[FishHead]

Congratulations on the new baby!

Any comparisons with the T2?

 

[tomb]

As we stated earlier, the Torpedo II does not offer any appreciable benefits over the Torpedo I and it will not be pursued.

That actually means the Torpedo I is pretty d*mn good already. We were only looking for something that would offer an obvious increase in quality over the Torp I. The reason is obvious, because the Torpedo I is not an insignificant investment.

That also means the Torpedo III must really be stupendous. Well, it is! The bad news is that also means it will cost more. However, Beezar Audio and ECP Audio have always tried to maintain a logical correlation between price and performance.

We think the Torpedo III will justify all of this - it's that good.

 

[dsavitsk]

The short answer is that the prototype T2 was OK, but it did not really do anything better than the T1. And it would have cost a little more to build. The T3 on the other hand is really a nice amplifier -- maybe we're not supposed to say that, but we would not make it available if we didn't think that was the case. We put a lot of work into it, and we're very pleased with the sound we're hearing.

 

[diego]

So ... is it an AC coupled series push pull Mosfet mu follower?

 

[tomb]

So ... is it an AC coupled series push pull Mosfet mu follower?

As usual, it's best to let dsavitsk answer the technical questions.

That said, it is a tube hybrid DSHA. However, the differential output devices are BJT, NPN transistors, not MOSFETs. This is the second T3 prototype built (after two T2 prototypes). The first T3 prototype used the MOSFETs. However, we found that the highs were rolled off because the MOSFETs don't present a high enough input impedence to the tubes.

 

[dsavitsk]

So ... is it an AC coupled series push pull Mosfet mu follower?

I looked into the mu follower topology, but for a variety of reasons chose against it. The T3 is a Long Tailed Pair with a CCS tail with differential emitter followers. It can also use MOSFET followers, but I like the BJTs better.

 

[OJNeg]

Thanks to Tom and Doug for letting this guy loose. I'm curious as to how it stacks up.
 
Also have to say I'm very curious about possible Cinemag OPTs. I still think that even more performance can be squeezed out of the Torpedo I with better parts!

 

[diego]

Thank you for your answers. I am looking forward to see the final result as I have been wanting to build something like that for the last few months.

 

[FishHead]

LEDs under the tubes? I thought I saw one on the board in the pictures.

 

[dsavitsk]

Also have to say I'm very curious about possible Cinemag OPTs. I still think that even more performance can be squeezed out of the Torpedo I with better parts!

We are working on making this a possibility. We'll know more soon.

LEDs under the tubes? I thought I saw one on the board in the pictures.

Not quite. There is an LED between the tubes.

 

[tomb]

 

Also have to say I'm very curious about possible Cinemag OPTs. I still think that even more performance can be squeezed out of the Torpedo I with better parts!

We are working on making this a possibility. We'll know more soon.
 

LEDs under the tubes? I thought I saw one on the board in the pictures.

Not quite. There is an LED between the tubes.

 
Just an update for those interested and it elaborates a bit on Dsavitsk's post above ...
 
We decided to wait until the Cinemags come through to make absolutely certain that no other changes are needed on the PCB.   The Cinemags are spec'd to match the pinouts on the Edcors, but you never really know, especially when dealing with two different mfrs.  Of course, the PCB is just fine with the Edcors now.
 
Also - just an FYI on the LED.  I asked Dsavitsk for this as additional assurance in indicating whether the amp is powered on or not.  The 12AU7 tubes are pretty much completely surrounded by the case lid, even more so than the 6J6's in the Torpedo 1 because the tubes are a larger diameter.  If you have getter flashes on top, then the tubes are almost completely obscured and it becomes fairly difficult to tell if the power is on.  I've accidentally left the Torpedo 2 prototypes running for a couple of days for the same reason.
 
Anyway, Dsavitsk has eked out every bit of current using the existing Edcor PT.  To do that, he severely limited the current to the LED.  You may find that with cheaper, high-output LEDs, the current is not even sufficient to light the LED.  I put in a cheap pink LED just for the heck of it in building the prototype.  It wouldn't light, so I thought it was in backwards until I tried replacing it and then got the full story.  So, I'll test quite a few LEDs when we release the design and make sure there's a decent selection that will provide reasonable lighting under the conditions described.

 

[skeptic]

Sounds like an awesome amp kit!  Looking forward to reading more about it as you guys approach the release.  

 

[UKToecutter]

Ooh,
 
Not sure I like the sound of that.......
No headroom on the PT????
At all??

 

[tomb]

Ooh,

Not sure I like the sound of that.......
No headroom on the PT????
At all??

A few things:
1. Dsavitsk is very conservative.
2. Current is very limited within the tube circuit anyway. The high-voltage conversion through the OPTs is the real source of current at the output.
3. Referring to #1 again, Dsavitsk is simply ensuring that nothing whatsoever could introduce noise in the circuit, like a PT with too much current load.

"Eke out every bit if current" was poor phrasing on my part. Again, if I've said something technically incorrect, Dsavitsk will correct me.

I don't think you'll be disappointed. It drives Grados, Senns, and AKGs with equal authority.

 

[dsavitsk]

We are running it conservatively to ensure that noise and heat are minimized. In a sense, we are running it more conservatively than in the T1 as the B+ current is similar but the heater current is reduced. If you run a bit more current, nothing will be damaged (which I think is what you mean by headroom) but I don't think much will be gained either.

Think of it this way, the output portion of the circuit runs at ~12mA per side but that is into a 10K load. As that is stepped down into a 32 Ohm load by a winding ratio of ~18:1, the current available is multiplied by 18 (minus transformer losses). That is to say, power is conserved across a transformer. It is the equivalent of running a bit over 200mA per side. Into 300R phones, it works out to just under 100mA (with the requisite increase in voltage swing). And because the circuit itself has a low output impedance, it does not suck up much of that (which is the big advantage over the T1 and why this amp is a bit more powerful.)

Power transformers are rated not based upon what will make them burn up, but on how much current one needs to draw to make them meet their specs. So a 50W 50V transformer will deliver 1A @ 50V. If you draw a little more current, the voltage will drop, how much depends upon how it is wound. If you draw a lot more it might burn up, but we're not pushing that line.

The LED is at the end of a B+ bleeding resistor and it gives some indication that there is still voltage present. A run of the mill red or green led will light up with no problem. If you want it brighter, you can reduce the resistance, but again, there is really no benefit to doing so. I'll note to that the fact that the LED is not lighting up should not be taken as definitive evidence that there is no high voltage present. You should always measure before touching anything in a high voltage circuit, or really in any circuit.