I have a dac (circuit below) that I feel the need to tinker with again. I have replace the 2 x NE5532's with 4 x AD825 and done some direct coupling mods with dc offset adjusters in yellow highlight.
I've just recieved 4 samples of TI's THS4031's.
With the ad825's having very low dc offset and fet input, the trimmers only have to be adjusted very slightly off the replaced resistor value of 2.2k to give me zero offset at the dc coupled output. The sound is very good this way.
What do you guys think the difference would be with the THS4031's replacing the AD825's, they are bipolar input does anyone know if the THS4031's have low dc offset say below 20mv and more important would they sound better than the AD825's in this position, I've noticed that a lot of people here love them and a couple have prefered them over the AD825's in the ouput stage of their players/dacs.
Cheers George
I've just recieved 4 samples of TI's THS4031's.
With the ad825's having very low dc offset and fet input, the trimmers only have to be adjusted very slightly off the replaced resistor value of 2.2k to give me zero offset at the dc coupled output. The sound is very good this way.
What do you guys think the difference would be with the THS4031's replacing the AD825's, they are bipolar input does anyone know if the THS4031's have low dc offset say below 20mv and more important would they sound better than the AD825's in this position, I've noticed that a lot of people here love them and a couple have prefered them over the AD825's in the ouput stage of their players/dacs.
Cheers George
Attachments
Got the 4x THS4031 's in yesterday put them in the I/V stage and only had to adjust the 5k 25 turn trimmer a half a turn to get zero offset on the output, not bad, this opamp has good low dc offset for a bipolar input opamp, put it back in the system and wow!!! the mids and highs are so so much more transpaent than the AD825 's and it made mince meat out of the NE5532 's, but the bass it was a little anaemic, no balls compared to the AD825 's, but this was their forte.
So instead of getting on the merry go round I left the THS4031 's in overnight and left it turned on. What greated me the next day was a revelation, gone was the anaemic bass, it was even better than the AD825 's now, and the mids and highs were just as good as before even a bit better maybe. These bloody things like to get used to their suroundings, before they give their best.
I can now say that the THS4031 is the best for I/V conversion that I have tried to date, give them a go you will be surprised, but don't listen to them till their 24hrs old.
Cheers George
So instead of getting on the merry go round I left the THS4031 's in overnight and left it turned on. What greated me the next day was a revelation, gone was the anaemic bass, it was even better than the AD825 's now, and the mids and highs were just as good as before even a bit better maybe. These bloody things like to get used to their suroundings, before they give their best.
I can now say that the THS4031 is the best for I/V conversion that I have tried to date, give them a go you will be surprised, but don't listen to them till their 24hrs old.
Cheers George
Hi George,
Why do you follow your opamp with a pushpull amplifier?
Is it something to do with the 5532? Or the DAC output levels (I'm not very familiar with the Burr Brown DACs)? Or are you simply outputting a bigger signal?
If its just buffering? What does it sound like without it?
My DAC's output is straight from the LM4562 output into the preamp.
Cheers,
Phil
Why do you follow your opamp with a pushpull amplifier?
Is it something to do with the 5532? Or the DAC output levels (I'm not very familiar with the Burr Brown DACs)? Or are you simply outputting a bigger signal?
If its just buffering? What does it sound like without it?
My DAC's output is straight from the LM4562 output into the preamp.
Cheers,
Phil
Phil most dacs have an output opamp after the I/V opamp stage, my is just a discrete amp after the I/V opamp.
As good as opamps are for output stages, I would still rather have a discrete amp, this one is biased up strong into class A and it can output up to 6 watts, not that it ever does, and sounds better than anything I've heard. It's a Musical Fidelity A3.24, exactly the same as the MF Nuvister Dac with discrete solid state output instead of nuvister tubes.
And I've direct coupled it with some mods, you can find all my circuits here for the dac.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=57887
Cheers George
As good as opamps are for output stages, I would still rather have a discrete amp, this one is biased up strong into class A and it can output up to 6 watts, not that it ever does, and sounds better than anything I've heard. It's a Musical Fidelity A3.24, exactly the same as the MF Nuvister Dac with discrete solid state output instead of nuvister tubes.
And I've direct coupled it with some mods, you can find all my circuits here for the dac.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=57887
Cheers George
janneman said:
Ask your countryman Gudio Tent, he sent me an email saying he listened to one that he was doing some work on, and he was impressed with it, also. I think it's still better to have common disrete amp than a chipamp (opamp).
"arguably" is the correct word, as I've stated, I've had many different dac's come through my place, and nothing has touched this one yet.
Straight out of the box it had something very special, very powerful detailed sound without a hint of hardness, all the I/V mods and direct coupling have done is given it more transparency and a bottomless pit of detailed tight bass. Not that it means anything but it got Class A rating in Stereophile
http://stereophile.com/digitalprocessors/561/
But I do read their measurments, and they do a good job of them, and this dac had the lowest measured jitter of just about any dac they have measured.
Cheers George
Here is a link to the Burr-Brown PCM1738e dac chip( E I was told stands for selected), current output (2.5mA), that is in my MF converter, for those who want to read about it.
It does not get mentioned much in these forums, but it does a very good job. It made a fully optimized Marantz CD94mkII with double TDA1541 S1 double crowns sound dirty and flat.
And last month blew away the new Belcanto Dac 3 that I had here for evaluation, made it sound thin in the mids and up top, and gutless down low.
http://focus.ti.com/lit/ds/symlink/pcm1738.pdf
Cheers George
It does not get mentioned much in these forums, but it does a very good job. It made a fully optimized Marantz CD94mkII with double TDA1541 S1 double crowns sound dirty and flat.
And last month blew away the new Belcanto Dac 3 that I had here for evaluation, made it sound thin in the mids and up top, and gutless down low.
http://focus.ti.com/lit/ds/symlink/pcm1738.pdf
Cheers George
Just have been looking at the THS4031/2 datasheets. Strange thing is the dual 4032 has 6 - 7dB higher 2nd order harmonic distorsion. A double THS4031 on a browndog would be better for dual opamp applications then, or keep it 4032 when you like 2nd harmonic... But this 2nd harm. declares problably the "involving" sound i perceive with this opamp.
Another nice feature of 4031, between pin 1 and 8 a DC nulling pot can be added, wiper on minus PS. A dc coupled I/V can be arranged with this.
Listened a few days to the THS4032, i have a feeling it is perfect for I/V, but the buffer after it would be nice with a (single version?) LM4562, as analog filter a GIC filter with LM4562.
Another nice feature of 4031, between pin 1 and 8 a DC nulling pot can be added, wiper on minus PS. A dc coupled I/V can be arranged with this.
Listened a few days to the THS4032, i have a feeling it is perfect for I/V, but the buffer after it would be nice with a (single version?) LM4562, as analog filter a GIC filter with LM4562.
It;s called the LME49710.
I swim against popular opinion in that I don't think it works very well for purposes like LPF after dacs (and I've worked quite hard at it with some very refined PSus and other signal conditioning). You get great detail but it seems to turn great pianists into ...mere typists. No passion; something quite wrong with the resulting presentation that I can't quite articulate - seems like LF ambience comes and goes depending on signal level, I think. Definitely NOT one to make me pull disc after disc after disc...late into the night.
Hi Martin Clark
Thanks for info.
The fact is when i did a non-os mod on a tda1541 and had build a 4 transistor I/V after it, i didn't like a 3rd order filter either.
The current CD304mk2 cdplayer layout has a filter and i keep it that way. But i am considering to re-try a PCM56 non-os dac with glue logic on I2S and give the fine single LME a try. I used previously Pedja's AD844 I/V for that. Very clear sound.
The THS can sound some non-os like(in the positive way), even with digital filter still in place.
Tubee,I noticed on another thread that you wanted to run yourTHS into class A some more, are you running the 4031 or the 4032 as I have the 4031's x4 and their rails in my setup are at 15+ and 15-
I would not go anymore class A than they already are at with these rails, as I can only put my finger on them for 3-4 seconds before it get too hot (around 60c) running completely stable. They have a max temp junction temp of 125c which the case would be around 80c but that is too severe for me, so any more class A cold put them over the top. You could heatsink them but with mine single 4031 on brown dogs one is underneath and would be impossible to heatsink.
Cheers George
I would not go anymore class A than they already are at with these rails, as I can only put my finger on them for 3-4 seconds before it get too hot (around 60c) running completely stable. They have a max temp junction temp of 125c which the case would be around 80c but that is too severe for me, so any more class A cold put them over the top. You could heatsink them but with mine single 4031 on brown dogs one is underneath and would be impossible to heatsink.
Cheers George
Hi George
You mis nothing isn't it
I didn't check if they where getting hot, will palpate the chips soon.
I think, not sure, they run on +/- 12V.
Btw they are also deliverable in DGN version with a powerpad underneath in chip, then possble to dissipate 1800mW. The regular version only 629mW. Maybe this can solve some heatproblems. A adaption on the browndog has to be made then too.
I soldered the chips on a 8 pin socket, if they're getting hot will glue a small copper chunk under the chip and glue the chip's body on it. I think a class-A bias could benefit in sound of the THS.
You mis nothing isn't it
I didn't check if they where getting hot, will palpate the chips soon.
I think, not sure, they run on +/- 12V.
Btw they are also deliverable in DGN version with a powerpad underneath in chip, then possble to dissipate 1800mW. The regular version only 629mW. Maybe this can solve some heatproblems. A adaption on the browndog has to be made then too.
I soldered the chips on a 8 pin socket, if they're getting hot will glue a small copper chunk under the chip and glue the chip's body on it. I think a class-A bias could benefit in sound of the THS.
tubee said:
Mine are running stable, as I said in post 14, no, I think they are biased up into Class A quite high already and don't need anymore, this is why they sound so sweet and transparent 3D in the highs, and mine are running maximum rail voltages as well this also would add to it. Each amp is running @ 10-11mA where the AD825 was 7mA and it is a hot running chip.
Cheers George
Ok thats settled then George.
Maybe the powerpad might be a solution for temp problems, combined with heatsinking.
Diy heatsink: Bending all soic pins slightly down +/- 0.3mm, (keep solder surface in line with pcb surface) and glue a 3.5 cm lenght (or so) and 2.5mm broad strip of 0.2mm thick copper underneath chip. After soldering chip, the copper can be folded up above chip for heat radiation.
Glueing is best done with 2 component glue (araldite??) wich can have some heat. Cyanoacrylate (1 second glue) can be used too, but i am not sure if it can withstand the temperature.
Btw i had to bend all pins to adapt the chip on a dil socket. I earthed myself a few times for ESD safety then.
Maybe the powerpad might be a solution for temp problems, combined with heatsinking.
Diy heatsink: Bending all soic pins slightly down +/- 0.3mm, (keep solder surface in line with pcb surface) and glue a 3.5 cm lenght (or so) and 2.5mm broad strip of 0.2mm thick copper underneath chip. After soldering chip, the copper can be folded up above chip for heat radiation.
Glueing is best done with 2 component glue (araldite??) wich can have some heat. Cyanoacrylate (1 second glue) can be used too, but i am not sure if it can withstand the temperature.
Btw i had to bend all pins to adapt the chip on a dil socket. I earthed myself a few times for ESD safety then.
I've now been listening to the THS4031 for over a week and I have to say that it is the most cleanest sound, especially through the mids and highs I have ever heard of the multitude of opamps I've tried, which is every one except the hidiously expensive OPA627, this could give it a run from what I've read, but at $60 a pop in Australia, you would have to be mad to buy it, and I need 4, and I've heard it's a bit too laid back for my tastes.
Now that I've told you about the good, let me tell you the bad, and it's a big one.
The THS4031 has one glaring problem compared to my previous favourite opamp the AD825, it does not have the rhythm of the AD825, and that is a major fault to me, it sounds so good that you forget your not tapping your foot, yet now and again you do get a bit of foot tapping rhythm come through but most of the time it sounds constricted in that department.
Back to the AD825 and your feel that the rhythm beast has been unleashed, even though it's not quite as good in the mids and highs it is still my favourite.
Conclusion, I hate opamps, there's not one that does everything right, I'm now starting to look at the ready made discrete type of I/V stages, and that is another world of hurt and anguish.
Cheers George
Now that I've told you about the good, let me tell you the bad, and it's a big one.
The THS4031 has one glaring problem compared to my previous favourite opamp the AD825, it does not have the rhythm of the AD825, and that is a major fault to me, it sounds so good that you forget your not tapping your foot, yet now and again you do get a bit of foot tapping rhythm come through but most of the time it sounds constricted in that department.
Back to the AD825 and your feel that the rhythm beast has been unleashed, even though it's not quite as good in the mids and highs it is still my favourite.
Conclusion, I hate opamps, there's not one that does everything right, I'm now starting to look at the ready made discrete type of I/V stages, and that is another world of hurt and anguish.
Cheers George
Yes, AD825 has some problems, it sounded the same as you say for me...
Because the IV is a very critical position. On each sample transition, the DAC basically outputs a step between the old and the new sample with a very fast slew rate and settling time.
Old crud like NE5532 is certain to go into slew limiting, then, and fast opamps will sound better.
In the end I used a folded cascode...
janneman said:
Because the IV is a very critical position. On each sample transition, the DAC basically outputs a step between the old and the new sample with a very fast slew rate and settling time.
Old crud like NE5532 is certain to go into slew limiting, then, and fast opamps will sound better.
In the end I used a folded cascode...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.