ProtoDAC TDA1387 X8 project

[DRONE7]

A cautionary tale...make of it what you will.

I have 4 ProtoDacs... 2 with the module soldered in place and 2 with the module pressed in to soldered headers.

Both soldered versions have failed. Low or no output on one channel.
Both pressed via headers perform fine.

From several posts on this and other threads I suspect the modules are very temperature (solder) sensitive and would suggest that using headers is beneficial.

Today I tried to desolder an errant module that was hardwired.
Would not recommend !
I finally managed to remove the module and replaced with headers and a new module.

I now suspect the soldered module was compromised from installation and had deteriorated  prior to its dropout.
The replacement has all the life and sparkle of the original and with a now pressed fit I hope for it to continue as such.

However, I am now calendaring to unplug-replug modules in their newly installed headers and perhaps some Cramolin.
(have several modules in reserve just because)

[eculcea]

A cautionary tale...make of it what you will.

I have 4 ProtoDacs... 2 with the module soldered in place and 2 with the module pressed in to soldered headers.

Both soldered versions have failed. Low or no output on one channel.
Both pressed via headers perform fine.

From several posts on this and other threads I suspect the modules are very temperature (solder) sensitive and would suggest that using headers is beneficial.

Today I tried to desolder an errant module that was hardwired.
Would not recommend !
I finally managed to remove the module and replaced with headers and a new module.

I now suspect the soldered module was compromised from installation and had deteriorated  prior to its dropout.
The replacement has all the life and sparkle of the original and with a now pressed fit I hope for it to continue as such.

However, I am now calendaring to unplug-replug modules in their newly installed  headers and perhaps some Cramolin.
(have several modules in reserve just because)

I installed a socket on my ProtoDAC and I can easily change the modules. Recently I had couple of modules go bad and I started to think what could be the problem since it worked well for a while. The I realized lately I changed my setup to something different and for this setup I need to bring the ProtoDAC output volume to the max, I think this is what burned out the TDA1387x8 module.

[hifinet]

I have personally not experienced a module failure. I only use and sell modules that test with the highest voltage output. The modules are made with recycled chips, so they may be prone to failure. To date, with the hundreds of kits I have sold, no one has reported a module failure.

I think damage to the chips due to soldering the module pins would be difficult, because the trace length from module pin to chip lead is relatively long.

I recommend soldering only the 7 required pins (1, 2, 3, 5, 6, 25, 28) to make any module exchange easier. I would recommend removing a bad module by cutting the pins, and then desoldering each pin individually.

[hifinet]

I do not know if repeated clipping (due to too high a resistance I/V resistor) will lead to premature chip failure.

I have no explanation how a single channel in all eight chips could fail at once, as experienced by DRONE7. Of the modules that I reject for the kit (about 7%), most have a channel failure in one chip (manifest by low output in one channel). A much smaller number have low output in both channels.

The TDA1387 has a maximum supply voltage rating of 5.5V (and absolute max rating of 6.0V), which is close to the typical voltage. This is similar to the TDA1545, which are both CMOS. This compares to the TDA1543 with max rating of 8.0V (absolute max of 9V).

I obtain modules only from HuaXinHuiDianZi Store on Ali (w0174.jgvknymwp on eBay is the same seller, I think). He was the only seller of modules on Ali when I first obtained them (except for one other seller WhatSell DKK Store who sold in lots of 10; bought once from him, he shorted me and had high failures, never bought again from him).

[DRONE7]

From further investigation it seems the module that died in one channel was the result of a power brownout (I'm assuming here as we have had several recently).

This seems to have caused an intermittent fault in the power supply I was using for the ProtoDac and I have recently tracked it down and repaired.

I've decided to beef up the supply and instead use a Salas-L-Adapter to power the Pi/Kali/ProtoDAC stack.

However, I noticed the Pi showing under-voltage during boot and excess heating of the L-Adapter during use so have returned to powering the Pi from it's official supply and the Kali/ProtoDAC from the L-Adapter.

From your recent post on voltages I was prompted to reread the Data sheet for the TDA1387 and noted the  voltage range is specified at 3V-5.5V.

So I'm currently experimenting with de-rating and have reset the L-Adapter out to 4.85V.
Next step may be to power both the Kali and Pi from the Pi supply and the ProtoDAC from the L-Adapter and see what comes of de-rating lower.

Fun!

Thanks for the tip re cutting the old module before desoldering !. 
I must have been overeager and in hindsight should have removed the output caps to allow cutting the module free...but thought that with so few pins soldered it would be an easy removal.

edit after listening...Heavenly!

I've felt that (Qualified!) -in my system- the ProtoDac can be a 'little' bass forward and 'smeared' at the bottom end.

Today it is so much closer to perfect that I am hesitant to experiment further with dialing back the volts..

My go to album is Jacques Loussier Trio 'Play Bach' (1959)
https://www.discogs.com/master/3394279-J...-Play-Bach

Today it is magnificent!  The resonance of the lower registers of the piano ring out and the plucked string bass is fully there!

Where previously I dialed back the volume, today I pumped it. Detail, detail, detail and as Dick Shahinian said..
https://flic.kr/p/hsdndC

Find the original album as the later one here is too closed-in to be of use.

https://www.youtube.com/watch?v=9zO_v3HP7Wc

[itchycoo52]

Hi
I am building a protodac. I have just soldered an audionote kaisai non polar capacitor 2200uf in position c3, c4 in place of the suggested two 1000uf parallel electrolytics. I am testing components before insertion with either  a Peak atlas gold esr meter or a multimeter. I measured the cap before insertion with the esr meter and it was showing correct value with zero esr.

After soldering in it was reading no value only saying in circuit/leaky.
Can anyone advise please if the cap has gone bad due to the heat during soldering. I did in fact use heatsink clips whilst soldering. Or is it in fact because it is in circuit. The only other components on the board though are a socket for the module and the five resistors.

Would it be advisable to replace the cap now or leave it until the build is completed and maybe doesn't work and perhaps cause damage. I do have two panasonic 1000uf to replace it with if neccessary.
Steve

[hifinet]

The cap ESR needs to be measured out of circuit to determine if it is good or not with an ESR meter. If the cap was new, it would be difficult to damage by soldering. As long as it doesn't measure a dead short, it should be good.

[itchycoo52]

hifinet
Thanks for your reply. Just got a new multimeter with capacitance measurement range on it and tested the cap still soldered in circuit and its reading 2200uf so all good for the moment, will continue on with the build. I am building a stand alone protodac without using a rasberry pi, using a dir 9001 board to get a spdif input from my cd transport. I may need further help and advice about the i2s connections between the protodac and the dir board.