Tda1387....... the older the fiddle the sweeter the tune

[nige2000]

so what im a doing with a 21 year old dac chip
well during abraxilito 's visit we were discussing non sigma delta chips and this one was of interest although it is discontinued it is cheap (pulled from worthless sound cards i believe) and it is proper good

certainly less risk than butchering a soekris r2r



i botched it together in a rush just to get a feel for it
powering with 6.6v lifepo4
diyinhk usb to i2s and a synchronized relocked usb hub
ive 440r in iv and a 1uf acrylic cap to output
output is too low will have to get some 1-2k r's?
will have to address the frequency droop though i find it quite noticeable

[abraxalito]

Excellent work here - would you like to try my discrete I/V stage? If you're using passive I/V I found the bass suffers. How much capacitance have you got on pin7? I'd recommend 1,000uF as a start.

For I/V resistor you can go up to 3k3 if you're running from 6V.

[nige2000]

whats the iv stage ?
active?

only acrylic 1uf in vref (they were beside the soldering iron looking at me)
ive since tried merging pin7 on two chips which i think was a shade better but marginal
often with lifepo4 batteries capacitors dont behave the way would think
adding a shitty 1000uf cap actually made the sound a little spongey
a similar effect as i seen with vcom on the pcm1794

[abraxalito]

Yeah discrete transistors (about 10 per channel). Current feedback but has classA output stage to get around the SQ issues of IC CFB amps.

[nige2000]

what about using a resistor voltage divider to provide 1/6 of vcc to pin 7?
is that just a mad idea?

how many dac chips could be feasible to stack?

[abraxalito]

What problem are you hoping to solve with pin7? There's one sure-fire way to blow these chips up and that's to put overvoltage on pin7. Its set internally at 1/6th of VDD so there's no way to disconnect the internal resistors. You could try driving the pin with an external buffer, but beware overvoltage particularly on start-up/shutdown if the buffer's fed from separate analog voltage rails. Pin7 works as a volume control if you decrease the voltage there.

Stacking chips - I think I've created a mini-tower of 8 but there's no limit really as they don't get hot like 1543s do.

[jkeny]

Great work
Interesting results on the pin 7 electro cap
Richard, have you blown up some of these chips with overvoltage on pin 7 - how sensitive is it to overvoltage i.e 1.1V is 1/6 of the 6.6V VDD would 1.2V kill it?
I presume by organising the powering up of pin 7 after the main chip VDD & also ensuring it doesn't overvoltage we would be golden?


Combining the trafo ideas & this chip - is there any advantage to using a transformer as a filter & I to V converter? Can it also be used in some way to deal with the NOS droop, either actively or passively?

Pin 7 (voltage reference) seems to be a very critical element of the chips providing a reference to all output stages of the chip. Do you have experience of using voltage on pin 7 as vol control - any sonic downsides to this?

An idea might be to design a trafo to provide more voltage than needed & use lower voltage on pin 7 to bring the volume down to the level needed - preventing any overvoltage.

Is Using a stepped attenuator as shunt R for vol control a worthwhile idea? High wattage resistors needed to drop 5.5V - what current?

[nige2000]

messin with the vref here
to be honest its hard to judge differences
a little difference in the vref voltage and the volume changes :(
made a voltage divider at 1v with melfs
thought it sounded a little duller which it would coz its lower vol hmmn......
put in a trim pot at 1.1v then definitely a little duller
put in a 5f supercap which had to be pre charged along with the trim pot and def improved the sq
removed trim pot sounds same
and now trying to figure out whether supercap in or out sounds best

when i had supercap in i tried vref = 0-2v no smoke only different vol but my iv is only 400r til i get some more suitable resistors tomorrow

maybe no low hanging fruit for me at vref?

[abraxalito]

Richard, have you blown up some of these chips with overvoltage on pin 7 - how sensitive is it to overvoltage i.e 1.1V is 1/6 of the 6.6V VDD would 1.2V kill it?

I've blown one or two with about double the 0.833V which is normal for 5V on VDD. Can't speak about 6.6V as I've never gone that high. They're so cheap it doesn't matter though.

I presume by organising the powering up of pin 7 after the main chip VDD & also ensuring it doesn't overvoltage we would be golden?

I presume so - I've never tried external power, when I blew my chips up I think it was because I tried a resistor to VDD.

Combining the trafo ideas & this chip - is there any advantages to using a transformer as a filter & I to V converter? Can it also be used in some way to deal with the NOS droop, either actively or passively?

Can't think of any advantage of using a trafo as a filter - its filtering parameter (inductance) isn't stable or predictable. When I want a filter I'll use a dedicated inductor - this can then fix up NOS droop by making the filter underdamped. I show an example on the DIYA thread about the Taobao DAC using 8 TDA1387s. Nige is on that thread already....

Pin 7 (voltage reference) seems to be a very critical element of the chips providing a reference to all output stages of the chip. Do you have experience of using voltage on pin 7 as vol control - any sonic downsides to this?

I've not done any serious listening with pin7 set below its normal value but I experimented to hear if there were any obvious artifacts going lower with pin7 voltage and there weren't, its well behaved.

An idea might be to design a trafo to provide more voltage than needed & use lower voltage on pin 7 to bring the volume down to the level needed - preventing any overvoltage.

Can't follow your thinking here.

Is Using a stepped attenuator as shunt R for vol control a worthwhile idea? High wattage resistors needed to drop 5.5V?

Nor here....

[jkeny]

An idea might be to design a trafo to provide more voltage than needed & use lower voltage on pin 7 to bring the volume down to the level needed - preventing any overvoltage.

Can't follow your thinking here.

My thinking was that we can use an IV resistor of greater than 3K3 on the secondary of a say a 1:4 trafo - giving a higher voltage output. By also using a voltage divider on pin 7 at lower than 1.1V we can change the volume output to what we want 1V, 2V or whatever? With this scheme we are always lower than 1/6 of VDD at pin 7 - if that is the critical ratio?

Is Using a stepped attenuator as shunt R for vol control a worthwhile idea? High wattage resistors needed to drop 5.5V?

Nor here....

Substituting an attenuator as the shunt R in a voltage divider could act as a volume control