Using The TI PCM4222 A/D Evaluation Module

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Using The TI PCM4222 A/D Evaluation Module

Post by mediatechnology »

The TI PCM4222 Evaluation board makes a great DIY A/D convertor for Test and Measurement or Recording.

The PCM4222EVM, available for $149 US from Mouser, Digi-Key and the TI store, is a good value despite being a bare bones board.
Having been designed to showcase the PCM4222 the performance is quite good.

PCM4222 Datasheet http://www.ti.com/lit/ds/symlink/pcm4222.pdf
PCM4222 Evaluation Module User's Guide: http://www.ti.com/lit/pdf/sbau124
PCM4222 Evaluation Module at Mouser: http://www.mouser.com/search/ProductDet ... PCM4222EVM
PCM4222 Evaluation Module at Digi-Key: https://www.digikey.com/product-detail/ ... ND/1900031
PCM4222 Evaluation Module at TI Store: https://store.ti.com/PCM4222EVM-PCM4222 ... P1046.aspx

Ross Martin is a big fan of the PCM4222: http://rossmartinaudio.com/index.htm and it gets great reviews on Gearslutz.

The PCM4222 is capable of PCM sample rates up to 192 kHz at 24 bits.

The PCM4222 evaluation module signal path is short having only an OPA1632 differential op amp as modulator driver.
There is no combo mic/instrument/line input PGA to add noise, distortion and gain uncertainty to measurement.

Block Diagram

Image
PCM4222EVM Evaluation Module Block Diagram

Input Section

Image
PCM4222EVM Evaluation Module Input/Modulator Driver

Audio Inputs

The PCM4222EVM left and right inputs are balanced XLR/TRS with 560Ω/leg impedance.
0 dBFS is 4.2 V RMS or about +15 dBu.
Unbalanced 1/4 TS can also be used for single-ended sources.

The PCM4222EVM is supplied without AC-coupled inputs.
C1-C4 are not fitted and shunted by 0Ω resistors (R1, R2, R5 and R6.)

When the inputs are not connected and left floating there is a 1.95 VDC common mode (i.e. relative to ground) potential on the input connector's tip and ring.
If DC-coupled to the source, the source will have to sink about 3.5 mA from each input.
This will force op amp outputs driving the PCM4222EVM into class-A so its not as bad as it may seem.
Just be aware that there is current-limited DC on the inputs.

The differential voltage between inputs however is essentially 0V.
A differentially-connected transformer wired to pins 2 and 3 on this input does not see significant secondary DC current flow.
If the transformer has a center-tap it should be left floating.

If AC-coupling capacitors are installed do note that the the polarity markings on the schematic and the layout are incorrect.
They should be at least 47 µF or greater with the positive side pointing to the OPA1632.
High value resistors (100KΩ) to ground should also be installed on the connector side of the input capacitors to discharge them.

The bill-of-materials specifies 0.1% resistors for the input stage.
A quick check of common mode rejection (tip and ring shorted method) showed about 65dB CMRR.
With matched source impedances feeding the PCM4222EVM the common mode rejection will likely be worse.

For test and measurement applications a higher impedance input stage with attenuator would be helpful.

Digital I/O

The PCM4222EVM has two sets of AES-3 outputs, two AES Coaxial, Serial Audio, DSD and Modulator outputs.

A USB interface requires a third-party board and use of the boards' Serial Audio I2S interface.

AES Interface

I prefer using AES interfaces for test and measurement because the XLR and coaxial outputs are transformer-isolated.
USB is almost never galvanically-isolated and it often introduces spectral spurs at 1 kHz.

Only one AES port is required for stereo.
The second AES port can be used to carry the right channel in dual mono mode.
When the dual mono mode is not being used, the second AES port also carries left and right.

The PCM4222EVM uses the Scientific Conversion dual secondary AES isolation transformer. http://www.scientificonversion.com/dualZo.html
The balanced AES and coaxial AES outputs are electrically isolated.
The PCM4222 in effect provides a 4 output AES distribution amp on its output.

Scientific Conversion has done considerable research on the subject of jitter in the AES interface.
The site has several AES papers on the subject dating from the 1990s.
The 2014 SMPTE paper has some interesting eye pattern comparisons: http://www.scientificonversion.com/SMPTECables9.pdf
For the PCM4222 TI apparently chose the best AES transformer they could find.

Image
PCM4222 Evaluation Module

Powering the PCM4222EVM

Power requirements are +5V at 200 mA and +/-15V at 200 mA.
There are on-board regulators for 3.3V and 4.0V.
This Meanwell 25W desktop "brick" supply is ideal for powering the PCM4222EVM: http://www.mouser.com/search/ProductDet ... 25A14E-R1B
I can't measure a difference between the PCM4222EVM operating from a linear supply or the Meanwell GP25A14E switcher.

Sample Rate and Format Selection

Configuration, including sample rate, is by DIP switches.
This may limit its use when frequent changes in sample rate are required.
In my application I intend to work at 96 kHz with the board internally-clocked.

The PCM4222 can be internally-clocked or externally-clocked at Master Clock frequencies.
For 44.1kHz and multiples MCLK is 22.5792 MHz; for 48kHz and multiples MCLK is 24.567 MHz.
The on-board clocks are specified to have ±50 ppm accuracy.
Jitter isn't specified by TI but the manufacturer publishes 2.5 ps (10 Hz-1 MHz from the output frequency) and 0.6 ps (12 kHz-20 MHz from fo).

AES Physical Outputs
Image

Audio Serial Port
Image

Modulator and Digital Power Connections
Image

SW1 and SW2 DIP switch Settings (96 kHz AES)
Image

Where is RN13?

Note that in the pic below RN13 (below SW3) is missing.
RN13 is absent on both my PCM4222EVM Evaluation boards and I've never seen a picture of one with it installed.
(A possible explanation is that an error in the BOM shows RN13 to have 8 resistors. The layout uses a 4 resistor network.)
I was alerted to this by forum reader Frank Lacy. Thank you for pointing this out Frank.

RN13's absence prevents proper pull-up of the DITFMT and DITMONO lines when asserting them Hi.
In the "Lo" positions shown DITFMT (Left-Justified) and DITMONO (Stereo) the floating lines are pulled down by the switch.
With RN13 out-of-circuit, the PCM4222EVM DITFMT and DITMONO lines may float to an indeterminate state when switched "Hi."
For proper Mono or I2S operation, RN13, a 10KΩ X4 0603, should be installed. (Only two resistors in the network are used.)

SW3 and SW6 DIP Switch Settings (96 kHz AES). SW 4 and SW 5 reset the PCM4222 and the DIT4192.
Image
PCM4222EVM SW3 and SW6 DIP Switch Settings (96 kHZ AES). SW4 and SW5 reset the PCM4222 and DIT4192. Note that RN13 is missing.

The noise performance of the PCM4222 is outstanding.

This is the 96 kHz idle channel noise of the PCM4222EVM with open inputs:
Image
PCM4222 96kHz Idle Channel Open Inputs

With the PCM4222EVM inputs shorted:
Image
PCM4222 96kHz Idle Channel Shorted Inputs

The A-weighted noise of the PCM4222EVM with open inputs is within about a dB of the converter specification.

Image
PCM4222EVM 96kHz Idle Channel Open Input A-Weighted

To measure the distortion of the PCM4222EVM I had to build a prototype oscillator based on Viktors Mickevics design.

See: viewtopic.php?f=6&t=886&start=4

The oscillator is unshielded on a Protoboard with some hum.

Image
PCM4222 96kHz Distortion Using A Prototype Of Viktors Mickevics' Ultra Low Distortion Oscillator

The distortion measures 2 parts per million or -114 dBFS at -21.5 dBFS input.

See Also: "Super Low Distortion Ultra Pure Audio Oscillators Revisted" http://www.proaudiodesignforum.com/foru ... ?f=6&t=887

How To Mount The PCM4222EVM

One of the things you immediately notice about the PCM4222EVM are that there are no mounting holes.
This is a problem if you want to put it in an enclosure.

Examination of the Gerber files shows that the area on the bottom silkscreen that marks the location of the plastic bumpers is a safe area to drill.
PCM4222EVM Gerber files http://www.ti.com/lit/zip/slac510

Fortunately there are no active signal or power lines (except the internal ground plane) near this area.
Image
PCM4222EVM Circuit Board Layout

The added mounting holes will pass through the ground plane in the middle layer.
The light green circles around pads are where pads tie to the internal layer.
The four dark green circles are where they stick the plastic bumpers on.
Image
PCM4222EVM Circuit Board Safe Drilling Area

Drill holes in the center of the circles and then use nylon screws to avoid tying the board's ground to the screw.
The bumpers have some strong adhesive - pushing them sideways (adhesive in shear) seemed to work best.

Image
PCM4222EVM Circuit Board With Added Mounting Holes

It works! Now the board can be mounted. For now I'm using metal 4-40 screws and aluminum hex spacers.
One screw has about 10Ω from the internal ground plane to the spacer so some form of insulation either nylon screw, plastic spacer or a collar washer (from a heatsink mounting kit) might be a good idea.

Once the spacers were attached I wired the power leads to the underside of the PC board.
This is tidier and also allows the top-side terminal block connectors to be used to feed power to other devices.
The DIN power connector mates with the Meanwell desktop supply.

Image
PCM4222EVM With Added Mounting Screws And Standoffs.

All it needs now is a box. (And RN13.)
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Using The TI PCM4222 A/D Evaluation Module

Post by JR. »

Uh-oh digital... :lol:

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Using The TI PCM4222 A/D Evaluation Module

Post by mediatechnology »

Uh-oh digital... :lol:
Oops. I forgot to mention that another application of my PCM4222 is to digitize vinyl. :lol:

Hey John you might be interested to see that I nulled your P10 EQ to Olaf's IIR RIAA. (Post #2) viewtopic.php?f=15&t=885
My first set of tests revealed converter differences and not EQ differences.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Using The TI PCM4222 A/D Evaluation Module

Post by JR. »

Nice...

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Using The TI PCM4222 A/D Evaluation Module

Post by mediatechnology »

I've posted some more measurements of the PCM4222 Evaluation Module in the original post.

To test the distortion I had to prototype Viktors Mickevics Ultra Low Distortion Oscillator.

See: http://www.diyaudio.com/forums/equipmen ... ost3116055

I didn't build the shunt regulators and may modify it to provide a balanced output.
Using the shunt regulators it looks like the output ground is quasi-floating from the supply ground.

Image
Viktors Mickevics Ultra Low Distortion Oscillator 1kHz

The PCM4222 EVM and oscillator combined measure 2 ppm distortion.
harty
Posts: 12
Joined: Mon Nov 27, 2017 12:53 pm

Re: Using The TI PCM4222 A/D Evaluation Module

Post by harty »

hello

I own a PCM4222EVM evaluation board .

I believe from inspection that the XLR digital AES connectors J7 & J9 are mounted incorrectly at the TI factory.

The xlr's appear to be placed too far back. I do not think an adaptor will help as the pcb would get in the way.

Can anyone please confirm this? Is this an error on all versions of this board?

Thank you.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Using The TI PCM4222 A/D Evaluation Module

Post by mediatechnology »

They are too far back.

I think they need to be unsoldered and flying leads extended to panel-mounted male XLRs.

The front panel look to be correct though Left and Right are reversed.
harty
Posts: 12
Joined: Mon Nov 27, 2017 12:53 pm

Re: Using The TI PCM4222 A/D Evaluation Module

Post by harty »

Thanks mediatechnology for the confirmation. yes the analog ones are placed fine and the AES of wrongly placed. i wasnt sure if i was missing a bracket.

I wonder have you used separate external regulators 3.3v and 4v on your pcb? or just used the on board regs?
is it possible to improve sound by doing this or is it just more reliable to use external regulators?

thanks for infos ! : )
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Using The TI PCM4222 A/D Evaluation Module

Post by mediatechnology »

I've always used the on-board 3V3 regulator.

It occurred to me that you could leave the existing AES connectors and tack-solder flying leads to a panel-mounted one. I've never used the second AES port.
harty
Posts: 12
Joined: Mon Nov 27, 2017 12:53 pm

Re: Using The TI PCM4222 A/D Evaluation Module

Post by harty »

thanks for your reply. yes good idea, although i am right out of space in the box. i would need to reshuffle some parts. I hoped it would just slot in as the PDF states "chassis mounted for AES j7 j9"

i am in communication with TI so hopefully they will email something interesting.

i shall eventually find out about the benefits of the external regulators and so shall post that info here, just in case you find it interesting too.
There must be a good reason for TI to go to all the trouble to allow us the option. Hopefully improved sound, lower noise? But I suppose it could just be stability.

kind regards
Post Reply