I'm scheming on making internal mics for my accordion using Simplest P48. Using just a single internal mic would greatly emphasize some notes over others, so a typical design uses multiple mics internally and blends them.

I thought to use parallel capsules in a standard Simplest P48 circuit... some questions, if anyone has thoughts:

1) a standard accordion is about 18-20" (~48cm) wide. That means wide-spaced mics would be about 12" ~= 1ms apart, which comes to comb cancellation with first node at about 500Hz... I wonder how the average accordion pickup system deals with this... simply ignores it, maybe? Judicious spacing of mics to make a less offensive cancellation pattern? Corrective EQ?

2) if e.g. I put multiple 5024's in parallel in a Simplest P48, I don't need to make any other adjustments to the circuit (resistor size, cap size), correct?

3) would using many capsules in parallel, say four or six, alleviate comb filtering by averaging out the cancellations, or make things worse? Simplistic testing simulations shows that the cancellation pattern is somewhat unpredictable, and sometimes ugly, but I wonder how this goes in practice...

Thanks!

-Casey

I'd comment on the comb filtering only.

The conditions for the comb filter to appear are 1) identical signal, and 2) time delay (i. e. sound path difference). The more equal the signals are, the more pronounced the comb filter is. This leads us to the very useful statement: to minimize comb filtering you need to make one of the signals much lower in volume than the other. How much? I'd say 15-20dB. In general, no less than 10 is advisable. There's a rough 3:1 rule that says that the distance between the mics should be at least 3 times more, that the distance from the closest mic to the sound source.

To reduce comb filtering, you can place mics closer to the source. This reduces the mic coverage, but it also reduces the comb filter. So I'd start with the number of mics that I'm planning to put inside the sound source, and figuring out the distances. I'd then decide on the evenness of the coverage and work further from that point - either increasing the amount of mics or moving closer to the source. I'm sure you'll be able to work that out.?

Le 03/06/2022 à 21:37, Casey via groups.io a écrit?:

I'm scheming on making internal mics for my accordion using Simplest P48. Using just a single internal mic would greatly emphasize some notes over others, so a typical design uses multiple mics internally and blends them.

I thought to use parallel capsules in a standard Simplest P48 circuit... some questions, if anyone has thoughts:

1) a standard accordion is about 18-20" (~48cm) wide.

Really? I thought the width of an accordion was significanty variable as you squeeze it. Do you mean height?

I've installed mics in all sorts of accordions; I always put one set on the left and one on the right, each set being one on top and one at the bottom. Except on the small ones, like diatonics and bandoneons.

That means wide-spaced mics would be about 12" ~= 1ms apart, which comes to comb cancellation with first node at about 500Hz... I wonder how the average accordion pickup system deals with this... simply ignores it, maybe? Judicious spacing of mics to make a less offensive cancellation pattern? Corrective EQ?

Think of how a piano is close mic'ed, with one on the left and one on the right, spaced by about 25 inches.

Ah, right, forgot about the 3-to-1 rule/guideline, thanks! I think I might be OK, then... in a multi-mic setup the mics would be pretty close to the instrument so I should be able to beat 3-to-1. -c

Le 03/06/2022 à 21:53, ilya dontsov a écrit?:

I'd comment on the comb filtering only.

The conditions for the comb filter to appear are 1) identical signal, and 2) time delay (i. e. sound path difference). The more equal the signals are, the more pronounced the comb filter is. This leads us to the very useful statement: to minimize comb filtering you need to make one of the signals much lower in volume than the other. How much? I'd say 15-20dB. In general, no less than 10 is advisable. There's a rough 3:1 rule that says that the distance between the mics should be at least 3 times more, that the distance from the closest mic to the sound source.

It just doesn't work, because the mics are not in free field. They are in a confined space, where the act as pressure sensors. Distance does not result in significant attenuation._._,_._,_

Really? I thought the width of an accordion was significanty variable as you squeeze it. Do you mean height?

Yes, sorry, I think from the perspective of a keyboard player :-) -- I meant height.

I've installed mics in all sorts of accordions; I always put one set on the left and one on the right, each set being one on top and one at the bottom. Except on the small ones, like diatonics and bandoneons.

Yeah my original thought was to use two mics arranged as I believe you are suggesting: bass / treble, basically.

If you have any easy suggestions for level control of a simple p48 circuit, I'd love to hear it; I'm currently planning not to bother and just go to two mini-XLR jacks (one for each side), but if there was an easy enough solution it'd be nice to have some volume knobs and a single XLR out.

Think of how a piano is close mic'ed, with one on the left and one on the right, spaced by about 25 inches.

Yeah I don't know why I'm making this harder than it needs to be. Upon reflection it seems clear that it'll work fine. I think I got hesitant when I contemplated many capsules (some pro systems use 4 to 7 capsules on a single side) and the ramifications of all those mics.

Thanks,

-c

It just doesn't work, because the mics are not in free field. They are in a confined space, where the act as pressure sensors. Distance does not result in significant attenuation._._,_._,_

Ah, so maybe I should stick with 2 capsules per side? I wonder how the six-capsule pro solutions deal with it, then... maybe they don't worry about it...

if you have two mics inside the front of an accordion consider:

* each reed is a different distance from each mic, and so the intermic delay for each reed will be different resulting in a different comb filter for each reed.

* the sound will be bouncing around inside the case all over the place and mixing acoustically.

* the reeds closest to each mic are furthest from the other mic, so the reeds with the largest time difference also have the largest level difference and the reeds in the middle are more even.

So the short version is don't worry about the comb filtering, I don't think you'll notice it amongst everything else that's going on, and there's not that much you can do about it anyway.

It's really not that different from any other stereo pair in that sense. It is however different in that it is in an enclosed space with it's own internal air resonances.

Do, however, expect to experiment with mic position to get the best tonal balance and bass/treble balance.

You don't say if it is a piano accordion or B/C system or some other type, but reed layouts may vary, so consider the reed layout.

I would think two mics on the keyboard and one or two on the bass would be sufficient.

There are better circuits than P48 that aren't really much more effort if you are going to the trouble of installing multiple mics in an accordion.

If you draw the circuit you intend we can comment better on the mixing - seems to me in a mixer each capsule needs its own resistor into the summing amp, just shorting together the capsule fet outputs seems wrong to me.

开云体育

Thanks thet --

You don't say if it is a piano accordion or B/C system or some other type, but reed layouts may vary, so consider the reed layout.

It's a piano accordion, so the reeds are linearly arranged. Doing some tests (with a single Primo omni Simplest P48 inside the right-hand side) I do hear a strong difference for the reeds near the mic, as expected.

I would think two mics on the keyboard and one or two on the bass would be sufficient.

Glad to hear that this and Jerry's experience accord with each other.

There are better circuits than P48 that aren't really much more effort if you are going to the trouble of installing multiple mics in an accordion.

Interesting... what did you have in mind? I have made so many S-P48's and they've been great, and given how simple and effective they are, and that I don't need low noise performance in this application, Simplest P48 seemed the natural choice?

If you draw the circuit you intend we can comment better on the mixing - seems to me in a mixer each capsule needs its own resistor into the summing amp, just shorting together the capsule fet outputs seems wrong to me.

Yeah the plan was to literally parallel multiple capsules on a standard P48 circuit, so, .

IIRC it has been discussed a couple times here that this works (and gives a little less noise) so it seemed like the easy way to deal with mixing the signals from the capsules without having to engineer some kind of active mixer as well... but I have no idea if there are subtle consequences, so let me know if I'm missing something.

-c

开云体育

I agree with thet- don’t overthink it. There’s so much mixing and reflections and resonances that you can get away with quite a lot.?

开云体育

Hmm I'm no expert on SimpleP48.

AFAICT they don't work right with transformer input mic preamps, which I mostly use, so i haven't investigated. Plenty of others here would know better what you can do with them.

Is there a full circuit diagram somewhere with an actual explanation of how they work? All the docs I can see seem to omit the fet from the diagram, and are vague about how it actually works. Capsule datasheets are also vague on internal connections, especially how (if?) the gate is referenced.

Paralleling them is using parallel fets though - I would think the matching of the fet characteristics would matter.

It just doesn't seem that hard to wire each capsule as a source follower and then mix them to a buffer amp that can give a low Z balanced output - the extra circuitry looks like a lot compared to a S-P48 but its trivial compared to the effort of fitting a pickup system into an accordion. It would allow the addition of a balance or level pot easily enough as well.

You are right that noise performance is not a big issue though - the SPL inside an accordion is pretty high.

Is there a full circuit diagram somewhere with an actual explanation of how they work? All the docs I can see seem to omit the fet from the diagram, and are vague about how it actually works. Capsule datasheets are also vague on internal connections, especially how (if?) the gate is referenced.

I can't speak to the technical side, but once logged into the groups.io website you could check Ricardo's PDF in the files section for some more info; this link might work: /g/MicBuilders/files/Ricardo/SimpleP48/SimpleP48.pdf

It just doesn't seem that hard to wire each capsule as a source follower and then mix them to a buffer amp that can give a low Z balanced output - the extra circuitry looks like a lot compared to a S-P48 but its trivial compared to the effort of fitting a pickup system into an accordion. It would allow the addition of a balance or level pot easily enough as well.

I'm sure you're right; I've built a few circuits from schematics here in the group (with lots of help from members here) and I'm not generally intimidated by that (even made my own PCB's for the DIYGenericIntFET)... but I'm basically just following instructions; to build the simple mixer circuit you describe I would basically have to pester the list until it collectively designed the entire circuit for me, and I wouldn't feel good about that. So if there's some already-drawn-up way to do what you describe, that would be awesome... I would love level controls for the two sides... but my knowledge is not at a level where I can design it myself. (Also it feels a little off-topic for the list to design mixers, so I'm trying to be sensitive to that.)

-c

开云体育

Paralleling them is using parallel fets though - I would think the matching of the fet characteristics would matter.

Actually, the FET characteristics do not matter too much, as long as the voltage across them is enough to guarantee they are in their pentode region, but the sensitivity, being usually spec'd with a +/-3dB tolerance may result in too much difference. Unless using matched capsules, the system must allow for some kind of individual adjustment, which is not compatible with the simple P48.

Considering the modest price of electret capsules, it should not be too difficult to buy a batch of 10 and match them using a simple jig.

On 04/06/2022 00:58, Casey via groups.io wrote:

you could check Ricardo's PDF in the files section for some more info; this link might work: /g/MicBuilders/files/Ricardo/SimpleP48/SimpleP48.pdf

Yes I just read through that. It doesn't explain the circuit straight up, but one can glean quite a lot about how it works from reading that. TBH the floating balanced operation sort of does my head in.

It seems that it *will* work with transformer inputs, though the more subtle consequences of that still give me pause.

. but I'm basically just following instructions; to build the simple mixer circuit you describe I would basically have to pester the list until it collectively designed the entire circuit for me, and I wouldn't feel good about that. So if there's some already-drawn-up way to do what you describe, that would be awesome... I would love level controls for the two sides... but my knowledge is not at a level where I can design it myself. (Also it feels a little off-topic for the list to design mixers, so I'm trying to be sensitive to that.)

Well this is something I'm also interested in. I want a compact way to mix 2 pickups for a fiddle with two hiZ piezo film pickups - so the same sort of mixer might well work for both. An accordion has a luxurious amount of space for mounting circuitry compared to a fiddle so it should be easy.

I know I could design something that would work - I'm not sure if I would think of every subtlety to completely optimise it.

Maybe you should consider NOT using omnidirectional capsules...
Omnidirectional microphones usually have their back chambers sealed. Your average omnidirectional capsule may have trouble dealing with the constantly changing pressure, in fact alternating between pressure and vacuum inside the accordion...

Thanks -- are you thinking that there might be some near-DC signal fluctuation from these pressure changes? At any rate, the capsules will be on the atmospheric-pressure side of the reeds, not inside the bellows, so there shouldn't be any significant pressure swings of that nature, AFAIK.

But I do wonder about cardioid capsules just for the sake of preventing on-stage feedback...

-c

Le 04/06/2022 à 03:36, Casey via groups.io a écrit?:

Thanks -- are you thinking that there might be some near-DC signal fluctuation from these pressure changes?

Eletret capsule do not pass absolute pressure, only its variations.

But I do wonder about cardioid capsules just for the sake of preventing on-stage feedback...

When put inside a cavity, mics become omni regarding what's outside of the cavity. Feedback resistance is governed by distance. The capsules are very close to the reeds and inside the pressure field.

I have noted that the difference between omni and cardio capsules inside is just that the latter have less bass. Feedback resistance is the same.

It would be different if the mics were externally mounted.

Maybe you should consider NOT using omnidirectional capsules...
Omnidirectional microphones usually have their back chambers sealed. Your average omnidirectional capsule may have trouble dealing with the constantly changing pressure, in fact alternating between pressure and vacuum inside the accordion...

Omni capsules always have a pressure leak, in order to prevent the possible effects of change in ambient pressure.

I agree stick to the omni capsules for the reasons Jerry gave.

As far as designing a mixing preamp system, it might help to have specs on the specific fet in the specific capsule that you want to use. You mentioned 5024s - does anyone have specs on the fet?

(I really wish you could get little electret omni caps without fets, it would make designing good preamps a lot easier to use known good parts)

Can the trace be easily cut to turn it into a 3 wire capsule with the source separated from ground?

If I were starting from scratch I'd probably be tempted to deign around a primo EM273

I have some 10mm omnis that are supposed to be good but I can't remember what they are, they look like pics of 5024s though.

ie here:

and here:

Are there different 5024s with different specs?

For an accordion mic is capsule balance a set and forget thing? ie could it be a trimpot on the board, and with limited range, just to make minor adjustments to almost balanced levels?

Or does it need to be a full on 100% balance pot that is accessible to the player during a gig?

It may be useful to have separate outputs for left and right hands. The final balance can be made at the mixing desk.

As for a mic array. indeed, since the mics are inside, it should be possible to get away with 2-3 mics per side.