ProtoDAC TDA1387 X8 project

[eculcea]

I am just a hobbyst as well. I recently added the test method in Post #1 of this thread.

Make a simple module test rig with one of your spare PCBs with GPIO soldered, three I2S resistors (R1-R3), a socket for the module (14 pin DIP socket OK), two metal film 430R I/V resistors (U7/U8) and 2000uF power supply decoupling cap (C3/C4). There is no need for any other parts on the test rig. Download this 60s_Sine_50Hz_0dB.wav file and play on Moode with repeat track on. Measure AC RMS voltage at the two test points, which are the output coupling cap pads at the I2S resistor end of the PCB. Measure between test point and ground. Voltage should read 1.2-1.3 Vrms, and the two channels should read very closely. I reject modules for the kit that can't pass this test. I do think modules that read 1.10-1.19 are good, but I only include 1.2-1.3 Vrms modules with the kit.

I hot swap modules and wear a grounding strap.

Most DMMs are accurate at 50 Hz, and accuracy suffers as the frequency increases.

I apologize, I missed your first post. I have the module installed in a socket and I "tested" it by using my ears. After reading your post of 10% failure rate of modules it got me thinking, my ears might not be good enough to find a bad module.

[eculcea]

here is what I came up for using the 100R tantalum SMD resistors. One is the front and the other one is the back of the SMD adapter. I got 4 in serial for a total value measured at 400.1R

   

[hifinet]

That is a very nice layout. I wasn't sure if the 1206 parts would fit the pads on that particular board. Also, the 0.1% tolerance of the parts seems pretty accurate.

[trebtek]

Hello everyone, this is my first post on this forum. Glad to be with you.

A year ago, I built a ProtoDAC:
    Kit sold by Mark on EBay
    + RPI-4B
    + iFi Power supply
    + Roon Bridge renderer
    + No DSP, No Camilla or anything

I am very very pleased with the results, sound quality is great for the money, so a big Thanks to Mark and everyone involved in this great DIY project.

But last month, I started noticing some occasional crispy noise, that seemed to be random, not related to the level of the music being played, and often present at the beginning of a new song - but not always. I was able to isolate the root cause to the ProtoDAC itself after successively playing music without Roon, replacing the microSD card, and replacing the RPI. (I also checked and retouched a few solders without any audible effect.)

Only then did I search this Forum, to find out that others have apparently experienced the same problem, and that odds are the TDA1387x8 is misbehaving. So I ordered 2 new modules from AliExpress... awaiting delivery.

In the mean time, I decided to run some tests on the ProtoDAC, and share the results with you.

I connected an old RPI to the ProtoDAC via GPIO, installed RaspiOS on the RPI and used the ALSA speaker-test utility to generate a 400 Hz sine wave to the I2S interface. I then connected the RCA of the ProtoDAC to the input of my Behringer DAC/ADC and collected the signal into REW wich allowed me to use the scope and capture the following screenshots.

It looks like after a second or so of normal behaviour, some distortion appears, and then the P-P voltage strangely rises, with more distortion being apparent on close-up, and then after some time, PP voltage comes back to its initial value and distortion gradually disappears.

That’s it. I thought I’d share this with you guys. I’m sure some of you are much more competent than me to perhaps comment on this.

               
               

[hifinet]

Hello everyone, this is my first post on this forum. Glad to be with you.

A year ago, I built a ProtoDAC:
    Kit sold by Mark on EBay
    + RPI-4B
    + iFi Power supply
    + Roon Bridge renderer
    + No DSP, No Camilla or anything

I am very very pleased with the results, sound quality is great for the money, so a big Thanks to Mark and everyone involved in this great DIY project.

But last month, I started noticing some occasional crispy noise, that seemed to be random, not related to the level of the music being played, and often present at the beginning of a new song - but not always. I was able to isolate the root cause to the ProtoDAC itself after successively playing music without Roon, replacing the microSD card, and replacing the RPI. (I also checked and retouched a few solders without any audible effect.)

Only then did I search this Forum, to find out that others have apparently experienced the same problem, and that odds are the TDA1387x8 is misbehaving. So I ordered 2 new modules from AliExpress... awaiting delivery.

In the mean time, I decided to run some tests on the ProtoDAC, and share the results with you.

I connected an old RPI to the ProtoDAC via GPIO, installed RaspiOS on the RPI and used the ALSA speaker-test utility to generate a 400 Hz sine wave to the I2S interface. I then connected the RCA of the ProtoDAC to the input of my Behringer DAC/ADC and collected the signal into REW wich allowed me to use the scope and capture the following screenshots.

It looks like after a second or so of normal behaviour, some distortion appears, and then the P-P voltage strangely rises, with more distortion being apparent on close-up, and then after some time, PP voltage comes back to its initial value and distortion gradually disappears.

That’s it. I thought I’d share this with you guys. I’m sure some of you are much more competent than me to perhaps comment on this.

The problem is probably with the module. Given the low cost of the modules, it's not worth devoting a lot of time troubleshooting.

Is the noise simultaneously in both channels?

I would do easy, simple things. Reflow the solder to the pins of the module, GPIO power and ground (GPIO 2,4,6). A bad solder joint can show up years later, as moisture and corrosion enter the joint. The bad solder joint could also be in the module itself. Not worth pursuing.

You could isolate the ProtoDAC power supply. Pull ProtoDAC off the RPi GPIO. Use DuPont jumpers for the I2S (12, 35, 40) and ground (6) from the RPi. Use a separate power supply to the 5V/GND pads between the RCA jacks. Standard green screw terminals will fit there or you can solder or clip directly. Double check polarity before powering up.

[hifinet]

Other I/V resistor options:

Rhopoint General Resistance GG102 or GR102 (non-RoHS compliant), which is a noninductive wirewound resistor. 0.6mm (22 AWG) wire of the "highest quality copper alloy drawn from
melts of known resistivity and controlled temperature coefficient". The standout spec is "Noise Immeasurable". The "D" version with +/-0.1% should be just fine (to get the cost down). I think these need to be purchased directly from Rhopoint. Made in the UK. Not sure yet where to buy them or the price.

Update: Based on a 2021 diyA Group Buy thread, minimum order $250 and each resistor about $12.50 USD. Cost is probably about the same as TX2575. 

Knockoff GG102 by RARA South Korea. Inquiry emailed.

RMS South Korea. Appears to be a distributor for RARA. Inquiry submitted.

Newark has the 8G16D axial Econistor in limited values. Closest value is 350 ohms. $8.78 USD each. Not in Stock. Manufacturer Standard Lead Time: 17 week(s)
Newark has the 5G10D axial Miniohm in 350 ohm in stock.  250mW rating. About 8mm length and 4mm diameter. $11.02 USD each plus $9.99 shipping.

Manganin (84.2% copper, 12.1% manganese, 3.7% nickel alloy) "roll your own" resistors. The wire is available from eBay Eastern Europe (Ukraine - "This package may be delayed due to shipping carrier service disruptions."). Not inexpensive. Should be wound in a noninductive pattern such as bifilar or Ayrton–Perry. The smaller diameter wire has the highest resistance per meter. These are supposed to be very low noise to no noise. Beyond my capabilities.

[hifinet]

A comment going back to post #659 by thomaschan. The Lamizator SRPP circuit is essentially passive I/V with a 90 ohm I/V resistor (R1), and a tube preamplifier (SRPP) to provide voltage gain to return the DAC output to line level.

When the 90 ohm I/V resistor is substituted, gain is reduced -13 dB compared to the 420 ohm resistor. If you have a preamp (tube or solid state) that can provide that level of gain, you should be able to use the lower I/V resistor value. THD from the DAC itself will decrease from 0.01% to about 0.007%. The decoupling caps (C3/C4) will likely need to be increased in value.

[hifinet]

Order in to Newark for two RHOPOINT 5G10D 350R and two RHOPOINT 8G16D 100R . Parts need to come from the UK warehouse.

The 100R for a direct comparison to 100R TX2575 with preamp at the output of the DAC.

[hifinet]

I should mention that one other wire wound resistor I have tried with ProtoDAC was the Mills MR100, which is a noninductive nickel chromium wire resistor. I just happened to have two 422R in the parts drawer. It sounded dark compared to the TX2575, which is much better top to bottom.

[smokey_mountain]

Folks,
my Christmas gift arrived - Iancanada PurePi II 5V/3.3V Power Supply PurePi_II

I liked the sound of Moode with ProtoDAC so much that I have pulled my collection of MFSL/DCC/XRCD CDs and ripped them to FLAC to re-discover how wonderful moOde plays them on ProtoDAC. This software and hardware combo is nothing short of magical, thus Christmas money went to possible improvements of better power supply and poorman's reclocker - bought Innomaker DAC HAT to reclock.

I carefully read Ian's manual and there is something that confused me. I wonder if anyone is using Iancanada's PurePi II with ProtoDAC combo.

To use Iancanada power supply but without Iancanada's reclockers, I wonder if I can even use his power supply without cutting pins 2,4 on the stack spacer before connecting any HATs to Raspberry Pi's HAT that is connected to the spring pins of PurePi II?

Iancanada says 1,3 - 5V
ProtoDAC manual says that 2,4 - 5V