THAT1570/5171 DIP Adapter - Digitally Controlled Mic Pre

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

promixe

Well-known member
Joined
Jun 9, 2008
Messages
193
Location
Midwest, US
UPDATE: in 2012 this project was (re)named THATDIP adapter that is housed in Lucidity Mic Preamp PCB (the support thread for which is HERE).

This thread here is a mini-diary of the R&D process for what has become Universal Mic Preamp Controller (UMPC) and Lucidity Digitally Controlled Preamps.

Another relevant thread on how Lucidity Preamps and UMPC system shaped themselves into their current forms is HERE.



Hi All,

I've designed this a while ago but just now had time to build a prototype:

innersonix_initial_thatdip_prototype.jpg


Basically, this is a QFN to DIP32 adapter but with integrated bypass caps as close as possible to the package pins. I've also separated the power supply pins of the 1570 and 5171, so one could run the 1570 at +/-18V and 5171 at +/-17V to maximize headroom, etc... If you don't want to do this just connect the same supply to these pins and you're good to go.

For people not familiar with this killer mic pre chipset - check it out here - http://www.thatcorp.com/DigMicPre.shtml

Pinout below has original pin numbers in () as per THAT's datasheet. This is a 15.24mm wide DIP 32-pin package, very easy to pop into any breadboard, etc.

innersonix_thatdip_adapter_pinout_v3.gif


Given the DIY difficulties to work with QFN packages, would people here be interested if I offered this package for sale in the next month or so? It would be the entire assembled adapter as on the picture above:

THAT1570
THAT5171
0805 X7R bypass caps x6
0805 C0G bypass cap x1 (for RFI protection, per individual request could be omitted)
Gold plated pins for DIP socket

Thoughts?
 
Haha, I've been working on this as well!  Now I don't have to finish  ;D

I'm in for 2.

Thanks,

Mike

P.S.: What did you use to get the QFN soldered on?  The soldering looks very well done.
 
Thanks! I have an SMT reflow oven and had a stencil laser cut for this project.

Have you looked into the digital control of 5171? Would you use a rotary encoder or buttons to switch gain settings?
 
Nice!

I would go with rotary encoder.  I got this guy for easy prototyping: http://search.digikey.com/scripts/DkSearch/dksus.dll?lang=en&site=CA&WT.z_homepage_link=hp_go_button&KeyWords=gh6102-nd&x=0&y=0

It also has a push button switch which can be used for all kinds of neat things.

I haven't gone too deep into the digital control portion yet but it looks fairly simple. I've done some pretty complex microcontroller based projects in the past so this should be fairly easy.

Looking forward to brushing up on my mikroC!  ;D 8)

What about yourself? What have you got figured out for the control portion so far?

Cheers,

Mike
 
We should probably have a new thread about this at some point (when a few people have this DIP adapter in hands) =) But...

I'd use this one - http://www.bourns.com/data/global/pdfs/PEC11.pdf
It's over $25 cheaper, takes less space and has the pushbutton integrated as well. I think for this particular application there is no point in going this high-end (i.e. optical AND Grayhill =).

I found that having less detents per revolution is a better feel for gain setting. Also, as far as the quadrature decoding is concerned, there are dedicated chips for this (like this one - http://www.usdigital.com/products/interfaces/ics/lfls7184/) or you can go Atmel's Xmega as it has built in quadrature decoding interrupt-free system. Otherwise there are some limitations with regular uC implementations, although some people might argue. Like having a plain RC filter on signals before entering uC, etc... I think Igor did his Wiregain PGA2500 pre with a rotary encoder, let me see how that's done...
 
I agree about the encoder being more than is required for this.  I got it because it had the header already mounted on and has very distinct detent clicks.  It will probably last forever as well!  I find all that useful for prototyping.  :)

I was planing on letting the uC take care of the decoding (interupt driven probably).  What limitations are you talking about?

Cheers,

Mike
 
Well, in my case there would be limitations if I wanted to use the uC for zero crossing detection, as well as driving signal meters, there would be a lot of cycles used for debouncing, quadrature table look-up. Maybe I'm mistaking but there is a lot of info online where guys are concerned with the whole rotary encoder software-only method. Maybe for front-panel use it's still ok, I need to find out for myself. =)

Anyone else wants the boards? =)
 
promixe said:
Well, in my case there would be limitations if I wanted to use the uC for zero crossing detection, as well as driving signal meters, there would be a lot of cycles used for debouncing, quadrature table look-up. Maybe I'm mistaking but there is a lot of info online where guys are concerned with the whole rotary encoder software-only method. Maybe for front-panel use it's still ok, I need to find out for myself. =)

Anyone else wants the boards? =)

hmmm... sounds like this will be an interesting project (software wise) then.  I remember reading in the datasheet that the 5171 has zero crossing detection integrated which it uses to change gain settings to attain 'click-less' audio.
 
mikefatom said:
I remember reading in the datasheet that the 5171 has zero crossing detection integrated which it uses to change gain settings to attain 'click-less' audio.

Yes, you're correct about the 5171 zero crossing, but I meant a different project of "digitally controlled analog". As far as the quadrature decoding, AVR XMega series have built in decoder that frees the CPU for everything else. Also dsPIC33 series has 1 or 2 QEI's (Quadrature Encoder Interface) for hassle-free encoder implementation. I'm leaning toward the AVR, as I have more experience with these and have an ISP for it. There was a thread on embeddedrelated.com about the best software-only algorithms for Q decoding, if I have time I'll dig it out and post it here.
 
amra said:
I'm for two too but I'm not able to make smt soldering, this come with chip already soldered?

Yes, the adapter would be fully reflow soldered, so you'd only need to pop it into a DIP socket or solder the pins directly onto a PCB.

I'm into micros too and recently I drived my encoder with an LS7184 http://www.lsicsi.com/pdfs/Data_Sheets/LS7183_LS7184.pdf that take care of the quadrature decoding, but there's the LS7183 that can act as standalone up/dwn. I found some on evilbay and works pretty well.

Yes, these chips are very easy to use without much programming involved, but they are kind of expensive (relatively). Most of the time it'd be cheaper to have a dedicated little uC to drive encoders ($1-2 cheaper usually). I'm getting my AVR Xmega board hopefully next week, so I'll see how easy it would be to hook it up to a couple encoders. It's about $4.70 per chip at Mouser/Digikey, so it's cheaper than having a separate uC + Quad Decoder chip...
 
For sure use a cpu will add advantages like the ability to snapshot the current settings in the internal EEPROM and cost less than dedicated chip.
Btw this project is really interesting becaus let me experiment by using different processor approach and you designed a very nice and clever board, I will go for 2 board for sure!
Just for infos, I've purchase several encoders from an evilbay vendor, here's an example http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=290513150286. He has several different ones. Personally I use bourns because I hate the mechanic grid of those type of encoders (tipically 2 bit grey for automotive industry) but in some project I needed something small and I found those cheaper and better than any RS counterpart.
My opinion is use always an interrupt w encoders and with cheap ones a Schmitt trigger to clean signal.
 
amra said:
My opinion is use always an interrupt w encoders and with cheap ones a Schmitt trigger to clean signal.

That's the way to go with most uCs. I've just started to work with Xmega series uC and built interrupt-less encoder routine. Basically it increments or decrements a counter in background with 0% CPU processing and when you need to get the position of the encoder you simply compare the current counter value with the previously stored one. I know it sounds a bit like polling, but it's a lot cooler since you don't actually poll the pins to decode quadrature states but only poll already counted rotations to compare the actual positions of the encoder, past quadrature decoding. Therefore you can do a lot of other stuff in the CPU and once you're done you simply check the counter and see if anything's changed. Once in a couple of mS is good enough for front panel encoder use.

Oh, and it's literally 9 lines of code to make this work. =) Which is the best part I think...
 
I'm intrigued by this chip, but this is the first time I've heard of it.  I get to do a lot of mixed signal and programming at work so this seems right up my alley.  I guess my question is, what would a finished preamp using this chip be capable of?  Obviously you'd place your adapter into a circuit to add Phantom Power, I/O connectors, and some digital interface for gain, but what other capabilities are there?  What are the advantages over the traditional analog preamp designs?
 
millzners said:
What are the advantages over the traditional analog preamp designs?

I'd start with total recall of settings (unless you have a rotary switch with 56 positions =).

Remote control possibility (whether it's from dedicated hardware or software) - put the preamp on stage to minimize mic-to-preamp cable runs, run a line level signal from stage to recorder (100s of feet) and control gain and GPIOs remotely (wireless, ethernet, usart, etc.).

Use the preamp in studio and memorize its settings right in the session file (yet to be done by 99% of digital micpre makers =).

Use the preamp in a console and have a single knob to control gain for all channels in a group for instance. Imagine turning gain down on all 14 drum mics with one knob in a precise way.

Have a mobile preamp unit with more channels than its front panel allows having controls for.

I bet there is more but this is a quick sketch of advantages the way I see it.
 
Other advantages are:
Excellent ein at all gain settings
Switched ladders make the control law very smooth with constant 1dB-per-click.
Most of the transformerless mic pres require a special law gain potentiometer that can be bought only by special order in 5 or 10k quantities. That's why many designers choose rotary switches. Using a standard RevLog pot makes the gain jump up at the end of travel.
 
Hi, This is a really useful adapter. I am definitely interested in at least 3 of them if you get these done,

Good work
 
Any progress on this DIP adaptor?

I am still very interested in this for use in a specific application. (Location recording of acoustic and orchestral groups). I also want to compare this with the head amps in my Yamaha 02R96 (specifically the 1570 preamp chip).

In particular I would like to put together 8 channels of very transparent pre channels with very precise gain switching- which seems to be a good application of this chipset. Even with a limit of 18V power rails - I am sure it will have ample headroom for my purposes

I am also hoping someone else in the forum will be able to help with the software/controller side of this in order to use BCD rotary encoders for the resistor ladder part of this dual chip set (as well as the obvious need to handle momentary switches to toggle the "switch" channels)

I can potentially help in the C code, but am clueless about the electronics side of Micro Controllers as are all the members of my software development department so am hoping someone can assist in the ideas department re controller chips/interrupt controllers etc etc.
 
Hi Delphi,

A couple of updates:

- Yes, I've ordered a new version of the adapter PCBs and should be receiving them by the end of February. The new version is optimized for performance a bit more and has some additional features I've been requested to add. More on that later, when I receive the boards.

- I also designed a PGA2500 adapter as DIP-28 and it's being manufactured right now as well. The adapter has the same exact pinout as the chip itself, but it's quite more than just a SSOP-DIP generic adapter you can get elsewhere. It is made with audio performance in mind and features bypass caps as close as possible to power pins and a split digital/analog ground with an option to stitch them together (which I prefer, and which PGA2500EVM doesn't have).

- Both adapters will be available for pre-orders on my website, which I'm currently building for that purpose. I'm thinking hopefully in a week or so, I'll launch the shopping cart and will post here, so people can start ordering.

- I've been working on a Universal Mic Preamp Controller for a while and am now putting finishing touches on it. It's a very cool chip as it will allow one to control up to 8 channels of EITHER THAT1570/5171 OR PGA2500, all selectable through a configuration menu, within a single chip. All one would need to add to it would be a 2-digit 7-segment display, two 74HC595's, 7 resistors, 2 transistors and a rotary encoder. No more then $5-10 of parts. It will be IS available for sale along with the adapters at the end of February.

- I have made a front panel board for this controller and will be manufacturing it ASAP as well. The board is pretty flexible as it allows for two types of 7-segment displays (0.56" and 0.8" sizes could be used) and two types of buttons (MEC 3F and alike, or Omron B3W-9000 series buttons). It can suit several brands of rotary encoders as well and is just 120 X 36mm. Again, a single board/controller can work with up to 8 channels. You cannot mix PGAs and THATs in one design, but any number of one type up to 8 channels. There will also be a Front Panel Express template for this board so people can make their own front panels to align perfectly with the board.

innersonix_initial_umpc_prototype.jpg


Hopefully all of this will make some DIY people drawn more to these really cool chips, as the two hardest parts (that is reflow-soldering QFNs/SSOPs and UI/SPI-controlling them) will already be done with. =)
 

Latest posts

Back
Top