• 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.

Balanced interconnect to differential amp - how would you do it?

Status
Not open for further replies.
Have what amounts to a Raven style preamp, http://www.nutshellhifi.com/Raven-MarkII.gif, so the output is fully balanced, with the shield grounded to chassis.

This output feeds two devices: a power amp for the speakers, and a Behringer DCX2496 for bass. Behringer is the typical op-amp based differential input stage, with four 10k resistors, I believe.

Amplifier is (will be) a differential design, similar to the Karna, http://www.nutshellhifi.com/Karna.gif.

My question is, do I want to still use the input transformer in the power amp, or save the money and drive the interconnect directly into the grids? Does the answer have any bearing on the fact the Behringer is hanging on the same interconnect (I wouldn't think so) ?

Thanks.
 
You would think that removing a component as compromised as the input transformer should produce sonic benefits.
However there seem to be a few benefits in having the input transformer;
Transformers generally introduce a hard to define sonic signiture which is generally considered favourably.
There is ground loop isolation which may eliminate potential hum source.
The transformer will present a much lower DC load to the input grids than the output of the preamp (unless the preamps output is transformer coupled), which generally seems to respond better as the grid starts to draw a bit of current in overload. Not an issue if there is plenty of input headroom.

So on balance I would like to see a transformer either at the output of the preamp or the input of the power amp - though two is probably overkill.

Shoog
 
Thanks Shoog:

If you would briefly check out the links I attached, the preamp is already transformer coupled, so yes, I would be using two in series. Don't want overkill (in addition to the added cost), but if there is any advantage to the input tranny, I would like to know what that is.

I didn't mention this initially, as I wanted to see what the response was, but I also would like to consider a passive 6dB high pass filter on the power amp, as this will drive my MTM's. How to implement this filter will enter into this decision, I suppose.

And no, using the Behringer (at least in its stock form) for the HP filter isn't going to happen - sounds much worse when it's in the path. Works fine for bass, tho.
 
Hi Zigzaflux

I can not give an answer to a possible advantage of the input tranny, but maybe some hints for your other two questions.

I have seen people using the existing RC coupling in the amp to achieve the 6dB HP. Just select the combination of coupling capacitor and gridleak resistor that gives the desired frequency cut. Generally this would be the largest possible gridleak resistor so that the driver stage is not unnecessarily loaded, and a very small capacitor, probably in the pF range.

Besides, on connecting the DCX2496 to the system. I have been reading about Gary Pimm experience, and I think you should read it to! http://www.pacifier.com/~gpimm/behringer.htm

Good luck! Erik
 
For me it would come down to two things:
- Does the use/not use of the second set of Tx's mean caps will be needed/used?
- How long is the run between pre and power?

And possibly, if there is too much signal level available, then a step down somewhere offers some potential advantages.

I'd use the iron.
 
ErikdeBest said:
I have seen people using the existing RC coupling in the amp to achieve the 6dB HP.

Right, that's what I'm currently doing with my unbalanced input power amp. Once going to balanced, things get a little more sticky. I could just insert a capacitor in the interconnect line, and that may work sufficiently with the input transformer secondary load plus grid leak. Better than inserting two caps on the phase split side, I would think. There might be some resonance issues to worry about, but I guess only testing could confirm that for sure.

I have been reading about Gary Pimm experience, and I think you should read it to!

I read his site on a monthly basis (since it's come back). I especially like the Tabor design, and his discussion of the process and evolution it's taken. A lot to digest from the expert. I have been considering a DC coupled design similar to his, but the instabilities have frightened me away. Especially since my track record of obtaining well matched triodes hasn't been stellar. I will say this, tho. At some point in the future, I may upgrade in I/O and power supply of my DCX2496. At that point, I may insert it in the MTM path. But that project is the lowest hanging fruit, know what I mean?

Brett said:
For me it would come down to two things:
- Does the use/not use of the second set of Tx's mean caps will be needed/used?

Not exactly sure of your question. Regardless of tranny or not, I will be implementing a 6dB passive filter. I suppose the implementation of said filter will depend on tranny or not, tho. I think my dilemma really boils down to "am I compromising sound quality by inserting a potentially unnecessary transformer?", and "do the benefits outweigh the drawbacks?" Isn't this the million dollar question for DIYAudio ? Do I buy that Mundorf or AudioNote capacitor, etc?

- How long is the run between pre and power?

About 3 feet to the Behringer, plus a parallel 6 foot run to the power amp.

And possibly, if there is too much signal level available, then a step down somewhere offers some potential advantages.

Definitely. My preamp output uses a 4.5:1 stepdown for high drive capability, yet still puts out more than 3V rms at 0dB. When I get a good handle on the input sensitivity of my power amp, I may make that input splitter (if used) a 2:1 stepdown or something. My tendency is to keep everything 1:1 at this point, since with a non-GNFB DHT amp excess gain is somewhat of a luxury. No 12AX7's, thank you. Already have that, want to try something different.
 
zigzagflux said:
Not exactly sure of your question. Regardless of tranny or not, I will be implementing a 6dB passive filter. I suppose the implementation of said filter will depend on tranny or not, tho. I think my dilemma really boils down to "am I compromising sound quality by inserting a potentially unnecessary transformer?", and "do the benefits outweigh the drawbacks?" Isn't this the million dollar question for DIYAudio ? Do I buy that Mundorf or AudioNote capacitor, etc?
If the iron removes the need for a cap, then I'll go with quality iron every time. M or AN cap? Simple. No.
zigzagflux said:
Definitely. My preamp output uses a 4.5:1 stepdown for high drive capability, yet still puts out more than 3V rms at 0dB. When I get a good handle on the input sensitivity of my power amp, I may make that input splitter (if used) a 2:1 stepdown or something. My tendency is to keep everything 1:1 at this point, since with a non-GNFB DHT amp excess gain is somewhat of a luxury. No 12AX7's, thank you. Already have that, want to try something different.
12AX7 are something you have thrown in front of your shotgun for target practice.

If you don't need the gain, then don't implement it.
 
First order roll-off in balanced mode can be accomplished with just two resistors and a single cap - couldn't be much simpler. One resistor in series with each phase at the input, cap shunted across the balanced inputs after (downstream of) the resistor pair.
 
Thanks, Kevin, but I believe that would be a LP filter. I am interested in the HP filter. By the way, your ST70 mod is still doing well, and everyone that listens is impressed. Just time to move on to the next project.

I see a single cap in series with the balanced line, feeding into the single coil primary input transformer. Secondary of this input transformer is terminated with some resistance, say 2 x 50K. Reflected to the primary, this impedance would then form my 6 dB HPF.

That assumes a cap in this location is friendly with the transformer (resonances, leakage, parasitics, which hurt my head a little bit with audio transformers).

Without the input transformer, I would think I need two caps, no?
 
zigzagflux said:
Thanks, Kevin, but I believe that would be a LP filter. I am interested in the HP filter. By the way, your ST70 mod is still doing well, and everyone that listens is impressed. Just time to move on to the next project.

I see a single cap in series with the balanced line, feeding into the single coil primary input transformer. Secondary of this input transformer is terminated with some resistance, say 2 x 50K. Reflected to the primary, this impedance would then form my 6 dB HPF.

That assumes a cap in this location is friendly with the transformer (resonances, leakage, parasitics, which hurt my head a little bit with audio transformers).

Without the input transformer, I would think I need two caps, no?

Hi Zigzag,
Yes that is quite true, LPF only unfortunately, with a transformer as you surmise you can get away with a single cap. If you make the resistor in your HPF 600 ohms I think the interaction between your input transformer and the cap will be minimal - doing otherwise I can't really say, suspect you might have to do some tweaking. Can your line stage drive a 600 ohm load?

Kevin
 
kevinkr said:
Can your line stage drive a 600 ohm load?

Well, let's see. Rp of my 6n6p is around 1.8k, but being this is a differential stage, I would say the output resistance is twice that, so 3.6k. Stepping down through a 4.5:1 transformer gives me an output resistance of 3.6k / 4.5^2 = 177 ohms. Is this correct?

If so, driving a 600 ohm load seems a little iffy to me. What did you have in mind, such that such a low resistance is used? A 22nF with 100k works, too, for a cutoff of 80Hz. That was along the way I was thinking.

aletheian said:
How about tossing out common mode noise.

But doesn't the output transformer of my preamp do just that? How much common mode noise will I pick up in the 6 feet of balanced interconnect? If it's significant, I guess the transformer would be the way to go.

Then again, a differential amp has good CMRR, which is proportional to 1/RL, so I can now see some benefit to Kevin's suggestion of a smaller load.

I ordered the Tribute transformer; I guess I'll just have to try each and hear the differences.
 
Status
Not open for further replies.