own speakers, first time out.

A 220pF, 200V polystyrene cap across each feedback resistor, squared off the corners of the square wave performance quite nicely.

Very slight improvements to the resolution and textures of cymbals, finger bells, triangles and other HF instrumentation; must be down to the caps across RFb. Other than that, nothing dramatic. The amp remains as uncoloured, cohesive and "human" as ever.

Shan't be using anything with an output transformer again. New Year project will be to improve the power supply for the input/ driver stage and carry out the possibly difficult job of getting it into a box without losing any of the magic.

[Ray P]

I missed the postie while I was out this morning but found a card on the doormat so I'm pretty sure there's a pair of 13E1s waiting for me at the local sorting office. I'll pop round and collect them tomorrow.

[Ray P]

I've been having a look this afternoon to see if I can have my cake and eat it with respect to the power supplies. I'll be using a separate chassis for my power transformers and I'm pretty sure I can squeeze in transformers for B+ supplies for both 6C33C and 13E1 valves - would just need to swap a couple of wires over in the terminal blocks to switch between them. For filament supplies I have a 120VA transformer with two 12V secondaries so by switching between parallel and series connection I can accomodate both valve types too (I'm assuming that with the loading on it the transformer will deliver a little more that 12V or 24V - we'll see).

[Cressy Snr]

Tried the older 13E1s (1950s) in the amp for a brief period. These don't have the four extra cooling fins on the anodes that the newer 1960s valves have. One was OK but the other developed two lovely red stripes on its anode, right where both the cathodes were, IOW, it got a little bit hot and bothered.

Bias was rock steady, but the valve was clearly unhappy. It's clear from this why the extra cooling fins were added sometime in the 1960s.

Nothing wrong with the older valves, they'll just need to be run at lower current when the time comes to use them. I've made a note on the boxes to that effect.

[IslandPink]

Nice work Steve, glad it's working out so well.
Must be a contender for this year's best 13E1 OTL amp competition. Oh, hang on, I've just heard there are no other entrants ... in the world

[Cressy Snr]

Nice work Steve, glad it's working out so well.
Must be a contender for this year's best 13E1 OTL amp competition. Oh, hang on, I've just heard there are no other entrants ... in the world

Aye,
It's interesting that no-one else has thought of using 13E1 in an OTL amp. OTOH it's such a daft idea, why would they?

[Cressy Snr]

I've been having a look this afternoon to see if I can have my cake and eat it with respect to the power supplies. I'll be using a separate chassis for my power transformers and I'm pretty sure I can squeeze in transformers for B+ supplies for both 6C33C and 13E1 valves - would just need to swap a couple of wires over in the terminal blocks to switch between them. For filament supplies I have a 120VA transformer with two 12V secondaries so by switching between parallel and series connection I can accomodate both valve types too (I'm assuming that with the loading on it the transformer will deliver a little more that 12V or 24V - we'll see).

You'll be able to hold your own tube shootout!

[Ray P]

Indeed, it would be an interesting exercise.

Of course, it's only the power transformer section that lends itself to a bit of flexibility. The amplifiers themselves have differences that would preclude a simple switch over so it would mean building two seperate amps - the necessary outlay won't make that happen quickly.

[Ray P]

Steve, if you still fancy trying the 13E1s with a little higher B+ remember my offer of the transformers I have spare (I thought I was going to use them for my 13E1 project but it turns out I have some other transformers I had forgotton about...).

Secondary 1 = 0-135V @ 1.1A
Secondary 2 = 0-7.5V @ 1A

Secondary 1 should have sufficient current to run a 13E1 at around 170/180V @ 500mA (so around 85W dissipation) and, with a voltage doubler, sufficient to run your input stage too. Sec 2 will do for your input stage filament supplies.

BTW, they're potted into black plastic cases so look reasonable on the outside of a chassis.

[Mike H]

The best amplifier n the world?

[Cressy Snr]

Steve, if you still fancy trying the 13E1s with a little higher B+ remember my offer of the transformers I have spare (I thought I was going to use them for my 13E1 project but it turns out I have some other transformers I had forgotton about...).

Secondary 1 = 0-135V @ 1.1A
Secondary 2 = 0-7.5V @ 1A

Secondary 1 should have sufficient current to run a 13E1 at around 170/180V @ 500mA (so around 85W dissipation) and, with a voltage doubler, sufficient to run your input stage too. Sec 2 will do for your input stage filament supplies.

BTW, they're potted into black plastic cases so look reasonable on the outside of a chassis.

Perhaps you could bring them up with you when you visit in the New Year?

[Ray P]

Perhaps you could bring them up with you when you visit in the New Year?

Of course.

[Cressy Snr]

Have got the amp about as quiet as I can get it, in its current form.
I did a few mods to the voltage doubler power supply to the driver stages.
So changed the caps in the centre tapped array that does the doubling to a pair of 200uF 500V JJ caps, giving a 100uF 1000V array. This is followed by a 100uF-10H-100uF CLC, followed by a 350mH-40uF LC filter.

Now the ear-to-speaker noise is "mmmmmm" rather than "zzzzzzzzz" and the sound quality has taken another step upwards.
More treble liquidity is the main improvement, plus better separation between individual voices on choral works. With an amp as highly resolving as this one, changes to the power supplies are very audible, which makes assessing the results quite an easy job.

When this goes in a box, I'm having a separate transformer for the driver stages. A small, 230V - 2x115V isolation transformer, with series connected secondaries will give the necessary 300 odd Volts to power the drivers. I've come to the conclusion that voltage doublers are noisy bastards and best avoided if you want the best sound.

[Mike H]

Not ideal, I would have thought.

[Cressy Snr]

Brought the 13E1 dissipation down to 72% of maximum by increasing the value of the CCS current setting resistors to 2.7R 3W (462mA current is now set). This has caused no ill effects on the sound and will undoubtedly prolong the lives of the tubes. They are now running comfortably at about 70W plate dissipation; 25W under the limit.

An added bonus is that the cathode voltage is now at 14.5V as opposed to the 12.3V it was, when the tubes were running at higher dissipation values. So now we have the capability of nearly 30V pk-pk of cathode movement instead of the 24V or so, we had before; so a bit more elbow room, plus, due to the reduced current draw, a litle bit more B+ across the valve
(a-k now 150V)

As Wolfgang advised, earlier in this thread, it has taken a bit of fine adjustment of feedback, tube dissipation and B+ values in order to optimize the amplifier for maximum output with the least distortion into my own particular speakers, and helpfully, optimizing the amp has not actually required giving the output tubes a thrashing. It needs a sensitive approach, rather than one of the hairy-arsed variety to get the absolute best out of what little is available to play with.