Transformer coupled 6AS7/6080

[poynton]

Originally posted by Shoog


This Allows you to use easily availble mains toroidals as output transformers.

What VA rating were the transformers you used ??

Andy

[jarthel]

if I were to use a regular output xformer, what should be the specs?

also that xformer in the middle, is that what you call an interstage?

thanks again

[Shoog]

What VA rating were the transformers you used ??

For the outputs I used something about 120VA, but I have been assured that something around 30-50VA would do better.
For the interstagers I used what I could get which I think was 30VA or 50VA. In this position the smallest availble will be a destinct improvement.

if I were to use a regular output xformer, what should be the specs?

I think you would need something around 1K6. This might be very tough to source.

also that xformer in the middle, is that what you call an interstage?

Indeed it is, and it works very well.

any alternative to ecl82 in octal mounting?

Try looking at the range of TV deflection tubes. There are some that have dissimilar triodes in them that might work very well. The ECL86 is rated as a better tube, but is still 9pin.

The only thing you may have to tinker with is the two feedback resistors, all other values are accurate. I don't want to take mine apart to confirm the values.

For the power supply you will need silicon rectification as getting a valve rectifier to pass 300mA of current is not easy. You might get away with using a big 55V+55V toroidal for the power transformer. I used a Microwave oven transformer as a power supply choke (don't know how effective it is as an inductor, but it makes a nice big 50R resistor).

The whole motivation for this amp was to build a stereo PP amp with just 4 valves, it looks very well.
This will probably be one of the cheapest high end amps you could ever build. It runs in pure class "A" triode. Some of the most highly respected amps in the world use interstage transformers. It also has a differential output stage. If you can live with about 7watts then go for it.

Shoog

[jarthel]

why toroidal?

also for the output transformers if the primaries is rated 115V, what should be the secondary rating?

also, what's the power rating of the amp? would it be suitable for 8ohm speakers?

can you please show the schematics of the power supply? I have seen designs where tube rectifiers where paralleled (thorsten's website). I think this allows greater amount of current to pass through.

thanks again.

[Shoog]

Toroidals are cheap and they perform really well. The real motivation is because of where I live there is no readily availble source of "real" outputs without incurring huge postal costs.
The design requires 6V secondaries for 4Ohm speakers, and 9V with 8ohm speakers. Don't get to hung up on this, with such low output impedance the cgoice of output transformers is more forgiving than with most designs.
I can get 7watts per channel out of this baby, which is more than enough for me. This is about the theoretical maximum for this tube without going into AB class.

There is another 6080 amp on the web which uses toroidals but a more conventional valve phase splitter. It measures very well. It runs at a higher voltage - which may suit some people. Personally I think interstage transformers have been found to be the best phase splitters availble, though they do impose certain constraints on the design - namely the need for very low impedance high current drive.

Shoog

[Shoog]

The power supply is very specific to the power transformer I had availble (a huge 1000VA toroidal), and the other bits I could cobble together. Using the PSU2 power supply design utility should allow you to create something to suit what you can lay your hands on. To be honest my exact design wouldn't be tremendously useful to you.
I really think that in this design the extra effort of using valve rectification isn't warranted. With resistance in the power supply chain, the result should be nearly identical. Using parallelled rectifiers presents risks because one valve is likely to hog the load and burn itself out. The way round this is to place resistance in the anode of the rectifier, but then we start to lose even more voltage than the waistfull rectifier itself.

Shoog

[sonata149]

Hi Shooq,

Tks for sharing your 6as7-ecl82 schematic. I'm interested in building one and would like to know details of the CCS on the anode of the ecl82 pentode section.

Another question pls -- How is the interstage primary section connected in series or parallel (55+55V:110+110V)?

Thanks for your help.

Joe A

[Shoog]

Sorry for the delay in getting back to you.
I have done a radical overhaul of the amp recently and have a new schematic. I dropped the ECL82 in favour of a parallelled 5687 which does a better job.
I have attached the new schematic with the specified CCS.

The interstage transformer goes with the secondaries wired in "series" as the primary.

[sonata149]

Hi Shoog,

Thanks very much for your reply. That was sure quite a change. I was hoping to build mine with ECL86 front-end (because that's the closest I have in my box).

Are there any substitutes for the 5687 which you would recommend (ecc82/88 or octal) without componentry changes. (You see I'm not good at designing circuits, I mainly copy them.)

Also what type of transistors do you use in your CCS? Naming them would make it easier to select the correct ones.

Thanks for your help
Joe A

[Shoog]

Thanks very much for your reply. That was sure quite a change. I was hoping to build mine with ECL86 front-end (because that's the closest I have in my box).

Are there any substitutes for the 5687 which you would recommend (ecc82/88 or octal) without componentry changes. (You see I'm not good at designing circuits, I mainly copy them.)

You could use the original circuit with the ECL86, and I believe it would be an improvement on the ECL82. The reason I went over to the 5687 is because the feedback constituted to high a percentage of the plate current of the triode section. This was swamping its natural response. Reducing the feedback resulted in the triode overloading at low input voltages.
The 5687 is a simpler option which has plenty of headroom and can handle the required current. There are no reasonable substitutions apart from the 6H30p which will cost more than the 5687. Before I went to the 5687 I used the pentode (triode strapped) of the ECL82 at 40mA which worked well, but struggled to give the required gain. Married to a high gain preamp you could probably use the ECL86 in the same way - but this seems a terrible waste.

Also what type of transistors do you use in your CCS? Naming them would make it easier to select the correct ones.

I used pulls from an old computer monitor. I couldn't tell you the exact numbers and the only relevant spec is voltage compliance. 200V for the bottom one, and anything over 30V for the top one which should be high gain.

Hope that helps.

Will be away on holiday for a fortnight so wont be able to answer any more questions.

Shoog