Mid Reference Capacitor

[manureichel]

Hi. I have a question regarding the Mid Reference pin. (Third pin on the IC). I'm currently working on a project that's going to be pretty small and want to use smaller dimension parts as possible. The Datasheet states to put a 1uF polarized capacitor between the MID pin and GND. Although I've seen other schematics where bigger values are used. Can I use a ceramic multilayer capacitor on this one?. And just for reference, what's the purpose of this Mid Reference Pin? My guessing it's that is there to apply a capacitor for internal Vref stability or something like that?

Greetings!

[frank]

MLCC is fine, pin is part of the reference for the internal CODECs.

[Aion]

I would be careful with MLCC as most of them, except C0G (NP0), are microphonic. A problem if you design an effect to be mounted in combo amp.
There are solid tantalum capacitors in 0603 case, like TMCJ1A475MTRF from Vishay/Sprague, offered by Mouser.

[smear]

oh that's great to know! any excuse to avoid an electrolytic is a nice help.

i doubt the microphonics would matter, since it's unrelated to the signal path and connected to ground. we all use ceramic caps for the DVD+AVD pins and for the crystal, but that doesn't cause any trouble.

[Aion]

Frank explained that "pin is part of the reference for the internal CODECs". I fear that modulation of this pin voltage can have an effect on conversion results of ADCs and DACs. Let's wait for Frank's opinion.

[frank]

Frank explained that "pin is part of the reference for the internal CODECs". I fear that modulation of this pin voltage can have an effect on conversion results of ADCs and DACs. Let's wait for Frank's opinion.

Well my opinion is Aion may be right and I had not thought of the microphonic issue so I say give it a try but listen for any issue. It does set a reference for the converters so it is part of the signal path but it is also internally filtered so it may be OK. If not then a C0G as Aion suggests.

[Aion]

High capacitance C0G capacitors in small (0805) cases are actually available only up to 47 nF and relatively expensive.
I would suggest to use tantalum capacitors instead, like TMCJ1A475MTRF in 0603 case from Vishay/Sprague, offered by e.g. Mouser.

[smear]

oh i see! tantalum does seem safest, as is used on the dev board. though personally i'm going to try some standard standard ceramics on my next build to find out for sure, like frank suggests

also, you can get 1uf C0G caps for the low price of... $15 each https://www.mouser.com/ProductDetail/Kn ... 5HJXqk0%3D

by the way, aion, why do you recommend a 4.7uf cap? more filtering for a more stable reference? i was thinking to use the same 1uf part for both the MID pin and the REFP pin.

[Aion]

You can find even 22 uF in some commercial products. The 4.7 uF tantalum is available in 0603 case.

[smear]

interesting! do they use the same larger values for REFP?

speaking of commercial products, i did some poking around and zcat pedals use MLCCs for both MID and REFP: http://www.effectsdatabase.com/model/zcat/polyoctaver

[ice-nine]

You can find even 22 uF in some commercial products. The 4.7 uF tantalum is available in 0603 case.

All my designs use 10uf for mid reference but I use micro sized electros or tants.

[Aion]

@ smear "do they use the same larger values for REFP?"

Yes, 22 uF both for MID and REFP.

[Sweetalk]

@ smear "do they use the same larger values for REFP?"

Yes, 22 uF both for MID and REFP.

Any improvements using larger caps?

[ice-nine]

it is just dataheet stuff, the main datasheet suggests 1uf for both but the dev board dataheet has 10uf it doesn't make much difference. I use 10uf in both positions.

[Aion]

"Any improvements using larger caps?"

I don't think so. I think that manufacturers often simplify their logistics and stock many e.g. capacitors of the same value. Another example: there are sometimes strange values of resistors on commercial schematics. I suspect that banana counters press the designers to use stocked components.