OPT on the speaker side?

[dovla]

Considering building OB speakers to go with the tube amp and locating OPT in the speaker base.
If running the amp with 'modest' B+ of 250-300 V (i.e. EL84 SE or PP) plate(s) and B+ could then be connected between the OPT and the amp 2-3 meters away using the regular household power cord.
Any reasons why this would not work (possible interference, etc)?

[aardvarkash10]

why not just put the entire amp in the base of the speaker and run line level to it?

[eeyore]

Would be interested to know if there'd be any microphonic or feedback issue with the amp or the OPT in the same chassis, both with airborne vibration (sound) and structural vibrations (chassis).

[Mr. Triatic]

This solution have crossed my mind a few times but i have never actually tried it out, it's certainly a good way to keep the tubes away from the speaker (microphonicswize) and minimize the length/loss-issue of wire on the current side of the transformer.

Although it's not a good solution if your amplifier has a feedback loop from the secondary.

[marcuswilson]

The impedance on the output anodes of a tube amp is generally between 2000 and 10,000 ohms. The capacitance in even a short length of cable can be a significant load on the output tubes. This can cause resonance with the leakage inductance of the output transformer, and these resonances may be within the past band on the amplifier causing peaks or dips in the frequency response.

Furthermore, if you are using overall negative feedback, then you'll need to bring the secondary of the transformer winding back to the amplifier chassis.

During my building and repair of valve amplifiers over the years, I've noted that amplifier stability is very sensitive to the routing off the wires connecting the anodes of the output valves to the output transformer. In my opinion you have great difficulty producing a wide frequency response and highly stable valve amplifier whose output was fed from the anodes of the output tubes via long wires to the transformer.

The possibility of generating interference and feedback into other wiring within your system is quite real if you have cabling containing audio at 600 V peak to peak around the place.

The last problem would be if one of your animals decided to chew through the cable. I don't know if you have young children who may have a pet rabbit that I would think the rabbit would probably be electrocuted if they decided to chew through the cable.

[dovla]

Thank you all for replies. I suppose the best way is to try it out and then put the OPTs back in the amp if it does not work out. There'd be no global feedback in the amp so it may turn out workable with 2.5 m between the tube and the OPT.
Nobody/nothing around to chew on the cables so that worry is out of the way.
The idea was to make the amp smaller and use the space behind/below the speaker to hide the OPTs. Putting the entire amp on the speaker side would require another PS, plus it would take the tubes away from sight and also put them too close to the source of vibration so that
Nobody/nothing around to chew on the cables so that worry is out of the way.

[DF96]

I would take careful note of post #5. Putting the OPT by the speaker was sometimes done in valve radios, but the cable was only a couple of inches and the bandwidth stopped at about 5kHz anyway!

Low impedance connections (e.g. conventional speaker cables) can have losses due to resistance. High impedance connections, as you propose, have problems due to stray capacitance. That is why someone invented the 100V line - a useful compromise.

[tomchr]

I would be very concerned with the capacitance of the cable as well. When I started out in tube amp design (not that long ago) I started with a crappy OPT that had high parasitic cap on the primary. The frequency response was horrible and the amp had a real nasty tendency to oscillate at 60~70 kHz. This oscillation is well outside the audio band, so not audible by you. But it'll happily fry your tweeters.

If you go against the advice you've been given and place the OPT away from the amp, at least make sure you get some cable that's rated for at least 2x the B+ voltage. Yes, the voltage does swing that high. Also make sure it has a shield that you can ground for safety. You may not have critters who will chew on your cables, but I see no point in being reckless. Again, this way of operating is not recommended. I strongly suggest that if you decide to do this you try with a dummy load first. Monitor the output with an o'scope for a variety of input signals before committing to a final solution.

~Tom

[mashaffer]

I would be afraid of stability issues as well. Interestingly enough I am kicking around the idea of powered OBs as well. The idea in my case is to use a higher Z out amp for the helper woofer to help compensate for baffle losses. In essence I would be tailoring the amp to the speaker.

The microphonics are most important in the small signal tubes so it might be possible to use an approach similar to that used in some consoles where the output tubes were on one chassis and the preamp and driver tubes on another. One could use CF or mu follower to drive the lines between the preamp/PI and the output tubes and mount only the outputs in the speaker.

One advantage I think we have in the OB is that we are not trapping a high energy back wave in the cabinet so we might have less energy to deal with over all. I could be wrong on that of course. One approach I have considered is to put the entire PA in the speaker but suspend the chassis behind the full range driver and away from the helper woofer. One could damp the "cabinet" and attach the amp chassis and the speaker drivers to the "cabinet" using lossy material such as rubber mating or that green slime they make for that purpose (can't remember the name).

Just a few ramblings.

[Poindexter]

You gonna be running big voltage across the room.

Poinz