Strange 'balanced' output circuit?

[RuudNL]

I have been doing some work on an old Mackie 32:8:2 mixer.
At a certain moment I thought I had got everything working, but the 'balanced' (TRS) Aux outputs only produced signal on the tip.
So I took a look at the circuit diagram and thought at first glance that an OpAmp had failed.
However: after taking a closer look, I discovered the strange 20K/20 ohms combination around the 'upper' output OpAmp.
One would expect 20K/20K, so that the OpAmp would act as an  inverting buffer with 0 dB gain.
But the combination 20K/20 ohms will produce an attenuation in the order of 60 dB!
What could be the reason for this (IMHO) 'strange' design?
If the idea was to have an 'impedance balaced' output, only the 120 ohms resistor would have been sufficient.
Any suggestions?

 

[Matt Nolan]

Is it possible they made a typographical error on the schematic, which got carried through to production, and nobody noticed?

 

[moamps]

What could be the reason for this (IMHO) 'strange' design?

A mistake in redrawing the circuit at beginning and  low QC level at the end of  the production process.

 

[RuudNL]

There IS effectively a 20 ohms resistor in the circuit!

 

[ruffrecords]

There IS effectively a 20 ohms resistor in the circuit!

In which case I would be tempted to point this out to Mackie and see how red their faces go.

Cheers

Ian

 

[ruffrecords]

A quick squint at the user manaul reveals that the max L/R mix output level is +28dBu but all others are +22dBu. Might be worth checking th L/R output schematic to see how that differs.

Cheers

Ian

 

[RuudNL]

I have sent an email to Mackie support. Curious what their answer will be!...
(I think the 20 ohms should have been 20K. That would also explain the +22/+28 max. output level, since a working output stage would add the extra 6 dB.)

In the attachment is the main output. This makes more sense.

 

[micaddict]

Heh, heh, I know who I put my money on.  ;D

 

[RuudNL]

The answer from Mackie support (=no answer!)

"Hello,
We are currently unable to offer bench level support due to liability. For your convenience I have included the contact information for the service center that used to do our in house repairs and can answer bench level questions. They get around liability by charging a small fee. Let us know if there is anything else we can do."

 

[micaddict]

Classic ...

 

[JohnRoberts]

The answer from Mackie support (=no answer!)

"Hello,
We are currently unable to offer bench level support due to liability. For your convenience I have included the contact information for the service center that used to do our in house repairs and can answer bench level questions. They get around liability by charging a small fee. Let us know if there is anything else we can do."

That sounds like something you shouldn't step in around the farm.

I suspect Rick Chen designed that mixer, ask him. If you can find him.

JR

 

[abbey road d enfer]

In the attachment is the main output. This makes more sense.

This topology for an EBOS is probably the worst one could think of. To the point that it has been known as the "Tascam problem".
Since the main mix output generally goes to a more sophisticated piece of gear than most auxiliaries (or insert sends), there are statistically not many issues with it.
But for auxiliary sends, that are quite often connected to indiscriminate inputs, it makes "some" sense. I would guess the PCB layout was done, approved and sent to the fab, when somebody found that it was a risk of customer complaint, so it was one of the few possible easy fixes.
At least, it is balanced. And better balanced than a poorly implemented cross-coupled stage.

 

[kambo]

Mackie 32:8:2  ;D ;D ;D
i didnt know it was qualified as a mixing console... some strange reason people likes it...
i heard same songs mixed on M and SSL, and Neve,  by same very reputable engineer,
we picked SSL, Neve was too irony sounding for the style of music ( its was still great tho, if we havent heard SSl mix we would go with the Neve mix for sure)
anyhow, difference wasnt day and night. it was 20000000000000000000 light years....
every body scrapped the M mix, but some stupid deal on the record contract, M mix got released  :'( :'( :'(
songwriter partnership, studio name etc etc....
when we were listening  the same artists album songs, in a random order, we were all crying when M mix was playing...
it was so obvious, and crappy sounding ( great songs btw )....  DJ's were having real hard time to mix songs from that album....

now this topic explains better  8)

 

[abbey road d enfer]

Mackie 32:8:2  ;D ;D ;D
i didnt know it was qualified as a mixing console... some strange reason people likes it...

You must not neglect the historic role of the Mackie "8bus".
It was the right product at the right time for the advent of ADAT. The ergonomics were ok, and the performance adequate for the era (remember grunge?).
Not enough to worship it, though...

 

[kambo]

Mackie 32:8:2  ;D ;D ;D
i didnt know it was qualified as a mixing console... some strange reason people likes it...

You must not neglect the historic role of the Mackie "8bus".
It was the right product at the right time for the advent of ADAT. The ergonomics were ok, and the performance adequate for the era (remember grunge?).
Not enough to worship it, though...

remember grunge?

hahha u must be kidding me  ;D
i was right in the middle of it  ;D

edit/grammar : do you remember some of the Grunge albums that were recorded on 16 trk .... how crappy they were sounding !
big/deep dense cloud of noise all over... no separation at all.... thank God songs were so great...

 

[Humner]

what I also find funny is the cap across the 20ohm resistor.

If 20 ohm it would be a 16mhz corner frequency
If 20k ohm it would be 16khz

Just adds to the confusion I guess  ;D

 

[ruffrecords]

The AUX send level control R443 and its associated op amp U 47B are in a rather odd configuration too.

Cheers

Ian

 

[RuudNL]

Yes, I noticed that also.
Now R443 is more or less 'shorting' the output signal from U46B. (in the 'off' position.)
One would expect that the volume control would act as a voltage divider.

 

[JohnRoberts]

The AUX send level control R443 and its associated op amp U 47B are in a rather odd configuration too.

Cheers

Ian

That was just them trying to be clever.

On paper that topology improves kill and reduces the noise gain of the gain stage when attenuated.

It does trash any ability to use that stage as a differential amp (IMO differential gain stages are useful when moving signals around inside consoles).

======
I've heard stories about the sound quality but don't like repeating anecdotal observations. I never put one, or the Behringer version on the bench.

JR

 

[Audio1Man]

I did some consulting for Mackie for several years before almost all went to China.

Yes R48 is a 20k

U47B is an inverting amp, the R396 (5k1) and the R420 (39k) with a max gain of 17.6dB. R443 is a voltage divider and attenuates the signal, also note the taper on it.
Duke