Newbie Needs Help with Parts Needed for Eden Preamp Build

[ricardo]

Next is figuring out which pot to use
....
Any suggestions how I might solve this? (And economically)
....
Or admit defeat and go with two stepped attenuators. (Stereo balanced)

DEFINITELY use a switch rather than a pot.  You want better than 0.1dB matching between the 4 channels.

It's not so important with the SPS200 but you'll need it for a better A-format mike like TetraMic.

For less cost without compromising quality, use 2 stereo ganged switches or even 4 single switches.

 

[bruce0]

Use the rotary switch. 

Unless you are talking about trimmer pots, in which case any old multiturn trimmer will work. 

You have to adjust this on the bench anyway, you won't have any way to test the amplification in the field (a signal generator and a DAW or Oscilloscope. or DVM).  So just put multiturn trimmers on there and make them cover 5% of the range of the Switch or so.  I assume you will trim at typical mic gains of 40dB, and let the higher and lower ranges fall where they may (this depends upon the typical gain range you work in with this mic).

So you want something that covers say 50% of the range of that "typical" gain level's resistor, so you can push it around accurately.  Or just ball park it at 5%-10% of the total combined resistance value.

And you probably want to have the resistance values set up so that with the trimmer at middle position, the amp is properly adjusted (I.E the resistors are in series, make the first resistor soldered to the rotary switch or board smaller by half the value of the pot). 

 

[ricardo]

Unless you are talking about trimmer pots, in which case any old multiturn trimmer will work.

You definitely DON'T want trimmers on an Ambisonic mike preamp.

Your preamp needs 0.1dB matched gain AS DESIGNED.  Just use 1% resistors on your gain switch.

 

[outrecording]

Sorry for the lack of response from me the past couple days. I ran into a problem with either my recorder or SPS200. I'm getting a loud dose of noise on CH4 of the mic, regardless which input I use on the recorder.

I'm praying it's the recorder's problem...

I've plugged in four different mics of the same model and they work fine in the recorder. But the way the SPS200 is wired gives me some small hope that individual mics might not make this noise.

Anyhow, I know a guy with the same recorder, and will try it on his and cross my fingers.

Preamp on temp hold until I get this sorted. Crap!

 

[ricardo]

I'm getting a loud dose of noise on CH4 of the mic, regardless which input I use on the recorder.

This is a known problem with early SPS200.  If it's under guarantee, send it back for repair/replacement.

 

[outrecording]

Ah, this is "good" news. It's numbered in the upper 300's. Hopefully that's part of the early batches. Will give a call to TSL and see what they will do. Thanks for that info!

 

[bruce0]

Unless you are talking about trimmer pots, in which case any old multiturn trimmer will work.

You definitely DON'T want trimmers on an Ambisonic mike preamp.

Your preamp needs 0.1dB matched gain AS DESIGNED.  Just use 1% resistors on your gain switch.

I thought the idea would be to have trimmers internally to establish the gain as exactly the same.  This is recommended on the soundfield site (they describe a process to precisely trim individual preamp amps to the same gain).

The notion would be to have a master gain switch with feedback resistors for each amp, and then to trim them all to be identical (feed in a signal, measure the output) at a particular gain.  Any error would then be the difference in the resistors used for the gain switch.

Have I got this wrong Ricardo?  ( I am not talking about trimming them during recording, but just to make them identical  during a calibration step.  I suppose you might be able to get opamps to behave within 0.1% using 1% resistors... (it should take 10 tries <grin> but the values tend to cluster so maybe more or less), but a trimmer might be the way to go so you could calibrate on the bench).

 

[ricardo]

This is recommended on the soundfield site (they describe a process to precisely trim individual preamp amps to the same gain).

Have you got a link to this?  I can't seem to find it on their website.

The notion would be to have a master gain switch with feedback resistors for each amp, and then to trim them all to be identical (feed in a signal, measure the output) at a particular gain.  Any error would then be the difference in the resistors used for the gain switch.

Have I got this wrong Ricardo?  ( I am not talking about trimming them during recording, but just to make them identical  during a calibration step.)

If a 4 channel preamp doesn't have within 0.1dB matched switched gain using just bog standard 1% MF resistors ... something is seriously wrong.

Eden should do this easily.

 

[bruce0]

Ok..
I am being thick

But if gain in the preamp chip is 1+ ((Ra+Rb)/Rg) then if Ra and Rb are 1% high and Rg is 1% low then at a target gain of 30x V you get 29.67 vs 30.6 which effect gets more extreme at larger gains.

This is a .25dB difference and a 3% voltage difference and the effect gets worse at larger gains ( I don't know how hot these mics are but figure 40 Db typical?)

Anyway I would put in a couple if cheap trimmers if the don't exist on the Eden , or hand match resistors maybe

 

[bruce0]

This is recommended on the soundfield site (they describe a process to precisely trim individual preamp amps to the same gain).

Have you got a link to this?  I can't seem to find it on their website.

In this manual under "interfacing the sps200 section 1"
http://www.soundfield.com/downloads/sps200-manual.pdf
http://www.soundfield.com/downloads/sps200-manual.pdf

 

[ricardo]

This is a .25dB difference and a 3% voltage difference and the effect gets worse at larger gains ( I don't know how hot these mics are but figure 40 Db typical?)

Why don't you try it and see?  You need to be able to measure gain consistently to better than 0.1dB but it's not so difficult these days.

If you use 1% MF resistors from a reputable maker and from the same batch, you will find the 2 sigma distribution of values is easily 0.1% though the mean may not be within 0.1% of the nominal value.

If you use mixed batches from dodgy sources on eBay, you will have other problems ... so just ignore everything I say. 
____________

Thanks for the link to the sps200 manual.

Do you mean the bit on Page 8

1.  It is essential that the gains ...

This is for a device with a twiddle knob rather than precise switched gain.  I STRONGLY RECOMMEND YOU DON'T USE SUCH A DEVICE.

MOTU Traveler, Prism Orpheus, most RME, Metric Halo ULN8 are good preamps for Ambisonic recording.

Tascam DR-680 is the recommended portable device for TetraMic.

I'm surprised SPS200 reacts badly to DR-680 P48V noise.  Can you see if you can hear any noise with just 100R across pins 2 & 3 of each Mike I/P of the DR-680 with P48V?  That will tell you if its the Tascam or the SPS200.

 

[bruce0]

ricardo:

Apologies, I don't mean to oppose your guidance which is sure to be based on more experience that I have with ambisonics.

As I said earlier; this thread is my first exposure to the concepts of ambisonic recording.  I don't know the OP's preamp design except that he shows a picture of a Soundfield product as a reference and plans to use 4 EDEN preamps and wants ganged gain.

I acknowledge your point that using accurate components, and given a well designed preamp you could get the accuracy ambisonic (apparently) needs.

• I don't know what the Eden preamp adds in terms of variability and since the schematic appears to be unpublished, and they come pre-assembled except for certain (all?) gain resistors.
• Without a schematic and component tolerances I don't know how judge what differences there might be between individual assembled Edens.
• Also I know from making 23 step 2dB decades that finding resistors for all the positions in that stack for some values can expensive, and so I would end up with "small sets" that can have variability.
• I don't know what test equipment or components the OP has at his disposal
• OP did state he was on a budget

    .

My recommendation was intended to help the OP, a self proclaimed Newbie, and I suggested a 4 wafer 24 throw Chinese rotary switch with a decade scale of resistors as a ganged gain control (switch not twiddled knobs).  I also suggested a multi turn trimmer in series with each wafer to calibrate out any differences of unknown origin between the amps.

I acknowledge that to your point, the gain pot may be un-necessary especially if the builder could control preamp design, construction, and component tolerances.

I can't try it out because I am not building one, OP is.  I am not building one, just trying to help out. 

If I did build one I would put the gain trimmer onto the design and jumper across it if it was redundant.

Regarding the Soundfield document reference: Yes, you are correct it is on page 8

I am intrigued by the whole Ambisonic technique, in particular the ability to post process virtual mic placement direction after the fact and want to learn more.  I was going to look into seeing if I could get or build or cobble together a mic and use equipment I have in place to try the technique (among a zillion other projects).

I am sure the OP welcomes your suggestions, and I will withdraw from this conversation because I really am over my head with ambisonics and the issues that surround them.  I meant no offense.

bb

 

[ricardo]

I am intrigued by the whole Ambisonic technique, in particular the ability to post process virtual mic placement direction after the fact and want to learn more.  I was going to look into seeing if I could get or build or cobble together a mic and use equipment I have in place to try the technique (among a zillion other projects).

There is a treasure trove of very high quality uncompressed Ambisonic recordings at http://ambisonia.com/.  Look in particular at recordings by ajh, soundmanjohn, Paul Hodges & JHRoy.

Navigation is a bit wonky so some of the best examples are also at http://ambisonic.info/audio.html.

Download VVMic http://vvaudio.com/products/vvmic, get a pair of headphones and hear what an Ambisonic mike can do in stereo.  Harpex has fancy features which are convincing with some sources and not with others.

PS  Most of the recordings were made with mikes in which I had a big hand in design.
_____________

Eden is based on TI INA163, perhaps the best of the single chip mike preamps.  It has precision resistors on chip.  It 'only' needs a gain setting switch.  The gain switch is a VERY sensitive point and needs to be AS SMALL AS POSSIBLE.  The Soundfield preamp just uses 1% resistors .. as have all previous Soundfields.

 

[bruce0]

Thanks for the Ambisonic links.

Can you tell my why the gain switch needs to be small, is it just noise and crosstalk potential of large loops of wire, or something else?

Could't try the VVmic app, I am Linux and Mac only here.

Any suggestions on DIY Ambisonic mic's?  (With my current collection I would have to assemble one with two AKG 414's, a ribbon and an omni and it would obviously be BIG). 

For instance could I wire one up with 6 horizontally opposed tiny cardiod capsules, and an omni?

bb

I am intrigued by the whole Ambisonic technique, in particular the ability to post process virtual mic placement direction after the fact and want to learn more.  I was going to look into seeing if I could get or build or cobble together a mic and use equipment I have in place to try the technique (among a zillion other projects).

There is a treasure trove of very high quality uncompressed Ambisonic recordings at http://ambisonia.com/.  Look in particular at recordings by ajh, soundmanjohn, Paul Hodges & JHRoy.

Navigation is a bit wonky so some of the best examples are also at http://ambisonic.info/audio.html.

Download VVMic http://vvaudio.com/products/vvmic, get a pair of headphones and hear what an Ambisonic mike can do in stereo.  Harpex has fancy features which are convincing with some sources and not with others.

PS  Most of the recordings were made with mikes in which I had a big hand in design.
_____________

Eden is based on TI INA163, perhaps the best of the single chip mike preamps.  It has precision resistors on chip.  It 'only' needs a gain setting switch.  The gain switch is a VERY sensitive point and needs to be AS SMALL AS POSSIBLE.  The Soundfield preamp just uses 1% resistors .. as have all previous Soundfields.

 

[ricardo]

Can you tell my why the gain switch needs to be small, is it just noise and crosstalk potential of large loops of wire, or something else?

The input current flows in a critical INPUT LOOP.  In FIGURE 5 of http://www.ti.com/lit/ds/sbos177d/sbos177d.pdf it goes from XLR p2 to C2 to p5 of the ina163 to p12 to R7 to R6 to p3 to p4 to C1 to XLR p3.

The area within this loop MUST BE AS SMALL AS POSSIBLE.  Think of it like a loop aerial picking up crud cos that's exactly what it is.  The size of this loop is FAR more important than the amount of mu-metal surrounding the amp.  You can immediately tell if a maker understands Low Noise design by looking at how he lays out these critical INPUT LOOPS on his PCBs.

Anything you connect to the loop like R1, 2, 4, 5 & the diodes may be part of this loop too depending on exact values, layout bla bla.  The switch and resistors are definitely part of the loop.  If you can add trimmers without increasing the area within the loop you won't add additional problems 

I pontificate on this in LNprimer https://groups.yahoo.com/neo/groups/micbuilders/files/Ricardo/.  You have to join.

Could't try the VVmic app, I am Linux and Mac only here.

ambisonic.info has links to TetraProc & TetraCal http://ambisonic.info/tetramic/software.html These are excellent software packages.

Any suggestions on DIY Ambisonic mic's?  (With my current collection I would have to assemble one with two AKG 414's, a ribbon and an omni and it would obviously be BIG). 

For instance could I wire one up with 6 horizontally opposed tiny cardiod capsules, and an omni?

ambisonic.info also describes a Native B-format mike http://ambisonic.info/practical/recording.html.  See also Eric Benjamin & Thomas Chen's http://www.aes.org/e-lib/browse.cfm?elib=13348 & http://www.aes.org/e-lib/browse.cfm?elib=13444  I have to declare a connection with all these projects as well as some work on the Nimbus Halliday mike too.  Use the 2 x 414s for horizontal surround first before trying to do vertical sound.

These would be better than using tiny cardioid capsules.  I've played with a Native B-format mike using 2 x AKG414s and a B&K omni in the last Millenium.

Those who have done a lot of work on Native B-format mikes invariable end up with TetraMic but you'll have a lot of fun along the way

The main problem is matching the frequency responses of the different mikes.  Have you got a good method to measure mike frequency response?

 

[outrecording]

Well, nearly a week since I contacted KMR (where I purchased) and several days since I contacted Soundfield and nothing back. Great customer service on both ends. Time for an international phone call it seems...

Tascam DR-680 is the recommended portable device for TetraMic.

I'm surprised SPS200 reacts badly to DR-680 P48V noise.  Can you see if you can hear any noise with just 100R across pins 2 & 3 of each Mike I/P of the DR-680 with P48V?  That will tell you if its the Tascam or the SPS200.

It has been working fine since October. Strange that it just now started doing this.

I'll give the 100R a try. Thank you for the suggestion.

 

[outrecording]

I am sure the OP welcomes your suggestions, and I will withdraw from this conversation because I really am over my head with ambisonics and the issues that surround them.  I meant no offense.

bb

I thank everyone for their input. I'm somewhat flying blind on the project as I'm coming up to speed on preamp building. So any help you can offer is a great help. Please do stay in the conversation. Thanks!

 

[outrecording]

For the preamp, the input and output modules from Fab Modules I'm considering have pad and polarity operated by a switch that takes 5V. Can I split the 12V input (this would be the third split) and send it to something like this:

http://m.ebay.com/itm?itemId=121208203049

There would be 4 input and 4 output modules. Would one buck converter work to switch all 8 modules?

 

[ricardo]

For the preamp, the input and output modules from Fab Modules I'm considering have pad and polarity operated by a switch that takes 5V.

Not sure why you need pad & polarity.

This is a A-format mike.  You won't be mixing it with anything else until post processing when polarity is easily changed in your DAW.

If you are doing mostly Nature recording, you will be struggling for level rather than thinking of pads.

 

[outrecording]

I recorded some fireworks on my rooftop on New Years. They were popping off real close. Low gain wasn't clipping but close. I thought it could be a handy feature.

And for polarity. The mic is reversing polarity for whatever reason. Doing it in the DAW is fine, but takes ages when you have 4 tracks of hundreds of minutes at 24bit 96k. And I've forgotten to do this several times. So I'd just assume take care of it at the source.