Off board I/V
I do it all the time. Keep things as short as possible. It is really the most practical way to do it. Grounds... Star the ground where possible. Hope that helps.
Here my progress improving the performance of philips CD723:
Implementing Joe`s suggestions like change cap between + and – of AD844 from 47N to 330N and changing the components from Pin5 to Gnd to 16K2 and 47P. Add a cap of 0.1F/5.5V to the supply pins of TDA1545. Adjusting offset due to change of gain.
After reading forum member JUANEDA´s description of Jundac Five and his contributions on this forum I decided to do a low voltage reference by my own in order to feed the Iref pin with a voltage rather than a current.
This would also allow for a higher full scale current (2mA instead of 1mA) as recommended by the datasheet and so a higher output voltage.
Results:
My TDA1545 has an internal resistor of 9.24K at pin7 and so I adjusted my reference to 1.386V in order to get around 2mA full scale.
The actual measurements give 4.3Vss @1kHz and 3.5Vss @20kHz output corresponding to a drop of 1.83dB. Good enough for me.
Sound:
Before these latest mods it sounded already very good. Now, is it even better? Difficult to tell, but I like the higher output which is easier to interface. Apart from that I like the sound, it just feels right and alive. I am missing nothing, it leaves me enjoying the music.
Further improvements:
For sure there are some like adding shut regulators to the supply of the AD844, adding better and more caps, buy a superduper reference and so on. And….clear up the “birds nest”.
A last it is time to thank everybody here on the forum whoms work and ideas fed my inspiration.
Attachments
For good or for bad, you are doing your own thing, even if after listening to others. Adding a cap of 0.1F/5.5V to the supply pins of TDA1545 can only help, but it does not have the same level of improvement that it gives delta-sigma DACs, only because d-s has a weakness that 1545 does not.
Cheers, Joe
Doing your own thing - in my understanding this is the nature of diy. I admit I should have done one change after another but for me its time that matters. I just am not willing to spend hours and hours for a "scientific" investigation jeopardising to fool my ears and brain. My ears are for sure not the "golden ones" and after all it`a spare time hobby.
Despite I appraise your opinion and expertise - just see my equipment (Elsinore, FVP5A, OPPO BDP-95 and different tube amps).
Regards, Jürgen
You could remove 330nF capacitors connected across DAC output quite easily - unsolder one leg only and see what happens....
Nick
Agreed. Should I just remove the 330N caps or go back to 47N where I came from?
Funny enough, although people are reporting that the data that SAA7378 is supplying (CD723) is not bit-correct and the DAC is operating at 4xfs, I can hear no flaws. Compared to my also modded Oppo BDP-95 the difference is getting very small.
Regards, Jürgen
.
To those many that PM'ed and emailed me asking if I did the 3 x AD844 and it's buffers stack yet.
Yes I got to do it, I've only listened to 1 track and was very pleased with what happened, the speakers now may as well be in another state, such was the sound it gave.
Just holographic is the word that came to mind, nothing saying to my brain and eyes that the sound was coming from the speakers, they just vanished
I've ask Art Vandelay seeing he's just up the road for us to do a 2x (his) vs my 3x before he changes his to 3, on his B&W 800d's. As we both have the PCM1704 based Cary 303/200 cdp's
I will get back again with more descriptive differences between the two, after we've done the A/B.
Cheers George
To those many that PM'ed and emailed me asking if I did the 3 x AD844 and it's buffers stack yet.
Yes I got to do it, I've only listened to 1 track and was very pleased with what happened, the speakers now may as well be in another state, such was the sound it gave.
Just holographic is the word that came to mind, nothing saying to my brain and eyes that the sound was coming from the speakers, they just vanished
I've ask Art Vandelay seeing he's just up the road for us to do a 2x (his) vs my 3x before he changes his to 3, on his B&W 800d's. As we both have the PCM1704 based Cary 303/200 cdp's
I will get back again with more descriptive differences between the two, after we've done the A/B.
Cheers George
Just a little hint that may make the sound a bit better. Make sure the power supplies +-5vdc for the PCM1704s are equally loaded. if you have "super regulators" feeding the 1704s then the equal output impedance of the power supplies is not in issue. I am using dedicated emitter followers for the 1704s. Since the 1704s are asymmetrical load, that is the negative supplies is X6 the positive load, I have loaded the positive supply to match the load on the negative supply. Thus equal output impedance. Emitter follower supply output impedance is affect by the loading level. If you are tinkering with you player you might try checking to see if the source for these supplies are load balanced even if they are 3pin regulators to insure that the output impedance is fairly close. Most units will not have dedicated supplies so your YMMV. I don't know how closely the match should be, but try it if you have the inclination.
Yes optimationman, Jocko Homo also found this out.
"They draw a ton of current, on the (-) rail. Someone who used to work there 'splained there are actually several DACs, operating in parallel, but slightly time-shifted. They had to do that, to get the settling time down to where they wanted it."
If anyone tries to load the + rail the same as the - rail , please let us know of if anything is heard for the better when done?
Cheers George
"They draw a ton of current, on the (-) rail. Someone who used to work there 'splained there are actually several DACs, operating in parallel, but slightly time-shifted. They had to do that, to get the settling time down to where they wanted it."
If anyone tries to load the + rail the same as the - rail , please let us know of if anything is heard for the better when done?
Cheers George
Attachments
Last edited:
Have a working 1704 now...
Hi George, Happy to report that I am now listening to a PCM1704 based dac. Well for 3 days to be exact. The dac consist of a WM8804 receiver, DF1704 digital filter and a pair of PCM1704's. In the short term I am using an I/V based on a video opamp however it is not an AD844. 7th order GIC Bessel filter with an OPA627 output buffer. Getting very good results. The dac offset measured +0.1 mV and -0.1 mV's. I have the J grade. I will need to rework the triple stack to try it. Just a matter of time. I can clearly hear why people like this dac. I actually hear some similarities to the TDA1541 family. Hopefully I will get that delicate treble that I like with the 1541. I think it is just a matter of I/V and output stage. Cheers
Hopefully I will get that delicate treble that I like with the 1541. I think it is just a matter of I/V and output stage. Cheers
You will get that and more once you've implemented just the AD844 stack, and get rid of the I/V based video opamp, 7th order GIC Bessel filter and OPA627 output buffer.
Cheers George
Hi George, Happy to report that I am now listening to a PCM1704 based dac. Well for 3 days to be exact. The dac consist of a WM8804 receiver, DF1704 digital filter and a pair of PCM1704's. In the short term I am using an I/V based on a video opamp however it is not an AD844. 7th order GIC Bessel filter with an OPA627 output buffer. Getting very good results. The dac offset measured +0.1 mV and -0.1 mV's. I have the J grade. I will need to rework the triple stack to try it. Just a matter of time. I can clearly hear why people like this dac. I actually hear some similarities to the TDA1541 family. Hopefully I will get that delicate treble that I like with the 1541. I think it is just a matter of I/V and output stage. Cheers
Hi,
Could the AD1862 (20 bits) a better bet for the tones your are speaking about
! (ok I don't talk about the pcm63 not sourcable anymore without knowing if it's a fake or second choice) !Ad1862 can be found at Rochester, and I saw DIYhink sells a little dac with those two dac chip (mono like the pcm1704)... but at this price I ask myself where they sourced the chips
! Frankly with a tda1541, I don't see where the 16 bits can be a limitation but on the paper !
AD1862
Hi Eldam, Don't know if it is a bits related thing. Installed some Polypropylene decoupling caps on the dac board as seen on the PCM1704 datasheet. My dac build is based on the DF1704, PCM1704 evaluation board. With the added decoupling I am getting close with the treble I am looking for. I have some NOS AD1860's I will try some time in the future. AD1862 as you mention are 20 bit. The fake thing is really out of hand. Probably more then just Chinese scam artist. I picked up a few TDA1541A's from a Swedish friend that are used pulls. Probably the best way to get the real thing today. I am totally gun shy where purchasing anything off Ebay from China is concerned. When you consider that they fake Wima capacitors, it isn't hard to assume that the dac you mention is mostly fake so the price probably reflects that. A sad state of affairs....Hi,
Could the AD1862 (20 bits) a better bet for the tones your are speaking about
Ad1862 can be found at Rochester, and I saw DIYhink sells a little dac with those two dac chip (mono like the pcm1704)... but at this price I ask myself where they sourced the chips
Frankly with a tda1541, I don't see where the 16 bits can be a limitation but on the paper !
Triple stack
Hi George, After some additional decoupling with Polypropylene caps I am getting close. The video opamp I refer to is the best transimpedance chip I have tried for standard feedback I/V. I will need to get some parts on order to adjust the AD844 stack from TDA1541 to PCM1704. I forgot to order the 20K pots for nulling. It is possible the stack will get me that last bit of quality I am looking for in the high frequencies.
Hi George, The video opamp I refer to is the best transimpedance chip I have tried for standard feedback I/V.
What I found that for I/V duties any opamp no matter how good, if they have feedback networks around them, just can't cut it when compared to the zero global feedback AD844.
Someone please correct me if I'm wrong, but I think that the extreme HF rubbish and glitches that comes out of dacs somehow have adverse effects on I/V stages that have feedback around them, sending them maybe into spurious oscillations or ringing? Where the AD844 without feedback remains more stable with these nasties coming into it.
Sort of like amps tube or solid state which become more stable when feedback is greatly reduced (or removed in tubes).
Cheers George
Last edited:
Hi Abraxalito, thanks for chiming in, I thought it would be very hard or even impossible to duplicate the rubbish that comes out of a dac to monitor and see what happens to feedback I/V's, this is why I queried it.
Cheers George