jaudio said:
I always like to test with a 'real world' input impedance.
If you will have a volume pot before the pre, then test with a volume pot (of the same impedance you are going to use), and at several positions, from min no max.
If it works, it works.
If it only works with the inputs shouted, it's not optimal.
A DC servo should always work, removing DC offset from the output, anyway.
Btw if you can detect differences in sound just by using higher value resistors (and lower value film caps), it tells a lot of how a DC servo can impact the audio performance. After all, it is in the signal path, and it is more feedback.
It has to be optimized, as everything.
I’m sorry I didn’t explain myself clearly. I put all the components for the servo circuit on a board and I was wondering if I could test it without connecting it to the preamp. I would like to check for mistakes.
Hi Jaudio
you will have to connect it to the preamp. it works by checking the voltage on the output and compensating for it on the input. (just dont connect the preamp to you power amp)
you will have to connect it to the preamp. it works by checking the voltage on the output and compensating for it on the input. (just dont connect the preamp to you power amp)
safetyman said:rudi, how did you mount the heatsink on to the AD815 chip? BTW, which version of the LM317 and LM337 is the correct one. There seems to be many, with different letters behind them?
rudi, thanks for the encouragement... how the hell am I going to attempt what you did when I don't even know what is a DC servo? 😀 When I can get someone to help me make boards, I am going to try your X-Caliber... that's for sure. 😉
The heatsink is mounted with a spring clip that mounts through the board. BUT i found that is not the right way to do it. if you allow enough copper around the heat tabs it actually runs cooler. The heat tabs should run directly into a large area of copper without any narrowing of the copper before it gets to a large area.
Regarding the LM317 and 337, just make sure it is a TO220 package then you will be OK.
The X-Calibre and the AD815 is a BEUTIFULL combination. I left the thread to die off due to animosities (I prefer to keep this beauty for myself and for the people that would appreciate the true value of this adopted redesign), but if you need the latest PCB layout, just mail me.
carlosfm said:
Carlos I agree with you (eventually 😉 ) I have to add the differences were very subtle maybe the diff between two high end polyprop caps 🙄 but I had to make a call on this for myself.
I finished the OPA2134 version last night. And it sounds good. I think the caps needs some settling but it sounds the same as the two OPA627 version (maybe just a touch cleaner but that could be due to the fact that I used different caps this time round)
rudi said:
Rudi, bigger caps may have less impact, as they will act at a lower frequency, close to DC.
To use good film caps (say, up to 1~2.2uF) the resistors must be some hundreds of K. Also, it's a good idea to divide the output resistor (that goes to the input of the pre) in two resistors and put a cap to ground after the first one.
Opamps have low PSRR at high frequencies, so the servo is injecting high frequency noise to the input of the pre. It must be filtered. The problems (and solutions) are the same as when injecting PSU voltage.
Btw exactly what values are you using now for the cap and resistors?
Hi Carlos. I used a 820K with a 1uf but i went back to my 23.5uf with the 33K. these values did work better . I left the 10K as is
but i will wait to hear what values you come up with and try it out.
btw. is the values of the caps after the regulator essential? what combinations will work just was well.
but i will wait to hear what values you come up with and try it out.
btw. is the values of the caps after the regulator essential? what combinations will work just was well.
rudi said:
Rudi, I think the servo is not a bad idea (although I was reluctant to use it at the time), I will try it.
I just think it must be done with care.
Unfortunately I don't have the time right now, I'm just starting on a discrete phono pre, and I also have a CDP here waiting for a clock, etc...
Also, my pre is working fine... 😀
You are using a previous active circuit before the AD815, it's different from what I have (a volume pot) so naturally the dc null circuit I made doesn't work as well for you.
If you are in a hurry to finish that pre, I can't help you right now.
rudi said:
What caps?
Carlos, I see you also have to much hay on your fork aswell (i am renovating my house and there is dust everywhere)
I was actaully more interrested to hear what values you would suggest for the servo and also cap value on the output. can you do a quick schematic. The guy that designed this for me is a very competant electronic enigeer. so he did various sims with the AD815 and OPA627 and OPA2134 spice models to derive at what he thought to be optimum and it works very well but i am also interrested to see how you would do it.
the caps i am refereing to is the 33uf after the regulator followed by the 68uf. I used 47uf + 47uf (all i could get in SMD) but the bass were a bit fuzzy and then i changed the 47uf back to a 33uf as per your schematic and it improved the detail of the bass (this little change made a much bigger difference than other values on the dc servo) but all of this could be a artifact of my design and layout
I was actaully more interrested to hear what values you would suggest for the servo and also cap value on the output. can you do a quick schematic. The guy that designed this for me is a very competant electronic enigeer. so he did various sims with the AD815 and OPA627 and OPA2134 spice models to derive at what he thought to be optimum and it works very well but i am also interrested to see how you would do it.
the caps i am refereing to is the 33uf after the regulator followed by the 68uf. I used 47uf + 47uf (all i could get in SMD) but the bass were a bit fuzzy and then i changed the 47uf back to a 33uf as per your schematic and it improved the detail of the bass (this little change made a much bigger difference than other values on the dc servo) but all of this could be a artifact of my design and layout
rudi said:
When I have the time, I will make some tests and report, not now...
Last light I was crawling through the floor listening to Neil Young's 'MTV Unplugged' on vinyl, with my current phono pre and my Rega Elys cartridge. Someone has told me I could have better performance with a very simple discrete phono pre, so that's where my interest is right now.
I wanna see if it's possible that my crawling improves.😀
The line preamp is fine, thank you.
rudi said:
It has always been my experience that small caps right after the regs (78/9xx, LM317/37, LM338, etc) work better.
I did use 33uF, didn't I? Shame on you, you didn't follow my schematic. 😉
Actually, it can be smaller.
But the 68uF cap on the AD815's pins is ok, and it can be higher, say, up to 100uF.
jaudio said:
I have suggested on several posts, but I can't give exact values without trying.
When I'm sure what works best I will share it, what's the use of suggesting something now?
You guys make some research about servos and make your tests, if you are interested going that way.
For my part, I'll give news when I have them.
carlosfm said:
Carlos
I know this is quite a few posts back but I have a question about your tape buffer.
If you can remember what was the resistor value for biasing the opa627?
Thanks
jaudio i assume that you have a AD815 working. if so you can test it directly on the module, that is to say if you build it according to my schematic. if so then you will be ok and it should work first time
Would a single 160VAC, 0-18, 0-18 torodial be enough to power both channels or do you need one for each channel?
- Status
- Not open for further replies.