Subwoofer Single ended tube OTL amp?

stephie's place

Moderator: sbench

Post Reply
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Subwoofer Single ended tube OTL amp?

Post by sbench »

Here's an interesting little concept to amuse you on those snowy eves...

A DC restorer is simply a RC coupled circuit that has a diode hanging on it. With no AC, the output DC level is simply the return point of the resistor. As you insert AC signal, the diode pervents one side of the signal from changing (much) but the other side is free to grow. Thus the average DC point changes. Only if the signal changes quickly do you get any "distortion", but the average value does change with signal.

Now, for "inband" signals, this is audible and not necessarily very good. But for LF signals, this is not too much of a problem. What this gives us is the ability to bias a tube so that it is not overdisipating when idle, continue to maintasin class A operation, and develop some very high peak powers, which tubes don't mind all that much.

That allows us to build a single ended, class A transformerless (it uses a 220mH inductor on the output) tube amp that develops useful power to power a sub-woofer. 4 6AS7's for example produce over 20 watts in this manner, with each section idling at a mere 5.5 watts! (x8 sections = 44 watts). But this is amplifier is class A, and no output transformer.

The 220mH inductor will need to handle about 2.5 amps of DC, which isn't too bad. (it has 320mA idle current - 8 6AS7 sections going through it).

This circuit requires only one high current supply (output stage and driver) and 2 low current power supplies (bias, and input stage).

Something to contemplate on those cold evenings (I understand you New Yorkers have somewhere around 18 inches on the ground, and more coming?).

Enjoy,

Stephie
Attachments
6AS7_FOL.JPG
6AS7_FOL.JPG (88.83 KiB) Viewed 18846 times
tube_power_follower.asc
(8.24 KiB) Downloaded 595 times
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Hey Stephie,

Way cool... I'll digest... you could always tap that inductor and throw some voltage away for current :-)

dave
Get Your Fix
www.hifiheroin.com
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

Here's a slightly more refined version. The input stage is closer to what we would like, I think. Adds a little more sensitivity. Removes the 250 volt supply. Same power available, but I've detailed a little bit better what happens as you add tubes. The 5.1 ohm output impedance (with 4 6AS7's) ain't too shabby for a SE transformerless device. Harmonic distortion is primarily second, which might actually be advantageous in a subwoofer amp.

With 6 6AS7's the predicted power is 36ish watts, and with 8 6AS7's it is slightly over 50 watts. 50 watts is reasonably impressive for an SE tube amp, with or without an output transformer.

Tapping the inductor doesn't do a bleep of a lot for you... notice that the output impedance is actually pretty low.

Just an interesting concept.

The power supply might be a bit interesting too!

Enjoy,

Stephie (who is still getting used to typing with these acrylic nails) *giggle*
Attachments
6AS7_FL1.JPG
6AS7_FL1.JPG (83.18 KiB) Viewed 18801 times
Retsel
Posts: 47
Joined: Tue Feb 01, 2005 9:51 pm
Location: Ann Arbor, Michigan USA

OTL for Lowthers

Post by Retsel »

Hi Stephie, I admire your engineering ability.

I have 15 ohm Lowthers which are a good match for OTL amps. Many of the OTL designs have compromises that I would not like to live with.

I recall that you worked on a SE OTL design about a decade ago. Did you ever finish off that design? I imagine that this circuit could work as well, but the previous one used DHTs in the driver position. Anything that you could share about that design would be appreciated.

Retsel
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

Lowthers also have reasonably high efficiency, so the power output required isn't too high, making it nice for OTL purposes.

One could argue that having an output inductor required disqualifies things from OTL status.

The above amp would NOT be good for wideband; you'll hear the "breathing". I'm surprised no one picked up on that circuit, though, it produces a LOT of power relative to the devices used, and still qualifies as single ended.

The best sounding SEOTL was the inverted 5687 version. Since 5687's aren't THAT expensive, piling a lot of them in parallel and even matching them is tedious but practical.

Actually, the BEST sounding SEOTL was the 813 version I cobbled together. I didn't publish it, as it's pretty impractical; it required matching 10 813's per channel, and that's a kilowatt just to light the filaments. I'd have to go back to my notes, but driving it was a problem... you build what amounts to a 845 amp, running on 1000 volts, and capacitively couple into the paralleled 813's running inverted. So you end with absurdly high voltage in the supplies (845 resistive loaded from a 2.4kV), absurdly high voltages and currents in the 813's and a huge air-conditioner bill. I never built a stereo version, but even mono required a couple independent house circuits. It did sound pretty dang good, though from what I remember of it.

-Stephie
Retsel
Posts: 47
Joined: Tue Feb 01, 2005 9:51 pm
Location: Ann Arbor, Michigan USA

OTL for Lowthers

Post by Retsel »

Thanks for the very rapid reply!

I definitely do not want a whole house heater!

I have your writeup on the 6CG7 into 5687 OTL including a schematic. I need to study this a bit, but the electrical demand seems to be quite modest for an OTL amp. You even modeled the tube performance inverted! - very linear indeed. Parallelling 10 tubes though - Yikes -isn't there a loss of focus (I guess that is why matching tubes is important)? I recall that for one of these OTL designs, you used a 6SL7 (as a voltage regulator?) in the bias supply which helped to reduce distortion and increased focus. Would that work here?

While this design is cap coupled, I don't mind that (particularly since there is only one!) because I can use this cap to high pass the feed to the Lowthers which is also a plus because it keeps the low frequency bass demand out of the output stage. I may want to increase the total number of tubes above 10 to have some power in reserve. Although I may start with fewer tubes just to make sure that I can make it work.

Retsel
Retsel
Posts: 47
Joined: Tue Feb 01, 2005 9:51 pm
Location: Ann Arbor, Michigan USA

5687 OTL amp

Post by Retsel »

Oops, I remembered the schematic wrong. There are two coupling capacitors in the signal path and an input capacitor. All are necessary. This is one downside of this design, however, Stephie's comments on its sound quality are impressive.

I analyzed the power draw for the circuit and I estimate the power draw at idle to be 150 watts per side, ignoring the losses in the power transformers. This is really low for a single ended design, but of course, without feedback.

This circuit is very interesting to me. The inverted tube design is wild. I need to read more about this. I have a general understanding of tube operation, but I need to get my head around this design.

Retsel
Post Reply