Newbie Needs Help with Parts Needed for Eden Preamp Build

[outrecording]

Hi everyone,

A couple days ago I started a thread seeing if it's possible to build what I need; located here:

http://groupdiy.com/index.php?topic=55187.0

I've decided to go ahead with the project.

I'm a little stumped as where to begin. I've been reading up and looking at different designs, but once I think I know what components I need, I see a different design with some part I have no clue of its function.

Can I ask for someone to look over the features I need for this preamp and suggest what parts I'll need to make it happen? I would really appreciate the guidance!

Features:

*4 channels (step ganged; one pot to control all 4 channels)
*Portable
*Phantom power (one switch turns on/off for all channels)
*Polarity Reverse (one switch turns on/off for all channels)
*Fuse (is this necessary?)
*Powered by external dc battery pack (don't need mains power)
*On/off switch
*Extremely quiet (it will be used for outdoor ambience)
*Transparent as glass  (sterile sounding is perfectly fine)

*Possibly a pad if it's not too difficult (again, one switch for all channels)

I've decided to go with the Expat Audio Eden preamp for this build. Its size and reported clean sound works perfect for me.

I'm thinking to go with a transformerless design. (For a more colorless sound) But not sure if that's advisable.

With the above in mind, what parts will I need?

I apologize for being a dunce here. All the reading I've done adds just as many questions as it answers. I'm hoping your expert guidance can set me in the right direction.

Thank you

 

[Rochey]

Great News!

Phantom power is easy to control from a single source for 4 channels. Each of the channels will have a 6K8 resistor to the + and - inputs. One switch to pass the 48V to them is easy.

Polarity is a pain to control from one switch. You might consider adding individual toggle switches next to input connectors, or use a relay to do the switching, with one switch on the front panel. IF you use relays, be ready to burn around 100mW per relay when engaged. This means, that if you have no phase reversal, the relays won't consume power. Once you switch the phase reverse on, burn baby burn!

Your phantom power is best driven from a small array of batteries for lowest noise. The amount of current drawn on phantom is tiny. If an array of PP3's is too much, maybe consider a strip of coin cell batteries? If you expect to use the same microphones, maybe work out the minimum phantom voltage? most mic's do not need 48V.

Cheers

/R

 

[outrecording]

Thanks Rochey!

For power, I'm looking at Fab Module's FAB1215 - 12V to +/-15VDC Power Supply Module:

http://www.fabmodules.com/shop/index.php?route=product/product&product_id=51

If I use this along with their Fab1248 +48V converter:

http://www.fabmodules.com/shop/index.php?route=product/product&product_id=50

...and power it with 12V external rechargeable battery, will that work fine with the Edens?

I think for simplicity, I'd like to run everything with just the one power supply. So I can accept a little more noise out of the phantom power. We're not taking a major amount, right?

Thx!

 

[Rochey]

You should be fine.

Put some decent filtering on the 48V rail, you should be fine.

Cool modules, looks like R from Five Fish has been busy!

 

[outrecording]

Ah! Just looked deeper at that power module. Looks like it only has 2 outputs. Is there any way to operate two modules together (for four channels)? Could the 12V be split into the two power modules?

 

[Rochey]

should be plenty of current, just use a splitter.

 

[outrecording]

Excellent! Thank you.

 

[outrecording]

I'm looking at these input and output modules:

http://www.fabmodules.com/docs/pmwiki.php/FABModules/FAB2010
http://www.fabmodules.com/docs/pmwiki.php/FABModules/FAB3010

They'll take care if my pad, phantom power and polarity. Pad and polarity needs 5V. What's the easiest way to get the 5V to these with a 12V to +/-15VDC Power Supply?

Thank you

 

[bruce0]

Just read this thread and it opened my eyes to sound field recording. I was unaware of the technique, although I had used M/S technique which seems related.

I think you will want to put some gain trim multi-turn pots inside somewhere, because having the gain be identical is important according to the mic documentation.  So a linked pot or switch externally but trimmers internally might be appropriate.

Be sure there is adequate HF filtering on 48V supply because DC-DC converters operate at HF, the noise is generally out of the audible range but can have lower order harmonics that are audible (The high frequencies can be easily filtered with a small cheap LC filter).

It seems like the B format uses 3 figure 8's and an omni and if so, I can see why you are concerned about a quiet 48V supply, because some noise is added together when the signal is processed. So a single supply for the 48V is a good idea since the processing subtracts 3 of the channels from the omni, noise on those outputs might/would cancel. The other 3 channels are summing (or subtracting the inverse signal) so looking at total noise on all output channels you get 3X whatever noise is there. 

On the other hand if you had 4 separate 48V supplies the noise would not be coherent and you would get no cancellation and the cumulative output would have double the noise of the single supply.

This cancellation probably doesn't matter much, the noise level is very low, but it is easier and cheaper to use a single 48V supply so why not design it in.

 

[outrecording]

Thank you for your input bruce0.

Yes, ambisonics is quite interesting. It has its limitations, like any other recording method, but the benefits outweigh the limitations IMO - especially if using Harpex-b to decode the b-format signal. Off topic, but Harpex is extremely interesting. (And expensive!) it manages to simulate spaced pair stereo techniques, which is (was?) one of the limitations. I haven't compared the simulated to real spaced pairs, but to my ears they it sounds good.

My mic (sps200) is actually an A-format mic. Four sub-cards and no omni. The A-format signal must be converted to B-format. So I believe a uniform noise across channels would be important when converting. Although as you mentioned, the noise floor would prob be so low that it won't matter.

I found in my tascam dr-680, the phantom power is noisy and uneven across channels. After decoding, I get some uneven sounding noise. Only at high gain though. Hence, my desire to build this preamp for those situations.

I'll definitely look at putting the gain trims in.

Thanks!

 

[bruce0]

Cool... Keep us in the loop, we like pictures!

 

[outrecording]

I'm getting thing sorted out now. Next is figuring out which pot to use...

"Ideally" one monster sized ganged stepped attenuator to cover all 4 channels (balanced) would be great. The only one I could find, by DACT, will take up way too much space.

Any suggestions how I might solve this? (And economically)

I'm considering to use a balanced stereo attenuator and run my four channels unbalanced.

Or admit defeat and go with two stepped attenuators. (Stereo balanced)

I don't want to spend any more than $250 total for this part. Even that's a bit much.

 

[bruce0]

Going with switches or pots, here are some choices as low as $10 or so, and as high as $270

(I don't know much about digital approaches,  but you might want to use some sort of digital control if you know about that stuff, there are some THAT digital control chips and others. )



These are a bit stuff, but they work and the quality is ok, and they are cheap.

http://www.ebay.com/itm/4P-24-Step-Rotary-switch-Attenuator-Volume-Control-DIY-/260693714311?pt=US_Amplifier_Parts_Components&hash=item3cb28e6587

Beautiful goldpoint (elma) attenuators come in 4 gang.

http://www.goldpt.com/prices.html

Quad Revlog pot might be perfect, but may not get you the matching you want.

http://www.audiomaintenance.com/acatalog/potentiometers_carbon_16mm_diameter_triple_and_quad_gang.html

 

[Rochey]

Bruce! Great links!

I might be tempted to throw some small PCB's around those stepped switches with resistors already in place.

 

[bruce0]

Well those goldpoint one (I think you are talking about) feel great when you turn them but they are expensive - I think the 4 wafer one is $270... but it feels great.

The big cheap chinese ones even with 2 wafers turn pretty hard, but they are usable and feel ok (especially with a big knob), with 4 wafers they may be harder, but why not try it it is only 12 bucks and you can switch to the goldpoint ones if you don't like it.


Also....outrecording...

You talk about unbalanced, vs balanced attenuators.  Don't be confused by "balanced" attenuators on ebay (ladders and such).  You are not using these as attenuators, you are using them as gain setting resistors.

The 4 eden preamps would still be balanced.  To my understanding (ask expat) you only need 1 resistor per preamp (4 wafers=4 resistors, setable to 23 positions)  to set gain.

Switch would allow you to be precise, and keep all the preamps the same once they are trimmed the same.  And if there are slight variations in resistance, since you are soldering on the resistors you can adjust them to be accurate as you go with a ohmeter, if one wafer starts measuring a bit high, you can choose the lowest resistance to solder on to that wafer, and so on.  So they will be really close (and  even with 1% resistors you will be a lot closer than 1%).

Just make sure that you do the taper correctly (does it make more or less resistance as you turn clockwise, and what is the taper, log or reverse log).  Many gain setting switches need to be reverse log, again it is not hard to do with a switch, but it is easy to get confused and do it backwards or upside down the curve.


By the way... if you do go with the chinese ones, the blue ones are better than the brown ones.  They turn easier AND they are smaller in diameter.  Other than that they are similar.  In your case, since you are making this portable, you can make the case tall enough to work with those, they are not too big, I think they would fit in 1U pretty easy.

I think you will find the feel acceptable if you use a BIG knob.  With a small knob the torque is pretty high.

 

[outrecording]

Oh yes, how stupid of me. Thank you for pointing that out. At one point before deciding to build this, I had considered using a prebuilt preamp and putting an attenuator between it and my recorder. That's where the idea of needing 8 wafers came from. I see now why I only need 4 wafers. I think I'll give Goldpoint a try. It's about in my budget and should fit in my enclosure.

Thanks again.

 

[bruce0]

I think they have different taper's available, and different total resistances using low noise SMD resistors.  Alternatively you can build your own with resistors.  There may be a big cap in series there.

The gold points (elma's) will feel nice.

 

[Gustav]

most mic's do not need 48V.

/R

I don't know if this is just stating obvious, but as Rochey says, some mics don't even need 48V. If you know you will be using one or two specific models, you might find that you can simplify the supply with no loss of quality.

Gustav

 

[outrecording]

Is there any way to find out how much phantom power a given mic is actually using? Other than contacting the manufacturer (TSL Soundfield never replied to two of my emails on a different subject) is there a DIY approach, perhaps using a DMM?

 

[bruce0]

There is a spec somewhere, I have never seen it.  Typically the power is delivered through 6800ohm resistors (one each on pin 2 and pin3 of the XLR) and the return is through pin1 (the shield). These resistors should match each other very closely to avoid reducing common mode rejection of the input (use a dvm or a wheatstone bridge to pick a pair that match closer than 1%).

Often there are resistors in the mic too.  But the thing to note here is that  even if there is a dead short across the three pins, the maximum current that can flow is 14mA. That is why you see people mentioning 14ma as a spec. for phantom power.

However at 14mA the voltage available to run the mic (as others have pointed out) is 0 (because all the voltage is dropped across the resistors).  And there probably are not any mic circuits that will run on 0 volts.

As you can see, the mic's cannot depend upon a consistent voltage (I think the spec allows 52volts) because the voltage varies depending upon current use.  Many mic's can accept a variety of voltages.  Some mic's have that in the spec (I.E.  operates on 12V to 52V phantom power).  Some will tell you how many mA it draws (ballpark most draw 3ma to 7ma).  Your SPS200 does not have a spec page, just says 48v.  Furthermore your mic connects via a "multipin" connector (actually 10 pins).  They combine the pin 1's of two of the XLRs on two pins and all the other pins 2 and 3 are separate of course.  What this means is that potentially (by combining) your mic could have access to more than 14mA (instead it would have access to 2 sources of 28mA each) but I don't think that is why it is set up that way, the thing just runs 4 capsules right?  So perhaps they are making the choice out of convenience (they could only find a 10 pin connector) or to arrange for some noise rejection of phantom power.

Anyway, if you want to know how much current it uses you could put an ammeter on the pin 9 and 10 connections and double it (or on the corresponding xlr connections and quadruple it.

Why do you care?  Why not just provide 48V phantom through 6800 ohm resistors like everyone else does.  Engineer the supply to provide a maximum of 56mA (or less, because as I say nobody uses that) for your 4 mic connections.  I am sure you would be fine with 40mA and probably be fine with 30mA.

Run the supply at a higher voltage (maybe 52 volts) to allow you to put some LCR filter on the output a drop a few volts.