Question about wiring attenuator after Jensen JT-11p-1

[JW]

I'm trying to insert a passive attenuator after my DAC and before my speakers.

Because of my long cable run from the machine room where the DAC is (75 ft or so) and because the attenuator is dual 10K, Classic Audio Products, Inc. I'm using Jensen JT-11p-1 transformers to de-balance signal, so goes like this

DAC (IZ radar 50 ohm source impedance) > 75ft cable > Jensen JT-11p-1>Elma 10K attenuator>short unbalanced run to 10K input impedance speaker amp.

My question is, in the Jensen application notes https://www.jensen-transformers.com/wp-content/uploads/2014/08/jt-11p-1.pdf it has a damping network of 13k and 620p. Should I omit this if I have the Elma switch on the secondary? And what about "RL?"

 

[emrr]

Yes, omit.

 

[JW]

Thanks Doug,
so "RL" from the Jensen note would essentially be my attenuator?

 

[emrr]

Yes. You’re already fully loaded.

 

[scott2000]

You’re already fully loaded.

Is the input impedance of the amp in parallel with the pot as it relates to the transformer? I've never been able to comfortably understand how this works.

 

[ccaudle]

Is the input impedance of the amp in parallel with the pot as it relates to the transformer?

If you draw out the circuit:


I called the upper part of the pot Rpu, and the lower part Rpl (with the wiper being the divider between upper and lower); Ramp is the input impedance of the amp (assumed resistive for simplicity).
(Please excuse the crayon art, I didn't want to fire up a full schematic editor).
If you re-arrange just a little you can see that the upper part of the pot is in series with the parallel combination of the amp input and the lower part of the pot:


The exact values of Rpu and Rpl will of course change as you move the pot.

 

[scott2000]

Thank you!. Yeah, I was definitely lost..

 

[nsalis]

Assuming that the "Elma 04, 24-10k, Shorting Rotary Switch, CAS000947" mentioned by JW is the Elma A4 switch in the 2 waver/ 10k Ohm version (data sheet attached), I have to admit that I find their specifications misleading. Elma specifies an input impedance of 10k ohms. According to Elma's data sheet (attached), it is a simple series resistor ladder and not a voltage divider, especially not one with a constant input impedance. This means that the actual attenuation that occurs depends heavily on the input resistance of the subsequent device.
For the output of the Jensen transformer this means that it has to deal with varying load impedances, at least the input impedance of the amplifier and a maximum of the input impedance of the amplifier plus 10k ohms.
Long story short: The combination R1+C1 mentioned in the Jensen Application Note must be installed because the load impedance will definitely be greater than 10k ohms.
RL does not have to be installed, RL symbolizes the input impedance of the load.

If you prefer a ready-made transformer box, you could use the Jensen Pi-2XX model (enclosed). This has the necessary components already installed.

ns

 

[MagnetoSound]

No. The amp is in parallel with the attenuator (which is grounded at the bottom). The load varies between 5k (with the attenuator at full volume) and 10k (with attenuator all the way down).

As Doug says, fully loaded and no need for the Zobel.

 

[nsalis]

@MagnetoSound: Did you see the Elma attenuator schematic which I had attached? This shows that it is an asymmetrical construction whose ground connection is connected through between input and output. If an attenuation of 0 dB is selected, the audio line from the input and output is also connected through. There is no resistor to ground. This means that the amplifier sees the output impedance of the source (because the Jensen Jt-11P-1 is a 1:1 transformer).

When Elma's "10k Ohm" attenuator is switched to a position other than 0 dB, one or more resistors are connected in series with the audio signal and increase the input impedance of the load as seen by the output of the Jensen transformer.

According to JW, the input impedance of his active speaker (unknown brand/model) is 10k ohms.

The Jensen transformer therefore sees a load of 10k ohms (Elma attenuator at 0dB) or 20k ohms (Elma attenuator at max. attenuation) at its output. According to the data sheet of the Jensen transformer, in this case (Z Load > 10k Ohm) the RC damping network must be connected to the secondary side of the transformer.

I think that list members "emmr", "ccaudle" and "MagnetoSound" mistakenly assume that the Elma attenuator is a voltage divider whose input impedance is always 10k ohms regardless of the attenuation. That's not the case.

Study the Elma datasheet I included which shows the wiring diagram of the Elma attenuator.

If I were in the situation like JW, I would do the following:

I would connect a Jensen PC-2XR (or a DIY circuit with the JT-10KB-DPC) between the source (balanced 47 Ohm output 0VU = 4 dBu) and attenuator. In addition to isolating and unbalancing, this also ensures the required 12 dB attenuation (4 dBu to -10 dBV) which is helpful for the active loudspeaker with its unbalanced consumer level input (-10 dBV). As an attenuator I wouldn't use the Elma stepper switch but rather a desktop stereo potentiometer like Fostex PC-1 (see appendix).

(BTW: The Fostex PC-1 has a stereo 10k Ohm log. potentiometer.)

ns

 

[nsalis]

If you don't want to use the PC-2XR and the Fostex PC1 but want to make something yourself, Jensen AS012 (enclosed) applies, and you can use a good stereo potentiometer with approx. 47 kOhm resistance and a logarithmic curve (e.g. Alpha RV24BF-10-15R1-A50K-LA or Citec 27ESB473MMF50NF) as a volume controller between the Jensen and the active loudspeaker.

 

[nsalis]

I must apologise to the previous speakers. I had misunderstood Elma's wiring diagram. The Elma attenuator is actually a voltage divider, not just a chain of serial resistors. For example, at the position "-6 dB" you have a longitudinal resistance of (R23 + R22 + R21 =) 4940 ohms and a cross resistance of (R1 ... R20 =) 5060 Ohm. In fact, this results in a halving (-6 dB) of the voltage when the load is high-impedance.
In your case, where the load is only 10k ohms, the attenuation will be greater. Capi takes this into account in his data sheet in that " ... in a typical ... control room monitor circuit..." attenuator position 4 is -11dB (and not -6dB as on the Elma data sheet).

The load that the Jensen transformer sees is actually 10k ohms. This means that the RC damping network at the output of the transformer can be omitted.

Nick Salis

 

[JW]

Thanks folks. I just checked back in on this thread, after having wired it up. I must admit, there is some difficulty in understanding these multideck switches. Thanks for coming back to clarify Nick.