Stereo to mono for subwoofer?

[Jorj]

This should be easy, but it seems that it's not. I am using my Topping DX7s to directly feed my 2 channel system amp when the HPs are not plugged into the front. All is well. I'm wanting to plug an Elac Debut S10EQ subwoofer in for a bit of low end-fill, but I have no mono outputs, and the sub only accepts a single line-level input.

There is a lot of ignorant chatter on the teh interwebz about this, and some good stuff too, but I need illumination from those that are more technically adept on how BEST to approach this. I am certainly capable of DIYing a few resistors to a Y-cable or making a summing adapter box, but given the output impedance of the DX7s and the worry of this causing substantial crosstalk, I'm not sure if this is really the best option.

Given the rated output impedance of the DX7s to coax is 100 ohms (unbalanced) and referencing the linked schematics, what would be the best resistance values to use to ensure that the summing function does not create unacceptable crosstalk in the signals to the amp?

Also, there has been mention of inverting one channel on some forums, which actually does not make sense to me, as it seems that would cause cancellation issues.

Finally, if anyone knows of a competent and inexpensive solution for sale somewhere, I'd be grateful for that info also.

-edit-
corrected impedance values

 

[dc655321]

I take it you already have the Elac sub in hand, and purchasing something with different connection options is off the table?

 

[Jorj]

I take it you already have the Elac sub in hand, and purchasing something with different connection options is off the table?

Yes, it's one I'm playing with to see if a sub is any help at all. My floorstanders go quite low already, but fun is fun, right?

 

[dc655321]

My floorstanders go quite low already, but fun is fun, right?

IMO, you can never have too much quality down low
There is so much value added in the bottom 20-40 Hz of music. Well, at least the music I listen to.

 

[Jorj]

IMO, you can never have too much quality down low
There is so much value added in the bottom 20-40 Hz of music. Well, at least the music I listen to.

Agreed. The C-9's can go down to 34 hz in my room, but only on certain tracks, so I want something dedicated to handling those lower registers.

 

[dc655321]

I'll be watching this thread for possible solutions with much interest.
Looking to put a system together in the not too distant future, and I have been wondering about this corner-case myself...
Good luck!

 

[DonH56]

If all you want to do is sum the signal to the sub, you can build a simple passive summer, and determine crosstalk that remains from knowing the value of the resistors and the input impedance (resistance) of the sub.

If the source is 100 ohms and the divider Rs and sub input are all 10 k-ohm, then the signal from one side (e.g. R output) will be about -50 dB when it hits the other side (e.g. L output). If I did the math right (back of napkin). That would also match levels'ish. Larger divider Rs will reduce crosstalk at the expense of signal level (normally compensated by turning up the sub's gain control).

Or you can build or buy an active mixer circuit to buffer the outputs and reduce crosstalk to negligible (if -50 dB is too high). Or use something like a miniDSP and do the summing in DSP.

Or you could build the resistor divider but add a capacitor at the output to roll off the HF signal at the sub and reduce crosstalk above the sub's frequency range.

Etc.

 

[Soniclife]

I'm not following this, how does crosstalk matter going to a mono sub?

 

[DonH56]

Crosstalk at the input of the summer. Isolation from R output to L output of the preamp is reduced with a simple passive divider to the sub.

 

[Jorj]

Crosstalk at the input of the summer. Isolation from R output to L output of the preamp is reduced with a simple passive divider to the sub.

Precisely. The sub don't care, but all that care and attention in the signal chain to keep the channel separation numbers high can be wrecked if your sub's mono summer allows the summed signal to bleed back into the path.

The question is, what is the number you'd shoot for?

 

[Soniclife]

Crosstalk at the input of the summer. Isolation from R output to L output of the preamp is reduced with a simple passive divider to the sub.

Given the topping has 2 parallel outputs is that an issue if you are using the other set? I.e. Will the unbalanced summed outputs degrade the balanced outputs?

 

[Jorj]

Given the topping has 2 parallel outputs is that an issue if you are using the other set? I.e. Will the unbalanced summed outputs degrade the balanced outputs?

I wondered the same, but it would not make sense to (IMO) to have additional opamps for the various line-level output formats, or would it?

 

[Jorj]

On further consideration, I like Don's thinking. Just go active mixing to eliminate the worry of crosstalk altogether. A simple Behringer MX400 and a couple of adapters will do the trick, and the opamps in the MX400 should keep the channel separation without adding noise to the stereo signal path. The operative word is should. Amir has had acceptable measurements from their stuff in the past, so perhaps the little black box will fill the bill.

 

[RayDunzl]

I wondered the same, but it would not make sense to (IMO) to have additional opamps for the various line-level output formats, or would it?

I'm just guessing, but it looks like mine does...

Two opamps feeding two L/R analog outputs, two feeding the +/- of one set of L/R XLR outs.

---

Looking again, I see two more hiding in there...

Guessing, as I said...

 

[Blumlein 88]

Not as convenient, but as a stopgap measure you could have a stereo headphone to mono RCA adapter. Plug it into the headphone jack when using the speaker rig. Let that feed the subwoofer.

That assumes headphone out tracks the volume control for the other outputs.

 

[Jorj]

I'm just guessing, but it looks like mine does...

Two opamps feeding two L/R analog outputs, two feeding the +/- of one set of L/R XLR outs.

First, kudos to you Ray for tearing into your DX7s! Not to be a spoilsport, but without tracing the layers in the PCB, your sketch might be, well, sketchy.

The RCA outs are stacked, so the image you posted is showing the COAX in and the L channel out. There's also several other IC's that appear to be opamps, so I'm not convinced.

 

[RayDunzl]

First, kudos to you Ray for tearing into your DX7s!

Oh, sorry, that's an internet image of a Benchmark DAC2. DAC3 looks similar, DAC1 doesn't.

 

[Soniclife]

Ray is showing pics of his benchmark DAC.

I would have thought balanced and unbalanced would need different output stages, not that I know anything.

 

[Jorj]

I would have thought balanced and unbalanced would need different output stages, not that I know anything.

I'm right there with ya, buddy.

Does anyone see any issues with using a $25 mixer box and just moving on with the day? Any worries about phases being reversed? Some of the nicer passive boxes have phase inversion and can lift the ground if there is hum.

-edti-
kant spele gud

 

[RayDunzl]

There's also several other IC's that appear to be opamps, so I'm not convinced.

Some of those 8 pin devices are likely voltage regulators.

"The DAC2 has a distributed regulator system. There are 20 separate voltage regulators within the DAC2. Each is dedicated to one specific subsystem. This segregation minimizes crosstalk between subsystems and eliminates the need to deliver regulated voltages over long distances. Voltages are regulated at the point of load."