A few week have passed now, and I've decided there’s something not quite right with the sound of my roksan cd player op-amp replacement. I replaced the original NE5532 op-amp ICs with LM6172IN op-amps. I replaced the original power supply bypassing capacitors with 0.01uf silver mica and 2.2uf polyester as specified in the LM6172 data sheet. This was all the adjustments I did prior to installing the op-amps.
Once installed I checked for any oscillation and there was non, when bypassing the output capacitors I got around 150mV of DC offset on both ICs at the rca's but I thought this may settle down after running it for a bit, and the output capacitor would eliminate this dc offset. The input offset is less than 1mV.
the sound of the player hasn’t really improved much, its rather grainy and can be quite fatiguing at times with a slight imbalance in sound, symbols sound sibalent and sandy. and the DC offset has raised to 200mV is this to much offset?
Any help would be most appreciated.
Here’s my original postings when I set out to do the clock and op-amp mods. Elso's clock mod did the trick
http://www.diyaudio.com/forums/showthread.php?s=&threadid=23023&highlight=kandy+clock
Hesky
Once installed I checked for any oscillation and there was non, when bypassing the output capacitors I got around 150mV of DC offset on both ICs at the rca's but I thought this may settle down after running it for a bit, and the output capacitor would eliminate this dc offset. The input offset is less than 1mV.
the sound of the player hasn’t really improved much, its rather grainy and can be quite fatiguing at times with a slight imbalance in sound, symbols sound sibalent and sandy. and the DC offset has raised to 200mV is this to much offset?
Any help would be most appreciated.
Here’s my original postings when I set out to do the clock and op-amp mods. Elso's clock mod did the trick
http://www.diyaudio.com/forums/showthread.php?s=&threadid=23023&highlight=kandy+clock
Hesky
"I replaced the original power supply bypassing capacitors with 0.01uf silver mica and 2.2uf polyester as specified"
The 7812/7912 may need the big high-R elcaps that were in place before. They may oscillate or at least have a noise peak close to the audio band when they are loaded only with a low-value low-R film cap.
"around 150mV of DC offset on both ICs at the rca's but I thought this may settle down"
Such things do not settle down. Please forget the myth that time and running cure all (or anything) in electronics.
"and the output capacitor would eliminate"
Sure. But I thought you bypassed that cap already. With what, actually? If a piece of wire, then remove if you do not succeed in getting DC offset below 10mV. If another cap then investigate closely, as output DC must be 0 then. Unless the cap is leaky, or unless there is not resistor between the output hot and ground.
Check.
And may I suggest the AD826 as a suitable though less-temperamental dual opamp for digital audio applications?
The 7812/7912 may need the big high-R elcaps that were in place before. They may oscillate or at least have a noise peak close to the audio band when they are loaded only with a low-value low-R film cap.
"around 150mV of DC offset on both ICs at the rca's but I thought this may settle down"
Such things do not settle down. Please forget the myth that time and running cure all (or anything) in electronics.
"and the output capacitor would eliminate"
Sure. But I thought you bypassed that cap already. With what, actually? If a piece of wire, then remove if you do not succeed in getting DC offset below 10mV. If another cap then investigate closely, as output DC must be 0 then. Unless the cap is leaky, or unless there is not resistor between the output hot and ground.
Check.
And may I suggest the AD826 as a suitable though less-temperamental dual opamp for digital audio applications?
What you've described is very similar to my experience with the LM6172. I've tried it on a couple of cd players, a headphone amp and a preamp, and found it challenging. The highish output offset is expect (bipolar, higher input bias current than jfets), but an output dc block should control this to 0V..
Something like the BB OPA2132 or OPA2134 should be easier to work with in your case.
Recently I have returned to the LM6172 in two cases -- a Marantz cd upgrade, and phono amp. Perhaps I've learnt more about building clean supplies or supplies that match the 6172 better, but I'm get the clean, detailed open sound expected of the 6172, without the nasties. I guess it's just a tough opamp to get just right.
Something like the BB OPA2132 or OPA2134 should be easier to work with in your case.
Recently I have returned to the LM6172 in two cases -- a Marantz cd upgrade, and phono amp. Perhaps I've learnt more about building clean supplies or supplies that match the 6172 better, but I'm get the clean, detailed open sound expected of the 6172, without the nasties. I guess it's just a tough opamp to get just right.
Lm6172
Hi Hesky,
The LM6172 is way too fast for my taste. I am currently using the dual OP275 as an IV-converter and I am very pleased with it. Before I was using the OPA2604.😎
Hi Hesky,
The LM6172 is way too fast for my taste. I am currently using the dual OP275 as an IV-converter and I am very pleased with it. Before I was using the OPA2604.😎
Go ahead......hate me.....
Why a silver mica/polyester cap combo? Probably the last thing that I would have tried.
Judging by your comments......the sound.....the DC offset........the speed of the op-amps........your cap choices.....
Have you checked for high frequency oscillations?
Jocko
Why a silver mica/polyester cap combo? Probably the last thing that I would have tried.
Judging by your comments......the sound.....the DC offset........the speed of the op-amps........your cap choices.....
Have you checked for high frequency oscillations?
Jocko
Thanks for all your help everyone,
My original intension was to replace the NE5532 op-amps with the AD8065 jfets, but I was recommended to replace them with the LM6172, I liked the idea of keeping it bipolar. I new that the LM6172 was a cranky op-amp and was hard to implement but I hopped it would work out ok. The output capacitor was never bypassed as its still has no where near 0V dc offset, this was just to test the amount offset that was being produced.
Why is the choice of mica/polyester cap incorrect? is it because they have a very low esr value, so with the addition of an electrolytic will this help?
Hesky
My original intension was to replace the NE5532 op-amps with the AD8065 jfets, but I was recommended to replace them with the LM6172, I liked the idea of keeping it bipolar. I new that the LM6172 was a cranky op-amp and was hard to implement but I hopped it would work out ok. The output capacitor was never bypassed as its still has no where near 0V dc offset, this was just to test the amount offset that was being produced.
Why is the choice of mica/polyester cap incorrect? is it because they have a very low esr value, so with the addition of an electrolytic will this help?
Hesky
If you check the data sheets for both opamps, you'll find the LM6172 has a "much" greater input bias current, which probably has something to do with your DC offset dilemma. Perhaps sticking with a jfet opamp may improve your situation.
- Status
- Not open for further replies.