Impedance and related attenuation question based on using transformers for "colo

Hi all,

I plan on using a pair of 1:1 600:600 isolation transformers on my Neotek Series 1E console 2-mix outputs.

I would like to switch between running them at 1:1 and 1:2 (switchable secondary between 600 and 150 ohms).

Also, if it were possible to put an attenuator on the output before the input into the two-track (Alesis Masterlink).

I guess I am wondering about any problems with impedance, I ask because I am still learning about impedance.

Seems like the Neotek outputs are going to be a couple of hundred (100, 200, 300) ohms? And the input to the two track is going to be something like 1k or more?

And that I read going straight from the Neotek to the two-track should be fine (low output impedance to high input impedance = bridging).

But now I am introducing the 600:600 (or 600:150) transformers...

So I am going from < 600 ohms up to "only" 600 ohms and then out to 600 (or 150 ohms).

My concern: Would going out of the Neotek into the 600:600 transformers cause any problem/damage? I thought I read it's better to have something like a roughly 10x difference between output to input? e.g., 100 ohms to 1k ohms, etc.

Secondarily, what kind of attenuator would I need on the output, after the transformers? A "U"?A "T"? Not sure what output impedance I am wanting to have on that output before it hits the two track... which is dictated by the kind of pad to use... and I think I only want maybe 6 to 10 dB attenuation, so neither an adjustable nor a rotary with discrete steps but just a switch.

Help a newb understand?

Thanks,
Chris

fazeka said:

I would like to switch between running them at 1:1 and 1:2 (switchable secondary between 600 and 150 ohms).

Click to expand...

Why not? But don't expect significant difference between positions.

Also, if it were possible to put an attenuator on the output before the input into the two-track (Alesis Masterlink).

Click to expand...

It certainly is.

Seems like the Neotek outputs are going to be a couple of hundred (100, 200, 300) ohms? 

Click to expand...

The 70's is a pivotal period where audio systems switched from matching (600r output Z driving 600r input Z with 6dB loss, inherited from telephone systems) to bridging (output Z <100r driving input Z >10k with negligible signal loss). However, and though information is sparse, I would think the actual output impedance of teh Neotek is less than 100 ohms.

  And the input to the two track is going to be something like 1k or more?

Click to expand...

Likely to be 10k or more.

And that I read going straight from the Neotek to the two-track should be fine (low output impedance to high input impedance = bridging).

Click to expand...

Yes.

So I am going from < 600 ohms up to "only" 600 ohms and then out to 600 (or 150 ohms).

Click to expand...

You have to understand that there is nothing like a nominal impedance for a transformer; that's a convenient way of describing one of the possible usage of a transformer. Actually a transformer reflects impedances; the impedance presented by the primary is a reflection of the impedance that is presented to the secondary. If the secondary is unloaded, the reflected load at the primary is infinite. That is, for a "perfect" transformer, which just doesn't exist. With the secondary unloaded, the transformer is just an inductor; as such its impedance is variable with frequency, very low at DC, and higher up when increasing frequency, until parasitic capacitance takes its turn and shunts HF.  In the meantime iron losses absorb some of the energy, so the impedance is not infinite. A xfmr designer has to make compromises between all these constraints, and ends up with a product that has a well defined scope of applications; that's where the xfmr gains its calling as a 600 or 10k or whatever.

My concern: Would going out of the Neotek into the 600:600 transformers cause any problem/damage?

Click to expand...

No.

I thought I read it's better to have something like a roughly 10x difference between output to input? e.g., 100 ohms to 1k ohms, etc.

Click to expand...

This criterion will easily be achieved, even if the xfmr is wired as a 1:2. The input Z of the Masterlink would reflect as 2.5-5 kohm at the primary, which the Neotek can drive easily.

Secondarily, what kind of attenuator would I need on the output, after the transformers? A "U"?A "T"?

Click to expand...

Almost any kind of pad can do. An "L" or "U" pad should do.  An "H" or "T" pad is usually used when working in matching mode, where all the impedances must be equal, which is not the case.

Not sure what output impedance I am wanting to have on that output before it hits the two track...

Click to expand...

  Doesn't make a difference when driving a 10+kohm differential input.

  which is dictated by the kind of pad to use... and I think I only want maybe 6 to 10 dB attenuation, so neither an adjustable nor a rotary with discrete steps but just a switch.

Click to expand...

Why not a good old potentiometer?

Thanks Abbey for the validation of some of my understanding and lots of education I need to sit down with.

abbey road d enfer said:

Why not a good old potentiometer?

Click to expand...

I guess I thought that a pot would "mess up" impedance and thus frequency response... but with the above that you discussed about, maybe that would affect not a line level but more a microphone output...?

So, you're talking like a dual-ganged pot? What value would be ideal?

fazeka said:

So, you're talking like a dual-ganged pot? What value would be ideal?

Click to expand...

Dual-gang if you want a true-balanced attenuator, which is not a necessity. I would use anything between 2.2k to 10k. 2x1k to 2x5k for balanced.

Cool, thanks Abbey!