• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Super Triode Connection Amplifier Build with 6BM8 and 6AK5

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Well, I've been lurking about here for a while, also been doing a bit of research on the net about first tube/valve amplifier projects.

I’d actually been researching simple morse transmitters using a 6BM8 and stumbled across quite a bit of info on STC amps using this tube. It seemed a good contender for a first project as I had several 6BM8 on hand, inexpensive output transformers are able to be used and construction doesn’t look too daunting.

As they say, fortune favours the brave (or stupid), so here we go with some thought, but probably not enough, already put into the project.

The original schematic available here (http://www2u.biglobe.ne.jp/~hu_amp/amput3e.htm) shows a requirement of 6.3v at 2A and a HV transformer of 220v output rated at 30VA.

Not wanting to invest too much in a project which may or may not work, I'm going to back-to-back a couple of cheap transformers for a simple power supply.

In the picture, the two respective transformers are 30 and 15VA. I should be able to take 6.3V at 1A out from the first and then step back up for 240V at about 60mA of HV.

I know this is only half of the power requirement so I guess the simple answer is to build two monoblocks!

Being newish to all this, I'd most appreciate any advise along the way. I'll also be posting questions as I certainly do not have all the answers yet.

This may well be a drawn out affair, so I hope you'll bear with me as this project takes its course.
 

Attachments

  • power transformers.jpg
    power transformers.jpg
    89.6 KB · Views: 2,191
These are the output transformers I’ll use. They are 5k Ohm input with a tap for 2.5k and the output is multitap with 2, 4, 8 and 16 Ohms out. The original article above shows a transformer with 3.5k in/8 Ohm out but I’ve also seen a few schematics with two OPT’s per side.

This chap (http://www.atatan.com/~s-ito/amp/yakinori.html) is using two output transformers with specs 7k Ohm in/8 Ohm out with the inputs paralleled and the outputs in series.

As these were really cheap, I bought four so I guess I’ll use two each per monoblock. Not sure how I’ll wire these yet. Any suggestions? Has anyone else built one of these?

DerFly.
 

Attachments

  • output transformers.jpg
    output transformers.jpg
    81.1 KB · Views: 2,063
These "magic diodes" keep showing up in various schematics. I tried a "stopper diode" in the plate lead of one of my sweep tube SE amps. The only thing that it stopped was good sound.

I noticed a definite change in the output signal when the amp was driven to clip with a strong bass note. What was previously a single cycle event with the top whacked off of a transient, now affected about 3 cycles and there was some funny ringing effects. Yes there is some definite speaker interaction here. It may sound good to some people with some music, but I didn't like it.

I have an EL-34 SE amp playing now. I will put in a diode and see what happens.

Disclaimer:

My experiments have been in triode or UL mode. The authors experiments were in pentode mode.
 
I think we should distinguish the diode in the screen lead from the one in the plate lead.

Just from a rapid guess, the one in the plate lead looks very very suspect to me, the experience that George mentioned makes me think about interactions between the power supply and the power stage.

The diode in the screen path, however, puzzles me. Is the screen grid supposed to have a one-way current allowed, like the plate? So what's the effect of the diode?

I remember we already touched this subject:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=76960&highlight=
 
"300B pop-ups?"

Check Epay, search on "Kepco", its the listing that has girlie pictures for power supplies.

After reading the STC article, I think the diodes are some kind of wishful thinking to block back EMF from the speaker.

As far as the inverted triode feedback scheme goes, this is brilliant thinking, but they need to put more gain in the loop to make the output conform more accurately to the constant current triode NFB characteristics, and to improve the speaker damping factor. They are suffering from global NFB phobia to their own detriment. (Although I would leave the xfmr out of the loop)

I have just been tinkering with an active ultralinear circuit and have a new version that uses an inverted triode NFB diffl. amp stage to control the output pentode screen. Useable for SE or P-P designs. The days of using actual triodes for output tubes are in jeapardy. Inverted triode feedback using constant current only generates significant 2nd harmonic, whereas actual triode outputs driving a real load generate higher harmonics too.

Don
 
In order to use a small triode for the inverse triode NFB in the STC circuit, one needs to attenuate the output tube's plate signal before applying that to the FB triode's plate, or one will run into headroom problems in the triode as mentioned in the article. One might be able to use an UL tap on the xfmr for this maybe, but then one risks instability. A partial CFB/ pentode output configuration would likely be a better way to derive an attenuated feedback signal from the CFB winding, but the phasing will have to be attended to then.

The use of direct coupling in the STC every where seems to be causing big problems for voltage operating points on the tubes, better to use a coupling capacitor somewhere and get optimum operating levels for all the tubes.

Finally, it needs a diffl. ampl. stage in the loop feeding the output tube grid so enough gain is available in the feedback loop for accurate triode emulation and low output Z.

Here is a successful implementation of inverse triode feedback using a SS amp in the op. amp. configuration. (I think the op. amp. circuit model shown in the STC literature incorrectly shows the NFB triode's grid connected to the input signal, it should be to AC ground as done here.) This amplifier produces only significant 2nd harmonic distortion. It also can put out 300 Watts, no SETs around to compete with this baby!

http://www.diyaudio.com/forums/showthread.php?postid=875795#post875795

Another interesting design would be to use TWO inverted triode feedback loops (not the global loop, but inner loops), one for each output pentode tube in a P-P amplifier. Both of these would use the same output signal phase to feed the inverse NFB triodes. This way one output pentode would be made to emulate a constant current triode and the other to emulate an "anti- triode" (due to the phase inversion on the second output) so their even harmonic distortions would not cancel as usual for a P-P amp, but rather would add. Another way to get SET sound with power. (edit: Hmmm, actually just need one inverted triode feedback and a phase splitter to sum NFB with each output pentode's grid drive signals)
(Of course, one could just use one inverted triode globel NFB loop to get similar results from P-P, depends on semantics desired. Triode outputs emulation, or whole SET amplifier emulation. )

Don
 
SY,

The diodes? Good question. It may as well be smoke, mirrors and black magic as far as I'm concerned. I certainly do not presume to understand the STC topology.

In all truth, I'm building this more as an experiment, an exercise in amplifier construction, if you will. I'll keep on with it and hope you might grace me with answers to some of my stupid questions from time to time!

Anyway, I guess I can fit the components on a 200mm x 150mm chassis. (8x6 in old money!)
 

Attachments

  • stc layout.jpg
    stc layout.jpg
    96.7 KB · Views: 1,322
Yes, I have reservations about the diodes, but I think that the circuit has a chance. You can always try it with the diodes, and short them out with a clip lead to try it both ways.

I would try a bigger chassis if you can find one. Keep the output transformers a little farther from the power transformer. The magnetic field from the power transformers can be picked up by the output transformers causing a hum that you can't fix.

Also try to keep the input circuitry as far away from the power transformers as possible. It likes to pick up hum also.
 
SY said:
Even after reading the explanation, I'm clueless about the "stopping diode." Could you elaborate on that? What happens when it reverse biases? How does it function when it's forward biased?

I've seen that before. The idea is that the diode prevents reverse screen current. What really happens is that it makes tubes poof.

As for the diode in the plate Vpp, it's not necessary, nor desireable. For a SE design, the PS is in the signal path, and you'd do much better keeping its AC impedance as low as possible, either with really decent filter capacitors (PiO, polypropylene) or an active regulator.

As for 6AK5s, they're little and cute, and five or six of them make one helluva distributed wideband amp, but I have my doubts as to how good they sound when run so near cutoff.

I wouldn't take a chance on this one: someone's trying too hard to be "fancy" and "clever" here.
 
Tubelab,

Yes, I agree that the chassis I'm working with is probably too small with the opt's in too close a proximity to the power transformer. I'm sort of stuck with this size as I have a piece of aluminium sheet 300mm x 200mm to work with.

I suppose not cutting this in half might have given me a little bit more space to play with. Unfortunately too late now, I'm stuck with 6 x 8" for each chassis.

I was going to build both monoblocs at the same time but might just build the one for the time being and the other on a larger chassis at a later date if this one has a hum problem.

I've laid this one out with all transformers turned 90 dergrees to each other. I'll also use best practice as detailed in another thread when wiring up, i.e. twisted heater connections, etc.

As you can see from the attached picture, I've drop through mounted one transformer so that all primary wiring is below the chassis and secondary HV is innaccessible to roving fingers. This leaves me with maximum 15vac above the chassis which I'll try to insulate somehow.
 

Attachments

  • chassis top.jpg
    chassis top.jpg
    67.4 KB · Views: 1,284
And on we go!

I've finished wiring the PSU as shown in the attached picture. The heater wiring is twisted and the lights come on with tubes plugged in and powered up.

The HV is showing just under 270v which is a bit less than the specified 280v. This drops to about 250v with the 10k 10W resistor which I've soldered in temporarily as a bleeder.

The original schematic doesn't show a bleeder resistor so, what do you all think? Should I use one or not?

Also, as the PSU is two multitap transformers back to back, I could wire the 15v out to the 12v in to step the HV up to the required 280v or a bit more. Is this acceptable?

Thanks all for your suggestions and comments.
 

Attachments

  • stc psu.jpg
    stc psu.jpg
    98.1 KB · Views: 1,162
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.