difficulty with 1/2 clock to TDA 1541

[tubee]

Hello all

After a recommendation of Guido Tent i connected a TDA1541 directly on a 1/2 clock signal of 5.6 MHz, by a 74HCT74 flipflop in divide-by-2 mode. The 7220 is removed. I use an XO and sometimes a Kwak 7 via X-in of SAA7210.

The result was: digital noise instead of music, full scale, and far away was some music hearable.
Checked everything, saw a neat squarewave of indeed 1/2 the clock on a scope.

Connected 74/74: 1 and 4 and 14 on +5V, 6(Qnot) to 2(Data) 11.2896 to pin 3 and took 1/2 clock from pin 5.

So what went wrong here?

Maybe the timing is a little out of phase and will a 74HC04 inverter after 74HCT74 will do the trick?

Or any other comments?

[Guido Tent]

Originally posted by tubee
Hello all

After a recommendation of Guido Tent i connected a TDA1541 directly on a 1/2 clock signal of 5.6 MHz, by a 74HCT74 flipflop in divide-by-2 mode. The 7220 is removed. I use an XO and sometimes a Kwak 7 via X-in of SAA7210.

The result was: digital noise instead of music, full scale, and far away was some music hearable.
Checked everything, saw a neat squarewave of indeed 1/2 the clock on a scope.

Connected 74/74: 1 and 4 and 14 on +5V, 6(Qnot) to 2(Data) 11.2896 to pin 3 and took 1/2 clock from pin 5.

So what went wrong here?

Maybe the timing is a little out of phase and will a 74HC04 inverter after 74HCT74 will do the trick?

Or any other comments?

Hi

I never checked the proposed metod with the 7220 removed, so I can't judge on this.

What 1541 clock connection would you have used without /2 mode ?

best

Guido

[tubee]

Hello Guido

Thanks for reply, i contacted you by mail last week but received no answer.

I now reconnected the 1541A's pin 2 and 4 on the clock signal of the SAA7210, 7220 is removed. So now WS, Data and CLAB from 7210 goes to 1541 via 47ohm resistor.
Clock's used: XO and sometimes a Kwak 7, on X-in of 7210.

Regards:

[Guido Tent]

Originally posted by tubee
Hello Guido

Thanks for reply, i contacted you by mail last week but received no answer.

I now reconnected the 1541A's pin 2 and 4 on the clock signal of the SAA7210, 7220 is removed. So now WS, Data and CLAB from 7210 goes to 1541 via 47ohm resistor.
Clock's used: XO and sometimes a Kwak 7, on X-in of 7210.

Regards:

Hi Tubee,

Correct, I am somewhat behind in answering mails

Does above config work ? If so, what is the clock freq at pin 2/4 ?

[tubee]

It works, so this player is in Non-os now. At the moment Kwak clock, I/V by 4 transistor scheme of RBroertjes.

I will measure the signal-frequency on pin 2-4 of dac tonight if possible.

regards

[tubee]

Guido you helped me on rails again.
measured pin 2-4 from TDA: One waveform is about 0.24msec. and looks like a four times slower signal than 11.2896, so would be 2.8224Mhz.

Digged in the serv. manual also: CLAB clock from 7210 to 7220 is 2.8224, CLBD is I2S like 5.6448 MHz.

The Word Select signal does a lot the for the dac, otherwise there should be no music hearable in non-os.

Now we have to connect the 74HCT74 in divide- by-4 mode.

Thanks

[guido]

Originally posted by tubee
Guido you helped me on rails again.
measured pin 2-4 from TDA: One waveform is about 0.24msec. and looks like a four times slower signal than 11.2896, so would be 2.8224Mhz.

Digged in the serv. manual also: CLAB clock from 7210 to 7220 is 2.8224, CLBD is I2S like 5.6448 MHz.

The Word Select signal does a lot the for the dac, otherwise there should be no music hearable in non-os.

Now we have to connect the 74HCT74 in divide- by-4 mode.

Thanks

Hi, Not shure this would be the answer to the solved problem, but here's some calculations:

7210 out is 32 bit's for one (16 bit) sample (16 not used):
(32 +32) * 44.1 * 1000 = 2.8224 MHz

7220 out is 16 bit's for one (16 bit) sample
(16 + 16) * 44.1 * 1000 * 4 = 5.6448 MHz

The DAC can handle these spare bits (e.g. crystal receivers are 24 bits iirc).

As for pin 2 and 4, pin 2 (iirc) is not used with tda1541A (A!) in i2s mode.

GuidoB

[tubee]

GuidoB wrote:

7210 out is 32 bit's for one (16 bit) sample (16 not used):
(32 +32) * 44.1 * 1000 = 2.8224 MHz

7220 out is 16 bit's for one (16 bit) sample
(16 + 16) * 44.1 * 1000 * 4 = 5.6448 MHz

-----------------------------------------------------
Thanks for this simple but clear calculation GuidoB, now for me its clear why there is a only 2 times frequency-difference of clock between 7210 and 7220, would epect with 4 times with OS, but its because 7210 has to handle the 16 subcodebits extra.

Did some thinking of the 74HCT74, divide-by-4 is less easy then.

Better take a 74HC4040, you can choose different dividers also.

But with scope the waveform of 2.8224 MHz from the SAA7210 didn't look bad at all, and i can imagine the SAA7210 doesn't give a lot of pollution as a SAA7220 would on its transmitting clock signal.

Maybe we then have revealed here a little bit more of the quest: why does Non-os sounds better for a lot of people???

[tubee]

Just read here its not wise to choose a 74HC4040:

http://www.diyaudio.com/forums/showt...730#post706730

Anyone an other option for a good divide-by-4 schematic? Meanwhile will search myself also.

Oh and forgot:

pin 2 (iirc) is not used with tda1541A (A!) in i2s mode

I use 1541 'A' version but pin 2 is still on CLAB signal as with pin 4.

[rfbrw]

Originally posted by tubee
Just read here its not wise to choose a 74HC4040:

http://www.diyaudio.com/forums/showt...730#post706730

Anyone an other option for a good divide-by-4 schematic? Meanwhile will search myself also.

One 74HC163 or a pair of '74's.