DIR1703 vs DIR9001

cyl

Member
2006-07-09 4:53 pm
Due to the bug in DIR1703, TI releases a DIR9001 in Dec. as replacement. I just bought two from digikey and wire in today. I find that DIR9001 is more stable than DIR1703 but the sound is a little bit less transparent and extend. The sound is a little bit close to CS841X family but still much better than CS8414/CS8416. It is still the best choice using this as input receiver but it is not worth to upgrade if you are using DIR1703 and happy about the performance.
According to the datasheet, DIR1703 uses SpAct clock recovery architecture and DIR9001 uses conventional PLL. Is this the main reason of less transparent and extend? I don¡¦t know what is spAct.
I hope the sound can improve after enough warm up. Is there anyone try to use DIX4192? Can you share your experience?

My home blew DAC:
TI CD74HC04 +DIR9001+SRC4192+DF1706+4 * PCM1704K(Balance output) + output transformer
Running at 24bit 192K, 4x Oversampling
 

hwb

Member
2005-01-24 9:28 am
Re: DIR9001

Cobra2 said:
is it a direct replacement? (pin & size?)

Have a look at the Datasheet, pages 26 and 27 for the differences.

The pinout is the same, although pins are named differently. Package is annoyingly different, the 1703 comes in a MO-150AH, the 9001 has the smaller MO-153AE. Pitch is the same, perhaps you get it soldered somehow on the old footprint, but you have to change wiring of pin 19 and the values for the loop filter components.

HTH,
Holger
 

cyl

Member
2006-07-09 4:53 pm
The pinout is the same but the components use in the digital filter are different. One is using filter to ground and one is using filter to Vcc. The setting between pin15 to pin19 are different. You need to read through the manual before you replace it.

28TSSOP can still solder in the 28SSOP board. After the long run-in, DIR1703 is still better than DIR9001. I roll in back and planning to use the DIR9001 for the next project. My next project is TAS3103..
 

cyl

Member
2006-07-09 4:53 pm
banny_seng said:
Very happy to hear about the comparison of DIR9001 and DIR1703.
I am using the DIR1703 too and going to try the DIR9001 also, I would like to tell you that the 26LS32/26C32 instead of 74HC04 will be a very good upgrade, you can try it if you have time.


I don't have 26LS32/26C32 but I have MAX1487. I still don't have time to use it. Where can you buy the 26L32/26C32? Normally where can use buy the component in Macau?

74HC04 is the input buffer. For the transmitter, I highly recommend dit4192. It can upgrade any CD player. It turns my NEC6Xi CDRom into a high-end transport. This is PLL1707+DIT4192. It has digital balance output and optical output. So I really want to know the performance of DIX4192 or SRC4192, they support balance input.
 

Attachments

  • dsc03271.jpg
    dsc03271.jpg
    26.1 KB · Views: 1,918

banny_seng

New member
2006-09-30 9:26 pm
hwb said:
Has anyone done further measurements on the 9001 meanwhile?
I'd be interested how the clock quality of the 9001 is compared to the 1703 or the crystal chips. Does it suppress jitter as good as the 1703?

In my DACs, DIR9001 is much better than the DIR1703, it give very clear sound stage, very reasonable dynamic range, everything is more ture and real.
according to the datasheet, the jitter of DIR9001 is 50ps, but the jitter of DIR1703 is 75ps.
 

banny_seng

New member
2006-09-30 9:26 pm
cyl said:



I don't have 26LS32/26C32 but I have MAX1487. I still don't have time to use it. Where can you buy the 26L32/26C32? Normally where can use buy the component in Macau?

74HC04 is the input buffer. For the transmitter, I highly recommend dit4192. It can upgrade any CD player. It turns my NEC6Xi CDRom into a high-end transport. This is PLL1707+DIT4192. It has digital balance output and optical output. So I really want to know the performance of DIX4192 or SRC4192, they support balance input.

I bought the ICs (26LS31, 26LS32, 26C31 and 26C32) in China, Macau have no good component to be bought.

For my self made Philips CDM3 transporter, I used a 74ACT74 to reclock the spdif signal, using the same clock souce of the CDM3 but divided to lower frequency, then use the 26C31 for output stage, each IC has its own regulator. Additional, use my hand made pulse output transformer to make a SPDIF output. This CDM3 is better than my VRDS-T1 transporter, even the VRDS has modified a lot and better than original a lot.

At input stage of DAC, also use my hand made pulse transformer but in different turn ratio, then pass to 26C32 balance input, then pass to DIR9001 -> PMD100 -> PCM63K2 -> tube or OPA627 IV. In my case, PCM63K2 is much better than NOS1541A S1 or NOS1541 S1, also better than PCM1794 X 2....
 
cyl said:
The sound (of the DIR 9001) is a little bit close to CS841X family but still much better than CS8414/CS8416.

Sorry, I missed this thread before. BUT: Can anybody comment on the above quotation (appeared in Post no. 1)? Is this the case also by the "common" 16/44 implementation???


banny_seng said:
At input stage of DAC, also use my hand made pulse transformer but in different turn ratio, then pass to 26C32 balance input, then pass to DIR9001 -> PMD100 -> PCM63K2 -> tube or OPA627 IV. In my case, PCM63K2 is much better than NOS1541A S1 or NOS1541 S1, also better than PCM1794 X 2....

Very interesting! PCM63-K2 should sound indeed EXCELLENT. Can you tell how is the PDM100 doing - if compared to a BB DF-1704?

p.s. I Hope I am not disturbing the - yet unanswered - questions as appeared in Posts #12 and #13.
 

mako1138

Member
2007-11-25 11:52 pm
One interesting thing in the DIR9001 datasheet is figure 3 on page 7, which shows jitter vs fs vs SCKO. The best jitter recovery occurs with SCKO = 512 fs, if the data can be believed.

The WM8804 doesn't have an analogous chart, just "50ps" in the TYP column. It sure has a lot of options though.

To those of you looking for schematics, I'm working on a basic DAC that uses a DIR9001. It will be done soon, hopefully.