Norns - Audio In/Out Specs?

avenir · 2020-06-23T18:06:02.920Z

I always have a doubt that torments me😌
I’d like to ask a questions about how to set up routing in a live set using Fates.

Considering that it should not be a good thing to mix line signals with modular signals, I wanted to understand if I use my ER-301 as a normal mixer connecting for example:

2ch out from morphagene to 301
2ch from Intellijel uVCA where are routed two VCO voice to 301
2ch from FATES to 301
(whereas I would use a limiter or gain from the simple mixer channel of the 301 to bring FATES to an acceptable level)

I would risk problems with the Fates audio codec?
I would be happy if someone would help me understand.

Thanks

desolationjones · 2020-06-23T18:13:16.722Z

You will have no problems with that setup. The ER-301 has analog preamp settings for IN 1-4 so you can bring your Fates up to nearly modular level. I recommend that before patching internal (digital) gain so you get a better signal:noise ratio.

Be cautious with ER-301 modular level outputs to line level gear so you don’t damage your hearing. Even if you put attenuating units on every output, the sample preview functions will play at full volume. Brian at O|D has promised a universal output attenuation setting in a future update.

avenir · 2020-06-23T18:21:50.364Z

Thanks:)
the final signal (mixdown) from ER-301 then goes into a Magneto and then into a zdsp where I have the dry controls so I shouldn’t have problems with the volume too high.

zebra · 2020-06-23T18:36:25.868Z

i woud not recommend running modular signals directly into Fates inputs.

Fates uses this codec, which is not rated for input voltages of >5v or so

statics.cirrus.com

WM8731_v4.9.pdf

778.56 KB

since the Fates schematics are not published, it is not easy for me to characterize any voltage scaling or limiting that is happening before input audio hits ADC pins, so can only make conservative recommendations.

you can of course measure levels coming onto the board, and compare them with levels at the ADC pins, to make your own determination.

avenir · 2020-06-23T19:47:29.321Z

Thanks for the answer
I would never send signals from the modular not properly attenuated. my doubt was whether the route of other modular signals inside the ER-301 together with the FATES Outs could be some return signal.

in this configuration I only use the Fates’s OUTs

okyeron · 2020-06-23T22:02:14.380Z

zebra:

since the Fates schematics are not published, it is not easy for me to characterize any voltage scaling or limiting that is happening before input audio hits ADC pins, so can only make conservative recommendations.

Thanks for chiming in. For reference - the schematic is pretty much the same as the sample “RECOMMENDED EXTERNAL COMPONENTS” section of the data sheet on page 60.

xeric · 2020-06-25T13:09:17.426Z

Apologies in advance if this is off-topic – but I made the mistake yesterday of plugging my modulars outputs into Norns’ outputs (and monitored via headphones). I really need to add some labels on there, as I always assume the outputs would be next to the headphones

Anyways, what’s surprising is that you still get audio out, it’s just got a crazy amount of FM and distortion. I really thought my Mangrove, Cold Mac, or RIP was broken, but after removing variables one by one, I finally realized I had plugged into the wrong jacks.

Everything is fine now, but out of sheer curiosity… does anyone know how is that possible? And where is the FM was coming from?

dani_derks · 2020-06-25T13:15:55.186Z

oh! wild.

xeric:

I always assume the outputs would be next to the headphones

stock norns? they are.

image1397×924 16.6 KB

curious what’s happening

xeric · 2020-06-25T13:31:42.240Z

Oh boy – yes, stock Norns. I might need to do some debugging tonight then. I’m getting a clean signal through my outputs but a terrible signal through my inputs?!

xeric · 2020-06-25T17:08:05.841Z

Don’t think so – it’s going through RIP which should bring it down to line level.
Nothings changed in my setup recently

bmoren · 2020-11-19T14:32:00.068Z

Question for those with a bit more electrical knowledge than me:

For attenuating modular levels for input into norns, could you use a voltage divider on the modular signal to bring it down to line level?

If I’m correct in this:
Vout = Vin * (R2 / R1 + R2)

Vin: 5v (10v peak to peak)
R1: 2200Ω
R2: 1000Ω

Vout = 1.56v (3.12v peak to peak?)

This would be a super simple thing to build into a cable or a small box or 2 HP module.

Not sure how this would affect audio quality, but would love input from folks who know more than I do.

I’ve always been a bit mystified by Line vs Modular vs Headphone vs Mic levels…

corpusjonsey · 2020-11-19T14:45:32.264Z

A transformer is more appropriate.

bmoren · 2020-11-19T14:49:36.863Z

Sure, do you have any technical info you could share on this?
I understand this is what RIP & others do, but I don’t have transformers sitting on my bench atm

corpusjonsey · 2020-11-19T15:01:49.593Z

I like this site. It has some great audio circuit examples with the math behind them.

bmoren · 2020-11-19T15:08:58.849Z

This look like a great resource! thank you!

I suppose in the mean time, I’m curious if a voltage divider will work in a pinch / as a cheap stop-gap, and what the impacts / characteristics of it might be – if anyone has tried it. (I’ll probably have a go this afternoon and I can report back too).

I’ve used a simple potentiometer attenuated signal into my H1 recorder for ages and have never had an issue, but I’d love to have something that is ‘fixed’ to make it all a little more consistent in my recordings (and in this case norns input).

anon60283851 · 2020-11-19T15:16:30.770Z

I’d actually argue that a simple potentiometer based volume control, or a fixed resistor divider (both having the same effect), would be more appropriate than a transformer - unless you specifically need galvanic isolation or want to hear transformer coloration in which case you already know what you want.

Why? It’s both vastly cheaper, and does the job as well as you’d want to with an unbalanced, line level (or hotter than standard line level) signal and modern gear. Fixed resistor pads are ubiquitous, and you’re likely to have and use attenuators / mixers in your modular anyway. Low distortion, no discernible change in sound quality, just a quieter signal.

(Edit: I see transformers used for balancing and impedance matching a lot, but it’s pretty rare to see them used in circuits for attenuating a line-level signal. There’s nothing inherently wrong with that either, but a resistor divider will just be neutral, effective and super cheap, so there’s not much reason to use anything else unless you know you need something else than simple attenuation.)

jlmitch5 · 2020-11-19T16:27:57.808Z

A voltage divider or pot, for a single channel should be totally fine I’d think…not 100% sure but I think you’d want to use a pot so that your input impedance isn’t too low and your output impedance isn’t crazy high. My guess is 25-50k…as all the devices your working with are buffered, active (not like a guitar).
Source on the impedance question as it relates to volume pedals:

missionengineering.com

Volume Pedal Q&A – Mission Engineering

When you get into mixing is when active circuits should come into play. Mobenthey/Ciat-Lonbarde (synthmall) thread

I have the EXI[S]T which combines RIP-style cinnemag transformers with a passive volume attenuator. It seems to works great for me, I haven’t noticed any sort of tone suck when the attenuator is down (and sometimes that lift is definitely necessary.

desolationjones · 2020-11-19T16:40:52.984Z

You’re totally in the clear with that concept. A simple voltage divider attenuator is called an “L-pad” in the audio world. There are fancier configurations for passive attenuation if you are concerned with impedance matching and coloration.

bmoren · 2020-11-19T17:13:23.706Z

I haven’t had a chance to read everything yet, but I put this together since I posted as a possible idea for a a super simple passive modular/norns aux send box.

I realize that this is kind of strange since it uses 2 dual gang pots to control vol from each input instead of using a wet/dry knob (panning mixer). But from my quick test, that approach resulted in too much loss of signal in a passive setup. I think it would have to be active for that.

My approach is kind of like a pre-configured matrix mixer with the voltage divider built in.
lmk if I’m missing anything big, and I’ll try to build it this afternoon

update, working version:

Screen Shot 2020-11-20 at 11.51.16 AM2364×904 89.3 KB

RNI-Films-IMG-53F60FCF-D39F-47DB-B85E-E6354890A8FE3640×1380 951 KB

RNI-Films-IMG-3CDF1979-D4D7-4216-956E-9C64CDAF61842471×2470 1.16 MB

RNI-Films-IMG-F94246BA-CC4F-45D7-8538-AA832C0BEDE52608×2607 1.49 MB

I’m going to switch this to 2 dual gang A10k pots like these when I get a chance

I’ll probably also lay this out onto a PCB / kit soon

bmoren · 2020-11-19T19:40:27.682Z

update:

Voltage divider confirmed working as per discussions above.
I put together a little schematic in case anyone else wants to make a version for themselves.

Screen Shot 2020-11-19 at 1.37.13 PM1190×898 7.68 KB

RNI-Films-IMG-397C0416-E7EB-4CFC-9C1F-BE729E7B6E251658×2210 646 KB

I may continue to work on the more complex version from above, but will likely break it out into its own thread in that event.