The HC86 can be thought of as a conditional inverter - with data going into one input, if the other input is 0 then it won't invert, with this second input at 1, it will. Incidentally since the TDA1543 was originally designed to be used with inverting (opamp) I/V, it contains inversion as a matter of course. So to get true phase out of it in passive I/V (as in the Muse) you'd need to run with one input of the HC86 = 1 so that it inverts the digital data. I mentioned this point on the Head-fi (800+ posts long) thread on the Muse 
hi, sorry for resurrecting an old thread
i have hifiman hm601 portable player which occupies tda1543 in it. from what i can tell, it uses 5v voltage regulator to feed stable supply to the dac chip. the I/V stage is a low pass filter occupied by LM833 and JRC4560
what is the suitable bias resistor value in pin 7? is there any calculation for it? i can't find any useful reference regarding this Rref in the tda1543 datasheet. i want to achieve maximum sound quality and i don't care about the battery life (i can play with the I/V stage & headphone amplifier stage later to achieve longer battery life)
thanks in advance
i have hifiman hm601 portable player which occupies tda1543 in it. from what i can tell, it uses 5v voltage regulator to feed stable supply to the dac chip. the I/V stage is a low pass filter occupied by LM833 and JRC4560
what is the suitable bias resistor value in pin 7? is there any calculation for it? i can't find any useful reference regarding this Rref in the tda1543 datasheet. i want to achieve maximum sound quality and i don't care about the battery life (i can play with the I/V stage & headphone amplifier stage later to achieve longer battery life)
thanks in advance
Last edited:
The standard Vref feeds the voltage reference for the opamp.
To get high quality go passive, then you must load the Vref pin with a resistor and also load the output with a resistor. The values depend on the voltage you have (and the number of parallel units).
It can be done with 5 volt but you get higher quality with about 7 Volt (i have that) without warming the device.
I made a table once of most used values but must look it up.
Anyway, you must have a potmeter for Vref, and a variable Vb too (best is between 6-7.5V) so you can set the operating points. There is a point of equilibrium where the saturation of the top and bottom of a 0 dB sine wave are equally low. (You can see it on a scope).
You can find the info how to calculate Rref here: http://www.diyaudio.com/forums/digi...ng-dac-complementing-cd-pro-7.html#post173483
peter, sorry to bother again. but i need to ask something
it seems like I/V stage, supply voltage, and Vref are playing key role for calculating the Rref value
but as i mentioned earlier, HM601 uses active I/V stage (2 opamps constructing a low pass filter), and there is no I/V resistor between the dac output and opamp input.
can i calculate the Rref by simply remove the existing Rref and measure the current flowing from the Vref to ground, then simply divide 2.2V with the measured current to obtain the value of Rref?
and why did you suggest using 3R8 ohm for Rref in this post?
sorry for the noob questions..
it seems like I/V stage, supply voltage, and Vref are playing key role for calculating the Rref value
but as i mentioned earlier, HM601 uses active I/V stage (2 opamps constructing a low pass filter), and there is no I/V resistor between the dac output and opamp input.
can i calculate the Rref by simply remove the existing Rref and measure the current flowing from the Vref to ground, then simply divide 2.2V with the measured current to obtain the value of Rref?
and why did you suggest using 3R8 ohm for Rref in this post?
sorry for the noob questions..
Last edited:
3R8 is for voltage regulator, not V ref.
As to other question, I'm afraid I can't be much of help.
Last edited:
It depends what quiescent output voltage you want from your LM833s. There's a minimum value it can be - 910R (in E24 series), but no maximum.
To come up with an optimum we'd need to know the supply rails to the I/V stage (LM833) and its feedback resistor value. That would allow setting the quiescent output voltage mid-way between the rails.
i'll look into that later, but as far as i know, the rail to rail supply of LM833 is -5 to +5V. so we need to set the Vref at 2.5V?
If by Vref you mean pin7 of the TDA1543, its set internally to 2.2V. Hanging a resistor off that pin, provided its not smaller than 910R won't change the voltage.
Given that the LM833 has balanced supplies we'd want its output to sit at 0V when no music is playing. Are the positive input pins (pin3,5) connected to pin7 of the DAC chip? If they are (even via some series resistance) then the LM's outputs would sit at 2.2V with no current coming out of the DAC.
The mid-scale current for the 1543 is 1.15mA typically, to get the bias current to equal this (for no net current out) needs 0.575mA as the Vref current, meaning Rref=3.8k. However we'd not like 2.2V out of the LM833, rather 0V which is done by increasing the bias (decreasing the Vref resistor below 3k8). The amount depends on the gain of the LM833 stage but I'd make a guess that for 2.8V peak signal the I/V resistor will be 2k4 meaning the bias needs to deliver an extra 1mA roughly to get the output to zero volts. Thus I'd go for a 2k Vref resistor as first stab
Given that the LM833 has balanced supplies we'd want its output to sit at 0V when no music is playing. Are the positive input pins (pin3,5) connected to pin7 of the DAC chip? If they are (even via some series resistance) then the LM's outputs would sit at 2.2V with no current coming out of the DAC.
The mid-scale current for the 1543 is 1.15mA typically, to get the bias current to equal this (for no net current out) needs 0.575mA as the Vref current, meaning Rref=3.8k. However we'd not like 2.2V out of the LM833, rather 0V which is done by increasing the bias (decreasing the Vref resistor below 3k8). The amount depends on the gain of the LM833 stage but I'd make a guess that for 2.8V peak signal the I/V resistor will be 2k4 meaning the bias needs to deliver an extra 1mA roughly to get the output to zero volts. Thus I'd go for a 2k Vref resistor as first stab
If by Vref you mean pin7 of the TDA1543, its set internally to 2.2V. Hanging a resistor off that pin, provided its not smaller than 910R won't change the voltage.
Given that the LM833 has balanced supplies we'd want its output to sit at 0V when no music is playing. Are the positive input pins (pin3,5) connected to pin7 of the DAC chip? If they are (even via some series resistance) then the LM's outputs would sit at 2.2V with no current coming out of the DAC.
The mid-scale current for the 1543 is 1.15mA typically, to get the bias current to equal this (for no net current out) needs 0.575mA as the Vref current, meaning Rref=3.8k. However we'd not like 2.2V out of the LM833, rather 0V which is done by increasing the bias (decreasing the Vref resistor below 3k8). The amount depends on the gain of the LM833 stage but I'd make a guess that for 2.8V peak signal the I/V resistor will be 2k4 meaning the bias needs to deliver an extra 1mA roughly to get the output to zero volts. Thus I'd go for a 2k Vref resistor as first stab
thank you for the very clear explanation abraxalito. i have to check the pcb first with my multimeter to test the connection
to manually set the LM833 output at 0V, can i just play with Rref until i get 0V at the LM833 output? we can measure LM833 output by tapping the red and black probe to pin1 of the LM833 and active filter ground right?
will report back here again after i got something to report. thanks!
Last edited:
"Digital" Hook up wire
Peter,
I hope this is not to far afield from the topic here but I suspect others might be interested, too.
What wire do you /would you use for digital signals - internally from a dac board to a buffer? Not speaking of transport to DAC ...
Have you found wire to make much a difference here? And if so, what do you like?
THANKS,
Peter,
I hope this is not to far afield from the topic here but I suspect others might be interested, too.
What wire do you /would you use for digital signals - internally from a dac board to a buffer? Not speaking of transport to DAC ...
Have you found wire to make much a difference here? And if so, what do you like?
THANKS,
Hi Peter and Bobken, and all you other Nossers ,
I've very much enjoyed reading about your development of this humble little chip. I am about to try it out for myself...
I have a question of course ;-) I assume someone must have tried a potential divider instead of a single I/V resistor to ground ? If so, why was it rejected ?
I was thinking a 4.7R series resistor, with the resistor to ground unchanged.
It would increase output impedance, and lower output voltage a fraction, but not significantly and would isolate the dac from cable capacitance ?
In my application, the DAC will feed 1m upocc cables with no DC coupling cap at the DAC. I have 3.3uf Mundorf silver/oil input caps with 100K input impedance at the amp input.
My apologies if this has been covered somewhere, but in my defence, in all the posts and reading I've done on the TDA1543, I can only find one reference to a potential divider here :
http://myweb.tiscali.co.uk/g8hqp/audio/TDA1543IV.html
I hope Dave Kimber will not mind me quoting him. He says, "If people really must have passive I/V then the best option is to use a potential divider which will set a standing DC bias at the output, but very few designs do this."
Thanks for any comments.
Tom
,I've very much enjoyed reading about your development of this humble little chip. I am about to try it out for myself...
I have a question of course ;-) I assume someone must have tried a potential divider instead of a single I/V resistor to ground ? If so, why was it rejected ?
I was thinking a 4.7R series resistor, with the resistor to ground unchanged.
It would increase output impedance, and lower output voltage a fraction, but not significantly and would isolate the dac from cable capacitance ?
In my application, the DAC will feed 1m upocc cables with no DC coupling cap at the DAC. I have 3.3uf Mundorf silver/oil input caps with 100K input impedance at the amp input.
My apologies if this has been covered somewhere, but in my defence, in all the posts and reading I've done on the TDA1543, I can only find one reference to a potential divider here :
http://myweb.tiscali.co.uk/g8hqp/audio/TDA1543IV.html
I hope Dave Kimber will not mind me quoting him. He says, "If people really must have passive I/V then the best option is to use a potential divider which will set a standing DC bias at the output, but very few designs do this."
Thanks for any comments.
Tom
Last edited:
I never tried potential divider, so can't really comment on that.
What I found so far about digital signal path, is that it behaves exactly the same way as analog signal path, every single component influences the sound. I usually try to install all input sockets directly on a board so all connections are done through pcb traces. If I need to use wire, I usually go with DH Labs solid silver hook up wire or this one: http://www.diyaudio.com/forums/audi...-kit-building-instructions-2.html#post1515208
What I found so far about digital signal path, is that it behaves exactly the same way as analog signal path, every single component influences the sound. I usually try to install all input sockets directly on a board so all connections are done through pcb traces. If I need to use wire, I usually go with DH Labs solid silver hook up wire or this one: http://www.diyaudio.com/forums/audi...-kit-building-instructions-2.html#post1515208
Peter
looks like your curent dac schematic differ from http://www.diyaudio.com/forums/digital-source/64738-non-os-tda1543-pcb-layout-10.html#post782209
would you tell us why omit 3300 pF from filt pin - as I read It filtrate noise
what type resistor and cap recommend to use for PLL - Is ordinary Vishay Dale RN good for resistor and how about Vishay Rodersatin MKT for cap
looks like your curent dac schematic differ from http://www.diyaudio.com/forums/digital-source/64738-non-os-tda1543-pcb-layout-10.html#post782209
would you tell us why omit 3300 pF from filt pin - as I read It filtrate noise
what type resistor and cap recommend to use for PLL - Is ordinary Vishay Dale RN good for resistor and how about Vishay Rodersatin MKT for cap
This capacitor is not specified in CS8412 datasheet and was popularly known as "wildmonkeysect tweak": http://www.diyaudio.com/forums/digital-source/21544-cs8412-2.html
In our listening tests the modded DAC sounded better without 3n3 cap.
For PLL I'm using nude Vishay and BG NX
In our listening tests the modded DAC sounded better without 3n3 cap.
For PLL I'm using nude Vishay and BG NX
1543 Vref
hi all , finding this page yesterday http://myweb.tiscali.co.uk/g8hqp/audio/TDA1543IV.html
i wonder why many people use 8V with 1.5K and 2.2K to 2.7K output resistors and not the values seen in this page?
is it just about 2nd order distortions or something else?
hi all , finding this page yesterday http://myweb.tiscali.co.uk/g8hqp/audio/TDA1543IV.html
i wonder why many people use 8V with 1.5K and 2.2K to 2.7K output resistors and not the values seen in this page?
is it just about 2nd order distortions or something else?
Hi,
I have the "Muse Mini 4xTDA1543" and have succesfully used these resistors:
680 ohm Dac resistor replaced with 212 ohm (270//1k SMD)
390 ohm Bias resistor replaced with 243 ohm (1k//1k//1k//1k//10k SMD)
Vout = 0.675 Vrms
THD+N = 0.029 % (2. harmonic dominant) Measured with Asus STX, 96kHz
I used a 1K+100 ohm multi trimpot to find the optimum Rbias value. It is important to do this to find the sweet-spot, and then later replace with resistors.
While messing around, I tried to tie a capacitor (47uF) across the bias resistor. Some how this raises the hole noise floor about 12 dB, and adding some nice analog "noise" to the music. Using a smaller Cap value of 1uF gives rise to the noise above 10kHz.
Have anyone else tried this ? (the 2. and 3. harmonic stayed at the same amplitude while this cap was on)
Regards
Flemming Bach
Denmark
I have the "Muse Mini 4xTDA1543" and have succesfully used these resistors:
680 ohm Dac resistor replaced with 212 ohm (270//1k SMD)
390 ohm Bias resistor replaced with 243 ohm (1k//1k//1k//1k//10k SMD)
Vout = 0.675 Vrms
THD+N = 0.029 % (2. harmonic dominant) Measured with Asus STX, 96kHz
I used a 1K+100 ohm multi trimpot to find the optimum Rbias value. It is important to do this to find the sweet-spot, and then later replace with resistors.
While messing around, I tried to tie a capacitor (47uF) across the bias resistor. Some how this raises the hole noise floor about 12 dB, and adding some nice analog "noise" to the music. Using a smaller Cap value of 1uF gives rise to the noise above 10kHz.
Have anyone else tried this ? (the 2. and 3. harmonic stayed at the same amplitude while this cap was on)
Regards
Flemming Bach
Denmark
Hi, I also have the Muse Mini 4xTDA1543 and after several mods I successfully terminated the input DIR9001 receiver. Anyway, the problem with connecting a capacitor across the bias resistor is that the resistor is fed by a DC reference voltage. This reference voltage has an output impedance considerably lower than series Bias resistor. If you short the bias resistor out with a capacitor the gain goes up for AC noise. It is like having a capacitor across the resistor in the cathode of a tube circuit. This is done to intentionally increase the audio AC gain of the tube circuit. However you are trying to maintain a constant DC current in the reference circuit. This brings to mind putting in a small inductor in series with resistor to decrease more AC gain and noise, particularly at higher frequencies. regards, Gerrit Aartsen