'braxy mentioned in the Any good TDA1541A DAC kit? thread
It seemed to me at the time the model 777 came out that Thorsten was using a DAC which had technical specs similar (or identical) to the TDA1545. If such a humble chip really does deliver 95% of the SQ of the much more complex TDA1541A then it has to be a chip worth spending time with, no? In my own DAC designs based on this chip, I have recently found I've been limiting its performance by giving it a less than clean PSU so I doubt I've reached the fundamental limits of the TDA1545 yet.
One advantage it has over the TDA1541A from a DIYer's pov is it runs on a single power rail (3-5V). This makes getting good layout an order of magnitude easier. It also needs no fancy tricks with DEM.
http://www.diyaudio.com/forums/digi...any-good-tda1541a-dac-kit-68.html#post3874428
That is an exellent point!
So are there any good kits or competent boards for modding?
For me SPDIF is a can-o-worms and 'dead' plus I've moved to computer and USB so ...
a properly bypassed 1545 connected to a good low jitter USB chip with a simple opamp (high speed) output would be perfect!
The Monica USB DAC had it's issues (broken GND plane) but would be a place to start (esp with jocko style IV) but isn't available:
Monica does USB!
USB Monica DAC | diyparadise blog
I just ordered these:
TekDevice Philips TDA1545A Right Justified Only DAC Module Passive Resistor I/V
Tekdevice Philips TDA1545A Right Justified Only DAC Module Passive Resistor I V | eBay
and
TekDevice Philips TDA1545A WM8804 SPDIF Coaxial DAC Board Passive Resistor I/V
Tekdevice Philips TDA1545A WM8804 SPDIF Coaxial DAC Board Passive Resistor I V | eBay
Critique of there boards?
(I'll let you know when they arrive and take some pics)
Any other inexpensive kits?
The TDA1543 is almost pin compatible so good 1543 kits/PCBs would work too. Any?
Thanks,
Jeff
View attachment TDA1545a.pdf
It seemed to me at the time the model 777 came out that Thorsten was using a DAC which had technical specs similar (or identical) to the TDA1545. If such a humble chip really does deliver 95% of the SQ of the much more complex TDA1541A then it has to be a chip worth spending time with, no? In my own DAC designs based on this chip, I have recently found I've been limiting its performance by giving it a less than clean PSU so I doubt I've reached the fundamental limits of the TDA1545 yet.
One advantage it has over the TDA1541A from a DIYer's pov is it runs on a single power rail (3-5V). This makes getting good layout an order of magnitude easier. It also needs no fancy tricks with DEM.
http://www.diyaudio.com/forums/digi...any-good-tda1541a-dac-kit-68.html#post3874428
That is an exellent point!
So are there any good kits or competent boards for modding?
For me SPDIF is a can-o-worms and 'dead' plus I've moved to computer and USB so ...
a properly bypassed 1545 connected to a good low jitter USB chip with a simple opamp (high speed) output would be perfect!
The Monica USB DAC had it's issues (broken GND plane) but would be a place to start (esp with jocko style IV) but isn't available:
Monica does USB!
USB Monica DAC | diyparadise blog
I just ordered these:
TekDevice Philips TDA1545A Right Justified Only DAC Module Passive Resistor I/V
Tekdevice Philips TDA1545A Right Justified Only DAC Module Passive Resistor I V | eBay
and
TekDevice Philips TDA1545A WM8804 SPDIF Coaxial DAC Board Passive Resistor I/V
Tekdevice Philips TDA1545A WM8804 SPDIF Coaxial DAC Board Passive Resistor I V | eBay
Critique of there boards?
(I'll let you know when they arrive and take some pics)
Any other inexpensive kits?
The TDA1543 is almost pin compatible so good 1543 kits/PCBs would work too. Any?
Thanks,
Jeff
View attachment TDA1545a.pdf
Last edited:
I shall be doing a layout for a very simple TDA1545AT board to fit on top of one of my hexacaps. I'll put up the schematic and layout on http://easyeda.com/ so anyone interested can clone it or mod it. I already have an output buffer schematic (AD815 based) up on there. If anyone's interested in that I can share it. After the DAC board will come a passive filter board - 11th order Chebyshev LC.
Right now the design is in progress, eventually there will be PCBs (later this year). The idea is you can order PCBs using the EasyEDA website. A whole DAC can be built up from a succession of small PCBs.
Unfortunately the TDA1545A is EIAJ input, but I think I'll make the board generic enough that TDA1387T can be used instead, which is I2S input. I'll be working on a DIR9001-based S/PDIF receiver board too which can give out I2S or EIAJ.
Unfortunately the TDA1545A is EIAJ input, but I think I'll make the board generic enough that TDA1387T can be used instead, which is I2S input. I'll be working on a DIR9001-based S/PDIF receiver board too which can give out I2S or EIAJ.
Hi Abrax,
may I ask you where is the outsides load points on your TL431 reg shematic please, I 'm not used to read simulation ?
About your impressive Babel Hexacap : why a super cap 1 F or more with ultra low ESR 0,000X M ohms are not serious : size is big but minus in regards to Hexacap you maid. price is around 60 to 100 dollars for less than 10 V e.g. Is it the time of loading of such a cap if not multiple // small values as you do ?
thank you
P.S. : I like the pristine treble of the TDA1545a, but found myself mid-bass and tonality of the 1541 stay unique... in regards to the two others contanders
may I ask you where is the outsides load points on your TL431 reg shematic please, I 'm not used to read simulation ?
About your impressive Babel Hexacap : why a super cap 1 F or more with ultra low ESR 0,000X M ohms are not serious : size is big but minus in regards to Hexacap you maid. price is around 60 to 100 dollars for less than 10 V e.g. Is it the time of loading of such a cap if not multiple // small values as you do ?
thank you
P.S. : I like the pristine treble of the TDA1545a, but found myself mid-bass and tonality of the 1541 stay unique... in regards to the two others contanders
Hi Eldam
Sorry, can't catch your first question. Are you referring to a schematic on my blog? If so which one?
Second one - supercaps look fine until you get to the ESR spec, then you find its a bit high to make good use of all those Farads. I've recently found a 2.7V/100F with max ESR of 15mohms so I'll probably parallel at least 15 to get to 1mohm or below. Then have a listen. I parallel multiple small values to get low ESR and decently low inductance. I have tried using very big cans but not only are they bulkier they don't measure so well as arrays. I reckon the built up cost will be under $30 for a 1F array but I don't know for sure yet.
Regarding the TDA1545's pristine treble, I suspect you've got an I/V stage which is responding to the glitchiness of CMOS. Put a filter after it and listen again. Tonality is greatly improved with massive capacitance across it, also for pin7 decoupling it needs 2,200uF or more.
Sorry, can't catch your first question. Are you referring to a schematic on my blog? If so which one?
Second one - supercaps look fine until you get to the ESR spec, then you find its a bit high to make good use of all those Farads. I've recently found a 2.7V/100F with max ESR of 15mohms so I'll probably parallel at least 15 to get to 1mohm or below. Then have a listen. I parallel multiple small values to get low ESR and decently low inductance. I have tried using very big cans but not only are they bulkier they don't measure so well as arrays. I reckon the built up cost will be under $30 for a 1F array but I don't know for sure yet.
Regarding the TDA1545's pristine treble, I suspect you've got an I/V stage which is responding to the glitchiness of CMOS. Put a filter after it and listen again. Tonality is greatly improved with massive capacitance across it, also for pin7 decoupling it needs 2,200uF or more.
of course the ONE with the best tl 431 from on Semiconductor of your blog !
I saw some super cap with one zero lower impedance but maybe the farad size or leakage current maid it unusefull ? Or just behind a little smd polymer + smd X7R near the dac chip and the big cap just for the main PS ? I think with such a cap, lead battery is the ideal : we will not see the impedance of the battery with 10 Farads !
I really like the treble of the TDA1541A : it has something special for my ;own taste : pristine but good "sound" (if it has a sense !) and that from what I heard TDA1541 hasn't. But the qualities of the TDA1541 let forget such default, like the tones of guitars e.g.
Is it possible to use the TDA 1545 without its X2/X4/X8 oversampling ?
Some says than the soic version sounds best than the DIP one !
of your blog !I saw some super cap with one zero lower impedance but maybe the farad size or leakage current maid it unusefull ? Or just behind a little smd polymer + smd X7R near the dac chip and the big cap just for the main PS ? I think with such a cap, lead battery is the ideal : we will not see the impedance of the battery with 10 Farads !
I really like the treble of the TDA1541A : it has something special for my ;own taste : pristine but good "sound" (if it has a sense !) and that from what I heard TDA1541 hasn't. But the qualities of the TDA1541 let forget such default, like the tones of guitars e.g.
Is it possible to use the TDA 1545 without its X2/X4/X8 oversampling ?
Some says than the soic version sounds best than the DIP one !
Last edited:
Ah you mean the 'Force/sense shunt' ? The load points are marked 'force' and 'sense' on the schematic and there's a pair of those for each polarity so you should connect with 4 wires.
TDA1545A sounds best without any oversampling yes, run it NOS. I used the DIL version on one build but worried that I might have fakes. I suspect the supply impedance can be lower with the SOIC part as the wires to the die are shorter, so that's the one I prefer to use.
TDA1545A sounds best without any oversampling yes, run it NOS. I used the DIL version on one build but worried that I might have fakes. I suspect the supply impedance can be lower with the SOIC part as the wires to the die are shorter, so that's the one I prefer to use.
Hey 'Braxy, are we talking days, weeks or months for the 1545 PCB?
As well, tekdevice also has a TDA1387 (1545 with I2S) board. Search for:
TekDevice Philips TDA1387 I2S DAC Module Passive Resistor I/V
I'm only one person but I have a great, burning demand ... so that probably equals about 5 people with only a half-assed demand.
Eldam's enthusiasm counts for at least 2 or 3 more demands ... and his fumbly English could be misinterpreted as probably 1 more demand
5 + 3 +1 = 9!
Come on people, pony-up and let's get 'Braxy to shift the 1545 PCB to the front burner and start cookin'!
Last edited:
I have an amendment to this from recent playing with TDA1387 - it seems that the 1387 has a significant advantage in that its output compliance includes 0V. This means I don't have to provide a nice low-noise supply to run the I/V resistors from. Of course the post I/V amp then needs to be able to cope with an input voltage only a couple of hundred mV away from ground, but this seems OK to solve with a suitable negative supply generator (ICL7660 is my first stab). Early days yet but this configuration is sounding more dynamic than having a 2.5V rail sourced from a TL431 with oodles of caps on the output...
I'll take it in small chunks, going back to basics.
A current output DAC has a voltage compliance specification, that's the allowable range of output voltages permitted on its output pin(s). In the case of these 8pin chips, the outputs are pins 6 & 8.
Internal to the DAC is a digitally controlled current source (the DAC proper) and a further current source, called the bias current in the DSs. The DAC proper lies between the output pin and the 0V supply, the bias current source is between the output and the VDD supply (pin 5). I've found that in the case of the TDA1543 at least, it sounds best without the bias current source switched on, however while this trick is possible on the 1543 its not on the 1545 and 1387. The operation of the bias current source appears to be non-negotiable.
The magnitude of the bias current sourced from VDD is proportional to the pin7 voltage, the same voltage also scales the full scale current of the DAC proper.
On the TDA1545, the output compliance runs from 2V to 1V below VDD so is theoretically zero when the supply is 3V meaning to use passive I/V the supply voltage needs to be at least the output swing plus 3V. I've tended to go for 4-5V supplies and normally I've set the output voltage somewhere near 2.5V as this is simple to generate with a TL431 shunt.
To be continued...
A current output DAC has a voltage compliance specification, that's the allowable range of output voltages permitted on its output pin(s). In the case of these 8pin chips, the outputs are pins 6 & 8.
Internal to the DAC is a digitally controlled current source (the DAC proper) and a further current source, called the bias current in the DSs. The DAC proper lies between the output pin and the 0V supply, the bias current source is between the output and the VDD supply (pin 5). I've found that in the case of the TDA1543 at least, it sounds best without the bias current source switched on, however while this trick is possible on the 1543 its not on the 1545 and 1387. The operation of the bias current source appears to be non-negotiable.
The magnitude of the bias current sourced from VDD is proportional to the pin7 voltage, the same voltage also scales the full scale current of the DAC proper.
On the TDA1545, the output compliance runs from 2V to 1V below VDD so is theoretically zero when the supply is 3V meaning to use passive I/V the supply voltage needs to be at least the output swing plus 3V. I've tended to go for 4-5V supplies and normally I've set the output voltage somewhere near 2.5V as this is simple to generate with a TL431 shunt.
To be continued...
I'm only one person but I have a great, burning demand ... so that probably equals about 5 people with only a half-assed demand.
Eldam's enthusiasm counts for at least 2 or 3 more demands ... and his fumbly English could be misinterpreted as probably 1 more demand
5 + 3 +1 = 9!
Come on people, pony-up and let's get 'Braxy to shift the 1545 PCB to the front burner and start cookin'!
Well it's been a week and only me and Eldam have spoken up (not even that jambul guy from the first page)
So 'Braxy, I guess we'll just have to wait till you get around to it ... even thought I'd have to consider the 1545/1387 one of the best Bang-for-Buck in DACs ... nay, all of Audio!
Good luck,
Cheers,
Jeff
Back to the arcane matters of DAC output biassing for a while... An alternative with the TDA1545A which I've not tried is putting the I/V resistor to ground and then biassing it up to a point within the output compliance range. For a 50ohm I/V resistor (which I use because that's the input impedance of my filter) a fairly heavy current (50mA) is needed to achieve 2.5V so not really an option for where low power is called for (portable applications). The CCS should also be chosen carefully for low noise and decent PSRR.
The TDA1387 provides a much more elegant solution even though it calls for an extra -ve supply for the post amplifier (AD815 in my case) because most amp chips can't work with the input voltage so close to their -ve rail. There exist rail-to-rail input opamps though that can, so these are worth a try when the AD815's fairly hefty current draw can't be tolerated.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.