I've been studying the circuit with directly mounted decoupling caps in mind. Looking at where the ground traces go I noticed the decoupling caps ground to the 0v (nice thick trace) on the 17v trafo secondary, and eventually the star point on the main smoothing cap. The GND pin (pin 7) of the LM3886 connects to the star point on the main smoothing cap - which connects to the 0v on the 23v winding. The Ground trace from pin 7 on the power chip seems to go via all the main areas of the circuit - Opamps and input selector, phono sockets, etc.
Is this done to get improved decoupling on the power chips?
My consideration is that if I hardwire the decoupling caps then I have a longer path to ground which has a lot of vias along the way.
What would be the best way to hardwire my 100uF 50v BG STD's in the above circuit?
Mike.
Hi Mike,
Sorry I haven't been able to respond earlier.
Hope that helps, Chris
Sorry I haven't been able to respond earlier.
This is normal for most audio equipment. Historically too. You can reduce the range by turning the gain stage into a nice buffer. I'm not sure if the gain reduction affects the actual PSRR, but it won't be amplified as much. I guess that's what Greg is getting at. A better buffer and supply regulation / filtering will be of more use to get rid of supply noise.
Yes. This circuit will detect a DC voltage on your output lines. At some voltage, it will act by shutting down the output. This circuit will not affect your amp in normal operation. Therefore, leave it in. It's never a good idea to bypass protection circuits. Even an internally protected chip amp can fail DC. That wasn't very hard to explain.
Yes.
You may make things worse, depending on how well that transformer is shielded and how you dress the AC leads to it. To be honest with you, all you really need is good regulation and filtering between the digital and analog supplies. Again, try to look at all the options and the effects of each. Another transformer sounds impressive and may give you the idea that the supplies are totally isolated, but that is not necessarily true.
Decoupling capacitors are mounted near the "load". Any trace length or vias in between the load and power supply are the reason you are using decoupling caps. Some smaller capacitors will be found (or should be) near the rectifiers to reduce the amount of higher frequency noise before it gets into other circuits. Always deal with noise at the source. So, smaller capacitors can be decoupling capacitors or high frequency filter capacitors, depending on where they are located physically.
Hope that helps, Chris
Thanks Chris.
I dont fancy redesigning the input buffer so I'll give Gregs idea a try. Besides, with a LM4562 the amp sounds much better than with the NE5532. And later I'll be adding low noise discrete regs.
With seperating the microprcespor supply how about fedding the new trafo via one of those Shaffener filters? Im also considering adding a dedicated trafo for the preamp which might also benefit from one of those filters. I would shield any AC supply lines also.
What do you think?
I dont fancy redesigning the input buffer so I'll give Gregs idea a try. Besides, with a LM4562 the amp sounds much better than with the NE5532. And later I'll be adding low noise discrete regs.
With seperating the microprcespor supply how about fedding the new trafo via one of those Shaffener filters? Im also considering adding a dedicated trafo for the preamp which might also benefit from one of those filters. I would shield any AC supply lines also.
What do you think?
mikesnowdon said:
Schaffner filters, like any filter, have to be matched to the load. If they are not, which is very likely, they may indeed worsen the performance of the amp. They will certainly not help against hum, because they are basically low-pass filters.SimontY said:
Yes, shielding may be a waste of time, because shielding efficiency decreases with the frequency.SimontY said:
How would good passive components help against hum?
Oh are we talking about hum? lol. Sorry, I was just in sound quality mode. Why is there hummm on a commercial amp? That's very odd. Do you have a ground loop Mike?
Have you tried unplugging all inputs and listening? Have you tried only connecting CDP? I presume you haven't connected the CDP to safety earth. You should also check the routing of any interconnect leads, make sure they're not going too close to AC power leads. You could also short unused preamp inputs, thought I doubt this will help much.
When did the hum start? Was it after an upgrade/mod?
Simon
Have you tried unplugging all inputs and listening? Have you tried only connecting CDP? I presume you haven't connected the CDP to safety earth. You should also check the routing of any interconnect leads, make sure they're not going too close to AC power leads. You could also short unused preamp inputs, thought I doubt this will help much.
When did the hum start? Was it after an upgrade/mod?
Simon
Hi all. I'm planing to move into your discussion and make some little tweaks to my 340 
1. Swap Opamp to LM4562. It has normal price (about 5$) and doesn't need any adapters. Maybe some time later, i'll offer some BD Adapters and put AD8066 There.
2. Change input Cap to lower value polypropilene (about 1uF).
[optional]: change all decoupling caps to polyp. (not so pricey as Mike's Mundorf,but something cheap like Epcos or our russian k73-17)
3. [Optional]: Adjust tone_Control section. Maybe change bass frequency and bandwidth. This part is optional, because i don't use tone all the time, but sometimes I want to set bass pot to about 3 o'clock... I think it's my speakers fault , but as soon as have them, tone controls section is usefull for me.
So who has any ideas how to count bass freq ?
4. Change Main PS Caps to Mike's Stock Caps (2*6800uf => 2*10000uf)
[optional] change diodes.
5. [optional]. Insert small +5V powersupply for source_selector section. Do You think that separate small trafo (5VDC 1A) + 78T05 VR + Rubicon MBZ litycs and some small-size polyprop... will slightly improve sound?
P.s. sory for my "broken" English
Simonty, see this thread:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=130909&pagenumber=1
here you can see a lot of internal photos:
http://s3am.livejournal.com/2319.html
1. Swap Opamp to LM4562. It has normal price (about 5$) and doesn't need any adapters. Maybe some time later, i'll offer some BD Adapters and put AD8066 There.
2. Change input Cap to lower value polypropilene (about 1uF).
[optional]: change all decoupling caps to polyp. (not so pricey as Mike's Mundorf,but something cheap like Epcos or our russian k73-17)
3. [Optional]: Adjust tone_Control section. Maybe change bass frequency and bandwidth. This part is optional, because i don't use tone all the time, but sometimes I want to set bass pot to about 3 o'clock... I think it's my speakers fault , but as soon as have them, tone controls section is usefull for me.
So who has any ideas how to count bass freq ?
4. Change Main PS Caps to Mike's Stock Caps
(2*6800uf => 2*10000uf)[optional] change diodes.
5. [optional]. Insert small +5V powersupply for source_selector section. Do You think that separate small trafo (5VDC 1A) + 78T05 VR + Rubicon MBZ litycs and some small-size polyprop... will slightly improve sound?
P.s. sory for my "broken" English
Simonty, see this thread:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=130909&pagenumber=1
here you can see a lot of internal photos:
http://s3am.livejournal.com/2319.html
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.