MixWizard channel connections

[ribrdb]

I recently got an Allen & Heath wz3 16:2:2 mixer to use as part of my console.  It has a separate connector module, so I opened it up to see if I could wire the individual channels directly into my patchbay instead of running a bunch of xlr and trs cables.  Each channel has a 10 pin connector with ribbon cables running to the connector board. I've attached the schematic I've come with by looking the the board. I've got a few questions about this:
- What's the value of the 15k capacitors? They look ceramic, with no marking other than 15k on them. They're buried under the xlr jacks so I can't disconnect them from the board to measure them without desoldering a lot of stuff.
- Why are there all these capacitors for the xlr input, but not the line?

Since I've got a patchbay I'm planning make a simpler connector with just the direct out and xlr input. I'm wondering if I still need all the capacitors in that case.

 

[scott2000]

- What's the value of the 15k capacitors? They look ceramic, with no marking other than 15k on them.

Usually the small numbers are in picofarads.....Never seen one with a K though.......so who knows........Could be a tolerance (i.e. 10%),voltage???(i.e. 40v) or maybe even a material designation???? Not sure.... Got a picture? I'd say to go ahead and check the value but, those are pretty hard to read without a good meter anyhow..... I'd be interested in what they are......

- Why are there all these capacitors for the xlr input, but not the line?

They are connected to the line from your schematic???

 

[Bo Deadly]

The capacitors shunt RF noise before it enters too far into the enclosure. If you're patch bay snake isn't too long, then you don't need them. If the snake is going to be long then you should have them. Something in the realm of 1n would be fine. You should be able to measure the existing caps in place actually because there are no other DC paths or sources of capacitance. I don't think the 6k8 resistors will affect the results. But note that to stop RF from radiating inside, you would have to put them right at the point at which the cable enters the metal enclosure and connect them directly to the metal enclosure at that point.

A potentially larger issue would be running mic level signal through your patchbay because it could create opportunities for noise. You might want to be particular careful about shielding and ground connections.

 

[ribrdb]

The xlr connector is normalled to the line jack. When something's plugged in the line jack connects directly to pins 6 & 7, and the xlr runs through the capacitors to pins 8 & 9.
If nothing is plugged in to the line jack, the switches close and the xlr signal is also routed to pins 6&7.
Here's a picture of the "15k" capacitor. Sorry it's not very good, I had a hard time getting the camera to focus on it.

 

[ribrdb]

The capacitors shunt RF noise before it enters too far into the enclosure. If you're patch bay snake isn't too long, then you don't need them. If the snake is going to be long then you should have them. Something in the realm of 1n would be fine.

You just mean the length of the snake from the patchbay, right? It's close to 1m between the patchbay and the mixer. But probably at least 10m from the patchbay to the mics.

You should be able to measure the existing caps in place actually because there are no other DC paths or sources of capacitance. I don't think the 6k8 resistors will affect the results. But note that to stop RF from radiating inside, you would have to put them right at the point at which the cable enters the metal enclosure and connect them directly to the metal enclosure at that point.

If I measure them in circuit they're reading at about 100nF.
That brings up another question I had about grounding. The ground plane on the connector circuit board is marked "CHASSIS", but there's no connection to the chassis. It looks like there's some teeth on the jacks intended to connect the jack ground pin to the chassis, but both sides of the chassis are painted so neither the teeth or the screws/nuts will touch bare metal. Unless the teeth penetrate the paint, but it doesn't really look like they do.  There's also DGND and SGND coming from the power supply, but there's no connection between these and CHASSIS. One channel (15) has pin 3 connected to SGND.  There's also no connection between CHASSIS and the channel cards, so I'm not sure how phantom power works.

So how would I handle the grounds in my snake? Bus them together? Do I connect the bus to the chassis?

 

[scott2000]

Sounds like a lot of work....

Those caps could be reading wrong if your meter isn't up to the task...Have you tried reading them with the meter on different settings to see if it's consistent????

I did come across something where 15k could be 15000pf which would be .015uf????.....anyway, squarewave said it best and I wouldn't change anything and those caps will be good for a long long time in that spot.......

Not sure about the grounds. If you do some continuity testing, you'll be able to see if there is a chassis ground??? Most likely it's connected somewhere you may not be seeing with it disassembled....Signal ground may be referenced differently  so, it may not be chassis grounded??

Ddground???? some kind of digital ground for a processor or something???

Continuity test is pretty helpful to figure things out.......powered off of course...

This is a current model I'm seeing so, schematics may not be available but, have you tried asking for them? Are there older models available that may have similar layouts? Maybe a schematic from those would be helpful to figure out what's going on.....

Good luck!

 

[Bo Deadly]

You just mean the length of the snake from the patchbay, right? It's close to 1m between the patchbay and the mixer. But probably at least 10m from the patchbay to the mics.

If I measure them in circuit they're reading at about 100nF.
That brings up another question I had about grounding. The ground plane on the connector circuit board is marked "CHASSIS", but there's no connection to the chassis. It looks like there's some teeth on the jacks intended to connect the jack ground pin to the chassis, but both sides of the chassis are painted so neither the teeth or the screws/nuts will touch bare metal. Unless the teeth penetrate the paint, but it doesn't really look like they do.  There's also DGND and SGND coming from the power supply, but there's no connection between these and CHASSIS. One channel (15) has pin 3 connected to SGND.  There's also no connection between CHASSIS and the channel cards, so I'm not sure how phantom power works.

So how would I handle the grounds in my snake? Bus them together? Do I connect the bus to the chassis?

Actually you do need to consider the full length of the cables to the signal source (mic) and not just to the patchbay. So I would keep the caps or add them if necessary. But 100n seems high to me. Even if the mic is 150R that's like a 10kHz cutoff. Is it possible those caps are for the line inputs and not mic inputs?

For grounding I would not buss them together. Just pass pins 1,2,3 straight through to corresponding pins on the mixer.

But are you really going to just hardwire your patchbay to this mixer? Or are you going to substitute the on the internal ribbon connector?  I can understand wanting to minimize connectors. I'm a huge fan of patchbays for that reason. But it could be trouble hardwiring these things together. You need to be able to disconnect these things.

 

[ribrdb]

The "15k" caps are connected directly from the xlr pins to ground. I'm not sure I trust my cheap radio shack multimeter though

My idea was to use something like these connectors on the end of the snake, and just plug the snake into the channel cards:
https://www.pololu.com/product/1913
If I'm going into the channel card ribbon connector there is no ground pin.(except one channel does connect the otherwise unused pin 3 to SGND)
Maybe it's just not worth the effort to bypass the two extra connector pairs (xlr+ribbon).

One more question. There's some corrosion on one end of the connector board. Should I try to clean this off:

 

[Gene Pink]

The "15k" caps are connected directly from the xlr pins to ground. I'm not sure I trust my cheap radio

Those itty bitty little things? I'm pretty sure they would be 15pF. Does nothing in the audio range, but shunts RF noise to ground,  way up in the cell phone range. To complete the RF filter, there should also be a ferrite beads on each line right before the cap.

I have another version of the mix-wiz, good sounding board, the only problem is engaging the global phantom power switch. If you forget to bring the main faders down first, giant thump through the PA.

Gene

 

[Khron]

Even on the silkscreen, next to the component designation (C___), i can see a "15". By more-or-less standard, unless there's another multiplier mentioned (n, u), the value's gonna be in pF.

I have the service manual for the WZ3 14:4:2 (safe to assume, virtually identical apart from two inputs and the group / output section), see attachment. Sure enough, the caps off each input XLR are indeed "15P/DISK", ie. 15pF.

 

[Monte McGuire]

The '15K' designation is 15pF 10% tolerance. 'J' is used for 5% tolerance.

Those 'troublesome' XLR and 1/4" cables provide shielding and interference rejection. A ribbon cable won't do that. So, you could spend a lot of effort rigging up something that will work poorly and have no re-use value, or you could just wire up some cables that will always have usefulness. I'd just wire up cables and stop re-engineering the insides of the console.

 

[ribrdb]

Even on the silkscreen, next to the component designation (C___), i can see a "15". By more-or-less standard, unless there's another multiplier mentioned (n, u), the value's gonna be in pF.

That 15 on the board is because this is channel 15. Thanks for the schematic, I'd only found the older model which doesn't include these capacitors.

Those 'troublesome' XLR and 1/4" cables provide shielding and interference rejection. A ribbon cable won't do that. So, you could spend a lot of effort rigging up something that will work poorly and have no re-use value, or you could just wire up some cables that will always have usefulness. I'd just wire up cables and stop re-engineering the insides of the console.

Well I was trying to get rid of the existing ribbon cable by putting the connector on the end of the shielded cable. But yeah,  at this point I think I'll probably just use standard cables.

 

[pucho812]

I am sure it was answered already, I skimmed through

a 15K cap would b as follows. the 2 numbers indicate the value , the letter indicates tolerance.
Capacitors will have a number on them like 103, 104, 224 ... . The last number represents the number of zeroes. All values are in picofarads.
For example: 103 becomes 10 + 000 (3 zeroes) pF = 10000 pF = 10 nF
For example: 224 becomes 22 + 0000 (4 zeroes) pF = 220000 pF = 220 nF = .22 uF

A capacitor that has a decimal is typically measured in uF.
For example: .47 = .47 uF = 470 nF

The letter on a capacitor designates the tolerance.
J = +/- 5%
K = +/- 10%
M = +/- 20%

For example: .47K becomes .47 uF +/- 10% of .47 uF. Therefore the measured value of the capacitor can be within the range of .423uF to .512uF.

The voltage listed on the capacitor is its operating voltage.
For example: a capacitor with 250V marked on it may be operated at any voltage up to, but not exceeding 250V. If 250V is exceeded, the capacitor may not work properly. It is therefore possible to replace a 100V capacitor with a 250V capacitor if a 100V capacitor cannot be found. However, the 250V capacitor may be physically larger than the 100V capacitor, so it may no longer fit in the circuit with the other components.

Any other markings on the capacitor are probably lot or batch numbers, and are only used by the manufacturer.

Now in the case of 2 numbers then the cap is that value in pF so your 15K is 15pF +/-10%