MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

[New Module] nw2s::io Balanced line transceiver
MUFF WIGGLER Forum Index -> nw2s Goto page Previous  1, 2, 3 [all]
Author [New Module] nw2s::io Balanced line transceiver
scottwilson
It's primary purpose is to improve audio-path workflow between your modular and your digital multitrack, mixing console, or effects. That it's DC coupled makes it _possible_ to use it for CV/Gate, but doesn't mean it's perfect out of the box.

Even the expert sleepers software which is output only needs to be calibrated. If you want this to be pitch-perfect in and out, you'll need to do some calibration as well... For instance 0.5dB which is a very very small amount when it comes to audio, is the equivalent of 100mV in the CV world - that would be horribly out of tune.

Rather, you'd want to create a loopback, figure out what the overall gain is, and adjust that track's output by that much. Personally, I haven't yet tried it, but it's on my TODO list to be able to document what the process would be and to see if you can get the tuning to work well.

-s
Noha
thanks Scott, that answer makes total sense. I think the module is a brilliant idea and one I am very seriously considering integrating into my workflow, but it sounds like for my purposes I'd be better off using it only for audio. Perhaps for things like LFOs or more chaotic modulations it would be fine to route the CV through attenuation, but for precision CV I'll stick with direct cabling.
Noha
le double
7C
you probably have it on the far right in your case, which is not possible because of the big cables on the right side of the module. it´s just a minor problem, but you loose 2 to 4 hp and this is quite frustrating, when you´ve spent ages to setup a case very specifically and it doesnt work out with the hp´s wink

and: my unit´s first two ins were buggy, most of doepfer patchcables would just come out again, confused does work with other cables though.

apart from that - sound and built quality is nice!
thresholdpeople
So I've had the :b for a while now, but I'm a bit shamed to admit I haven't used it so much as things have been changing around in my studio...

At the moment, I'm looking to set up a computer to interface with some hardware, I've got a motu 828mk3 on the way.. I believe the MOTU puts out a max +/-2.5v.

Does the :b convert line level to modular level? I'm interested in both audio and CV applications.
scottwilson
You mean the 'io? (not the 'b)

Yes, the active version of the IO will increase the signal coming from your MOTU by 6dB - doubling the output voltage.

-Scott
thresholdpeople
d'oh! yep the io!

Great! Thanks Scott.

Is it possible to run the io at +/-15v?
scottwilson
Yes, it is possible with one caveat... I know of one customer who is doing it and customized the panel to fit in a Buchla (or Serge?)

The LED bias voltage is only correct when running at +/- 12V, so if you want to run at +/- 15V, you will need to swap two resistors on the power PCB.

I'd be happy to do that for you if you like.

I did correct this on the second run of the power supplies. I haven't tested them at 15V yet, but in theory, they'll run without any modifications.

Scott
thresholdpeople
Thanks Scott! Let me think about it.

I have an unused +/-15v power supply laying around at the moment, that's why I asked. I need to figure out how much current it can provide, I'd like to run some Expert Sleepers hardware with it as well. If it can indeed power everything, I'll get in touch -- I appreciate it! -- otherwise, I'll just get a +/-12v power supply, and keep things simple.
scottwilson
7C wrote:
you probably have it on the far right in your case, which is not possible because of the big cables on the right side of the module. it´s just a minor problem, but you loose 2 to 4 hp and this is quite frustrating, when you´ve spent ages to setup a case very specifically and it doesnt work out with the hp´s ;)

and: my unit´s first two ins were buggy, most of doepfer patchcables would just come out again, :confused: does work with other cables though.

apart from that - sound and built quality is nice!


Very sorry for the patch cables popping out. I'm getting new ones in stock finally, so if you wanted a replacement, I'd be happy to help you out.

Also, I had promised a photo of how to set up the cables if you wanted to align it right up against the edge of your case. I've got a couple of them and one is done like this - right against the Tip Top Z rails.

Done like this, the cables are completely within the panel dimensions and if you pinch them in enough, they don't push on the pinheaders at all.

fragletrollet
Am i right in that the balanced version is not available as a diy kit? Panel/PCB combo`s not available for the balanced version?
scottwilson
Correct - It's got a lot of small SMT to be able to squeeze everything in there. Goes down to 0402 for some of the stuff!

s
SamUK
I'm thinking of picking up a balanced nw2s::io, but I really don't get pro-audio terminology that well. I'm a clever(ish) chap but dBu, dBV, dBfs all drive me nuts. I'm okay with V, Vrms and Vpp.

So 10Vpp = -5V to 5V and 5Vpp = -2.5V to 2.5V.

If I take a 10Vpp signal from my Eurorack into the (balanced) nw2s::io, it will come out the DB25 as a 5Vpp balanced signal. Right?

If I then take that 5Vpp balanced signal put it back into the input of the nw2s::io it will come out the minijack as a 10Vpp signal?

Let's say I want to take a signal from Eurorack, via the nw2s::io into an Eventide Timefactor (it's unbalanced):
- 10Vpp into the nw2s::io comes out as 5Vpp balanced
- 5Vpp balanced into the Timefactor comes out 5Vpp unbalanced. Let's just assume that the Timefactor outputs at the same volume as the input. I'm also assuming that it's okay for the Timefactor to accept a balanced signal, it will either ignore the sleeve or short it to ground. What about the nw2s::io, will it be okay with that?
- 5Vpp unbalanced goes into the nw2s::io and comes out at....? This is the bit I'm not so sure about, is there some signal loss due to how the common mode rejection works?

Any guidance helping me get my head around this would be greatly appreciated!
C14ru5
SamUK, while most of the points in your proposed setup are okay, I'm concerned that your Timefactor may not handle a 5Vpp signal without clipping. So before anything, you should at least verify that the Timefactor has the headroom to deal with such signal levels. If it turns out that a 5Vpp signal is too hot for the Timefactor, a better option for i/o modules would be the Ladik A-520/P-520 combo which can be set to a -20dB/+20dB (10Vpp/1Vpp) attenuation/amplification. Note that the Ladik modules just use plain TS (unbalanced) minijacks.

I'd say that the typical intended scenario for the nw2s::io is to hook up your modular multichannel signal to your professional audio interface, which usually has significantly more headroom than guitar pedals.
SamUK
Thanks for the reply C14ru5, the main usage would be with my audio interface.

The Timefactor was really given as an example of an unbalanced device, though out of curiosity I did check just now, it copes pretty well, if you fiddle with some of the gain settings, it will output an almost 10Vpp triangle wave that's been fed in. Once you add feedback it gets a bit more hairy.. Pretty cool though.

Mainly I just wanted to get the maths correct, particularly with respect to the unbalanced signals, i.e. what happens if I put an unbalanced signal into the nw2s::io. Does anyone one know the answer to that bit?

Edit: Those Ladik modules look really useful btw.
scottwilson
There's a few things adding to the confusion...

Balanced signals aren't measured relative to ground, but as a differential signal - two inverted copies of the same signal relative to each other. When you're talking unbalanced (like with the time factor), typically you would only use one leg of the balanced signal and reference it to ground. This would cut your usable signal by half when going from balanced gear to unbalanced gear.

Also, the 'io doesn't just do a straight balanced to unbalanced conversion - It's got a fixed gain structure that makes the studio<->modular signals a little more natural.

So take modular to studio on the nw2s::io. There is a 6dB overall gain reduction.

If you put a 10V p-p signal (+5/-5) in the modular side, you will get two 2.5V p-p signals out - each one measured relative to the other is 5V p-p.

5V p-p is about 7dBu.

If you run that into a balanced studio interface that's calbrated to +4dBu = 0VU = -18dBFS, then your input meter would read about -15dBFS with a 10V p-p modular signal.

Let's say that instead of running that into a balanced studio interface, you run it into an unbalanced timefactor. The time factor manual gives no specific operating levels besides "line level" - Let's assume this means that 0VU = -10dBv and we don't know exactly how much headroom that gives you.

The eventide pedal will only see half of what the studio interface sees because we are dropping one leg of the balanced signal and measuring against ground instead of against a differential signal.

This will be 2.5V p-p. This means that the eventide pedal will get about -1 dBv at its input. Which is about +9dBVU for that specific pedal. Since they don't say what "clipping" is, then we don't know with absolute certainty without trying it out that the eventide pedals will take the "unbalanced" signal from the 'io, but signs would point to yes, it should be fine.

Running back into modular, there is a 6dB gain... which inverts all of the above... increasing the voltages by the same amount that we decreased them.

Hope that helps.

s
SamUK
scottwilson wrote:
If you put a 10V p-p signal (+5/-5) in the modular side, you will get two 2.5V p-p signals out - each one measured relative to the other is 5V p-p.


Perfect.

One final bit of confusion about inputting unbalanced signals. If I input a 2.5V p-p unbalanced signal, do I get 5V p-p out the minijack or 10V p-p?
scottwilson
Well, to take the same approach, let's say you put a balanced signal with two opposing 2.5Vp-p signals into the balanced input. This is a total differential signal of 5Vp-p. With the 6dB gain going from balanced to unbalanced, that will create a 10Vp-p signal at the mini jack.

Alternatively, if you put a 2.5V p-p unbalanced signal, that will be considered a total of 2.5V measured by the differential input. Add 6dB to that and you will get 5V p-p at the mini jack.

-s
SamUK
I thought that was the case, but thanks for confirming. So if I use an unbalanced effects unit I'll need to add 6db of gain (2x) on the euro side to maintain an overall unity gain.
Multi Grooves
scottwilson wrote:
This is primarily meant to be an interface - balanced to unbalanced and back. It's designed to be as simple as possible with a pristine signal path. At this point, the only thing in the signal path is just the 1246 or 1646 IC (input vs output). Adding more gain circuitry, particularly adjustable gain would increase the cost and compromise the signal path for the whole device rather than only those channels that need it.

If you have something that's particularly low level that you need to interface, you'll still need to provide some gain - either in the modular realm or prior to patching into the modular. The Gozinta is a great looking module that seems like would be great to have a couple on hand for when the need arises. Hopefully, of the things you're interfacing, you wouldn't need one for each channel.

6dB gain is actually just about right for the difference between pro audio and modular audio. I'm assuming that +4dBu is the nominal voltage level for balanced pro audio and 10V p-p is the nominal voltage level for modular signals. (Note: nominal = ideal compromise between headroom and noise floor)

Here's an example: you have a 0VU (+4dBu) signal coming from your console, sampler, synth, or effects processor. You're going to want that to end up being about 10V p-p. 6dB of gain gets us close without needing any extra circuitry.

Code:

4dBu + 6dB = 10dbU = 6.9V p-p


So if you have something that needs 12, 24 or 36 dB of gain, you're talking about something that's probably 150mV to a volt peak to peak which would be about -24dBu.

Hope that helps clarify a bit more.

Scott


150mV would be a very low level moving magnet stylus through phono amp stage i.e. a quiet signal from a turntable. Theres no way around it, it'd need another gain to be used with modular...


**edit I just saw the date- ignore this**
MUFF WIGGLER Forum Index -> nw2s Goto page Previous  1, 2, 3 [all]
Page 3 of 3
Powered by phpBB © phpBB Group