For those of you who contacted me about PhiDAC kits (who have been patiently waiting): I finally made some progress! I have all the parts broken into individual kits now. I apologize for taking so long on this! We had some big plans fall through this weekend, so I had a big block of unexpected time.
Hopefully next weekend I will have time to run these to the post office and send them on their way. (Emphasis on hopefully!) I will contact everyone individually for shipping details shortly.
Thank you all for your patience!
Hopefully next weekend I will have time to run these to the post office and send them on their way. (Emphasis on hopefully!) I will contact everyone individually for shipping details shortly.
Thank you all for your patience!
No, thank you Matt for giving your time to progress this and anyway, as the old saying goes, 'patience is a virtue'.
Just to echo @nautibuoy's words - thanks Matt for yours in extricating separate kits from the pack.
Dear Abraxalito,
I finally allocated the time to build 2 PhiDAC boards. I have a quite busy end of year, and still need to set-up an amp suitable to listen to ir either with my AKG K-1000 or my Tannoy. My idea is to hook it to a Chalize Tripath TA2020 amplifier that runs on a 12V battery. When the battery is fully loaded, voltage is around 13-14 V
Could I power the PhiDAC with the battery without additional regulator ? What would be the max rating for the power of the PhiDAC ?
Best regards,
JMF
I finally allocated the time to build 2 PhiDAC boards. I have a quite busy end of year, and still need to set-up an amp suitable to listen to ir either with my AKG K-1000 or my Tannoy. My idea is to hook it to a Chalize Tripath TA2020 amplifier that runs on a 12V battery. When the battery is fully loaded, voltage is around 13-14 V
Could I power the PhiDAC with the battery without additional regulator ? What would be the max rating for the power of the PhiDAC ?
Best regards,
JMF
Since @sajunky asked the question I've been mulling it over in my mind and have come up with a solution, of sorts. I've prototyped it on a fully tricked-out PhiDAC SE but can't think why it wouldn't also work on PhiDAC aswell.
The key to adapting the design to work with multiple DAC chips is to reduce the current output to that of a single chip. Which is fairly easily done with a step-up transformer. With the I/V and filter circuits seeing the same current coming in as from a single DAC they don't have to be modified at all - changing their gain and impedances was a big headache and a significant cost-up with all the NP0s that would be necessary.
It turns out that the transformer to do the current reduction can be a rather small one - my prototype is using EP10 (roughly a 1cm-on-a side cube) and it doesn't need thousands of turns (like GrossDAC's EP17) rather just 400 or so. This is because the filter impedance of PhiDAC is much lower. So this trafo is much easier to wind yourself.
If anyone's interested then I'll give more details for how to create your own 'FrankenDAC' - a hybrid with the head of GrossDAC and the body and tail of PhiDAC (or PhiDAC SE).
Last edited:
I measured the 400 turns, it was just a smidgeon under 20ohms DCR. But its not really the DC resistance which matters in the circuit, rather the inductance (800mH). I'm still doing listening to see if having multiple DACs makes a significant improvement to my PhiDAC SE - I'm curious to see if it has a hint of the improved transparency of the GrossDAC.
A quick check shows that Mouser has these : https://www.mouser.com/datasheet/2/400/ep_10-1527669.pdf
T38 or T66 cores look as though they have enough inductance. No, not toroids, those are a big pain to wind more than a couple of dozen turns on!
T38 or T66 cores look as though they have enough inductance. No, not toroids, those are a big pain to wind more than a couple of dozen turns on!
Well my experience is I haven't so far detected anything that detracts from enjoyment of the music from using ferrite-cored transformers, with the caveat that they mustn't be exposed to DC. In experimenting yesterday with the 'FrankenDAC' I neglected to AC couple the transformers from the input of the AD815 and I got a 'hashy' sound on strings as a result, transients seemed over-emphasized. With the caps in circuit to protect the 400-turn secondary from the bias current into the CFA (measured at 89uA) that effect disappeared.
Having listened more I am forming the impression that paralleling DACs does improve transparency - presumably from the lower noise floor. However the effect of the AD815s shows itself on choral works in that individual voices don't get properly separated when compared with my memory of GrossDAC (which has ClassA SE buffers).
Having listened more I am forming the impression that paralleling DACs does improve transparency - presumably from the lower noise floor. However the effect of the AD815s shows itself on choral works in that individual voices don't get properly separated when compared with my memory of GrossDAC (which has ClassA SE buffers).
Same experience as Sumotan (power supply or rca cables)
Abrax, it could be interressant you have an exchange with John from ECDESIGN, he tried multiple parralel dac with tda 1543 and 1545A and came to the conclusion difference of consistency in chips production gave to him bad results at ears and measurements. So he came back with more classic I/V active solution and standalone dac chip. As for traffo, be it before I2S inputt or dac chip outputt, I surmise the too cheap smd or little bulk one we often see in commercial design is more a pain than helpfull.
I'm not aware wether he tried your outputt dac chip filter.
Hope that helps (but always consider I may be deaf
)
Abrax, it could be interressant you have an exchange with John from ECDESIGN, he tried multiple parralel dac with tda 1543 and 1545A and came to the conclusion difference of consistency in chips production gave to him bad results at ears and measurements. So he came back with more classic I/V active solution and standalone dac chip. As for traffo, be it before I2S inputt or dac chip outputt, I surmise the too cheap smd or little bulk one we often see in commercial design is more a pain than helpfull.
I'm not aware wether he tried your outputt dac chip filter.
Hope that helps (but always consider I may be deaf
)
Last edited:
@diyiggy - do you have a link to ECdesigns experiments? I know he has a mega-long thread on TDA1541 but searching that thread is a bit much for me. Seems to me that 1543 has production variations but I can't see how 1545A could have as its a continuous calibration chip. I know ECD has now moved away from COTS DAC chips and rolls his own.
Yes when EC started the thread it was single to double to quad TDAs.
I build a parallel TDA way back using passive I/V but compared to what I have now a highly tweaked AyA dac, my conclusion is what feeds the dac & components used is far far more important & for me an active I/V is way better. Credit goes to Pedja for a great design , looking around, I believe it is the best implemented TDA 1541 dac .
I build a parallel TDA way back using passive I/V but compared to what I have now a highly tweaked AyA dac, my conclusion is what feeds the dac & components used is far far more important & for me an active I/V is way better. Credit goes to Pedja for a great design , looking around, I believe it is the best implemented TDA 1541 dac .
why does Rdc matter? You are dealing with signals, at 20Khz the primary impeadance is 100K so you would get 100V across it for 1mA out of dac. This cant be right, what am I missing?