The input stage based on the Dan schemo... PCB will be on top of the 1632 PCB
Any opinion about OPA1632 for D/A output amp? ???
Any opinion about OPA1632 for D/A output amp? ???
DCX2496 audio mod
- Thread starter Moby
- Start date
Help Support GroupDIY Audio Forum:
OK than 
Let's say that Input board is solved with few options to try and measure in the real life. What about output DA section?
I tried to simulate stock schemo just to check filter response and here is what I get. First stage as Dan notice looks correct (I suppose that it can be better with some good receiver THAT or INA) but filter creates some phase shift from 10k . Also, I'm confused with crossover produced by Transistor mute. Maybe that can be sim bug but... Also It seems that rest of topology lack of headroom. I skipped the balancing stage because it's obvious that signal is already degraded before. Any opinions about -3db point of the LP filter? Is there any better option? What about the rest of circuit?
Let's say that Input board is solved with few options to try and measure in the real life. What about output DA section?I tried to simulate stock schemo just to check filter response and here is what I get. First stage as Dan notice looks correct (I suppose that it can be better with some good receiver THAT or INA) but filter creates some phase shift from 10k . Also, I'm confused with crossover produced by Transistor mute. Maybe that can be sim bug but... Also It seems that rest of topology lack of headroom. I skipped the balancing stage because it's obvious that signal is already degraded before. Any opinions about -3db point of the LP filter? Is there any better option? What about the rest of circuit?
Polite bump
Very busy at work....
V1 and V2 are +-5V, try about +-15V, it will work much better.
The -3db point could stand to be a little higher.
I would check the phase response is still Ok down at 20Hz as well, bottom end problems due to phase shifts are not exactly a new game.
Regards, Dan.
V1 and V2 are +-5V, try about +-15V, it will work much better.
The -3db point could stand to be a little higher.
I would check the phase response is still Ok down at 20Hz as well, bottom end problems due to phase shifts are not exactly a new game.
Regards, Dan.
Thanks Dan, my mistake ;D I definitely need a second pair of eyes . My comments about crossover and headroom are out of question now but I'm still looking about improving this stage. Cool about moving -3db point , changing tr mute to relay, what can I do more?
. My comments about crossover and headroom are out of question now but I'm still looking about improving this stage. Cool about moving -3db point , changing tr mute to relay, what can I do more?
What you think about stage like this?
![Soldering Iron Kit, 120W LED Digital Advanced Solder Iron Soldering Gun kit, 110V Welding Tools, Smart Temperature Control [356℉-932℉], Extra 5pcs Tips, Auto Sleep, Temp Calibration, Orange](https://m.media-amazon.com/images/I/51sFKu9SdeL._SL500_.jpg)
I don't think that works correctly...
In particular it does not attenuate common mode noise particularly well at high frequency, and will pass quite a lot of common mode noise to the next stage.
Try connecting the signal source between ground and both input terminals (with the inputs wired together) and then running a sweep from say DC to 2Mhz to see what I mean.
I tend to feel that unless you are going off module (like say a mix bus or external IO), it is normally best to convert to single ended early on as it makes component matching rather less critical, then convert back to balanced if needed.
Even if that filter was right (and I don't think it is, but I have not simulated it), you are asking a lot of the component matching.
Also, is the OPA604 really capable of driving into a 10 ohm buildout?
Regards, Dan.
In particular it does not attenuate common mode noise particularly well at high frequency, and will pass quite a lot of common mode noise to the next stage.
Try connecting the signal source between ground and both input terminals (with the inputs wired together) and then running a sweep from say DC to 2Mhz to see what I mean.
I tend to feel that unless you are going off module (like say a mix bus or external IO), it is normally best to convert to single ended early on as it makes component matching rather less critical, then convert back to balanced if needed.
Even if that filter was right (and I don't think it is, but I have not simulated it), you are asking a lot of the component matching.
Also, is the OPA604 really capable of driving into a 10 ohm buildout?
Regards, Dan.
Clear.. Well it works somehow but since I'm not sure how I will stick to your recommendations. Than, forget about previous schemo, and please tell me (if you find some time) what can I do about this? Of course , all comments are welcome. BTW, I found the way to purchase THAT chips so it's welcome too
Well it works somehow but since I'm not sure how I will stick to your recommendations. Than, forget about previous schemo, and please tell me (if you find some time) what can I do about this? Of course , all comments are welcome. BTW, I found the way to purchase THAT chips so it's welcome too
dmills said:
I think that I'm stuck in a mud with this. Especially with DAC output stage. I searched a lot about this but besides some minimalistic "audiophool" approaches like TX output or similar to Bhrngrr stuff I can't find something appropriate. Or I can't find it :-[ Well, TX output is really cool for some purposes but since I'm looking for something transparent and clean I think that is out of question. I believe that first behringrr stage looks OK but is there a chance for something "better"? You mentioned THATcorp chips like something welcome in this design. Can that be BB INA/DRV ? And hoe to implement that recievers-drivers in this situation? I don't see any DAC design (on the web) with that . Can I put INA134 in the place of 1 opamp (stock schemo) and expect some improvement?
Really stuck :-\
Ok, I'm still trying ;D how about this?
