Question about Bridging

[eeyore]

I finally committed and bought a Decware SE84 kit amp. This is what I want to try the HV with, prior to building my own DRD45 (of which I have all parts and iron for).

Just a question that came up reading the site, what are the effects of bridging the amp (as suggested on the site) to achieve more power? I appreciate that it will have more power (a good thing), but what are the other negative or positive effects of this arrangement?

Thanks!

[SY]

Well, if you're going to bridge, why not just do it properly and build a push-pull amp? Bridging merely adds the supposed defects of push pull to the known inefficiency and transformer issues of single ended.

[Eli Duttman]

Originally posted by SY
Well, if you're going to bridge, why not just do it properly and build a push-pull amp? Bridging merely adds the supposed defects of push pull to the known inefficiency and transformer issues of single ended.

SY,

In the case of the DECWARE SE84, things may be a little less straight forward than you think. SE84 O/P "iron" has a single tap that floats. You "strap", rather than "bridge", a SE84 into a mono block. The I/Ps are connected in parallel and the O/P trafos wired in series. Phase splitting is not required.

I can tell you from 1st hand experience with a pair of SE84Bs that "strapping" works well. A single SE84 produces about 2 WPC into an 8 Ohm load. The pair of "strapped" amps produce roughly 8 WPC into 8 Ohms. Combined with my "homebrew" 94 dB. efficient speakers, the sound is quite nice.

[eeyore]

From reading the various items on the Decware website, it appears that there a number of different ways to 'strap' or brige the two SE84 to function as one?

One appears to be simply connect the + to - of the two OPTX. However, in your post Eli, you mention something looks to be more like the 'Taboo' topology?

Am I right in assumig that 90% of the products of Decware is the same design but slightly different implementations and operating points?

SY, in your post you suggest that by bridging two SE amps together, you get the worst of both worlds of SE and PP.

But would the 'Decware' method of bridging/strapping actually counteract that, and rather, combine the best of both worlds?

[kmaier]

The Decware has an unusual set of OPTs... 9800 ohm primary and 6 ohm secondary. The secondary winding is full-floating, they have polarity of course, but neither end is tied to ground or any other part of the circuit.

As transformers reflect impedance, this is how they can manage what in most amplifiers would not really work. I would suspect driving an 8-ohm load would be poor as the 9800 ohm load on the output tube would increase close to 13K. Using a 4 ohm load as the manual states drops the load closer to 6.5K. So if you put the two secondaries in series (which is typically not possible on most stereo amps as the negative lead is wired to ground) you can get a 6dB increase as Eli stated.

Overall a fairly novel approach but not really optimal loading on the output tube. As for bridging the two channels, most SET amplifiers have no global negative feedback, so their gain is strictly the result of the gain of each tube in the respective stages. In other words, unless you have extremely well matched output and input tubes, the gain of each channel will be different. As a result, if you parallel the outputs windings, you have the two channels working against each other as the gains are unmatched. Putting them is series eliminates this. Again, not perfect but less detrimental.

Overall a neat little design and pretty flexible.

Regards, KM

[SY]

eeyore, I can't answer that without prejudice. I am decidedly not a fan of single-ended tube amps.

[Eli Duttman]

I would add that I use a matched quad of SV83s and driver tubes whose 4 sections are closely matched, in the "strapped" amps. As KM indicated, steps must be taken to maintain symmetry and balance, in the absence of loop NFB.

-----------------------------------------------------------------------------------------------------

Whenever a stereoblock is tweaked into a monoblock, the object is to sum the voltage swings of the 2 sections in the load. In an amp with a common ground for both channels' O/P trafo secondaries, the only good method available is to drive the I/Ps with 180⁰ out of phase signals and connect the load to the 2 "hot" speaker terminals. This is "bridging".

Simply parallelling channels yields BAD sound and generates heat.

[kmaier]

Yes... tube matching is a bit critical for amplifiers, even push-pull ones which employ NFB. I have always done my own matching in actual operating conditions, not via a tube tester... even a mutual conductance one doesn't cut it. A nice Tek curve tracer would... but that's beyond a personal budget.

For my SET amps I match the outputs and input/drivers for gain, distortion and noise and usually can better 0.1dB matching, but I also use all NOS tubes, nothing of new manufacture.

As Eli pointed out, true bridging with typical tube amps is not that simple and not as effective as one would like. I would suspect the easiest method for true bridging would be a line input transformer with a split secondary to get the phase inversion.

I built a couple modified ST-70 amps for a friend/colleague many years ago. He decided to strap them to mono to drive a little pair of Maggies... we paralleled them as that was the easiest. That also implies we paralleled the outputs too, albeit the 16 ohm taps with an 8 ohm load. Power output was 90 watts into 8 ohms. The modified design used all 1% MF and included balancing so it was a bit better than the norm. Of course, YMMV depending on many variables.

Regards, KM