I know, you had nothing to do with this; but maybe you can find the guy who did and slap him silly for me.
I am reworking a CD-80 for a customer. As I usually rip the SAA722 out, I pay not attention to how it is laid out. This customer needs a digital output, so it has to stay in.
The bypass cap for pin 24 first goes through a 15 mm staple. Then, a zigzag PCB trace at least twice as long.
How is that cap supposed to do anything that far away?
I looked at the service manual for the CDB-472, and -582. They place the bypass cap about 3-4 mm away. Much better. With as much attention to detail other places in the CD-80, how did that one slip by? Just curious.
Jocko
I am reworking a CD-80 for a customer. As I usually rip the SAA722 out, I pay not attention to how it is laid out. This customer needs a digital output, so it has to stay in.
The bypass cap for pin 24 first goes through a 15 mm staple. Then, a zigzag PCB trace at least twice as long.
How is that cap supposed to do anything that far away?
I looked at the service manual for the CDB-472, and -582. They place the bypass cap about 3-4 mm away. Much better. With as much attention to detail other places in the CD-80, how did that one slip by? Just curious.
Jocko
Jocko Homo said:
Jocko,
Your observations are very right. How did that slip the mind ? I'd say not sufficient attention, most of the times the knowlegde is in place, but the awareness not......
By the way, I have seen worse. Imagine the supply of the decoder and radial servo driver being shared, with only a single electrolytic haf way in-between.......
From your posts I understand you very much know what you are doing, but others may be happy to read
http://members.chello.nl/~m.heijligers/DAChtml/Supply_decoupling.pdf
Designing and building CD players is a difficult process, as you need to know about
- optics
- mechanics
- electronics at AF, RF, digital, EMC, the whole range......
enjoy the weekend
Jean-Paul:
Do you have any idea how many of these that I have worked on, and just now discovered it? I feel dumb, but I really do tear out the filter, so I never looked. Wonder what else I have missed...
Guido:
I have to confess to doing things like that when space is critical. I have just considered my self lucky when it still worked. But cetainly not proud of it.
Jocko
Do you have any idea how many of these that I have worked on, and just now discovered it? I feel dumb, but I really do tear out the filter, so I never looked. Wonder what else I have missed...
Guido:
I have to confess to doing things like that when space is critical. I have just considered my self lucky when it still worked. But cetainly not proud of it.
Jocko
Hi Jocko, I usually decouple the SAA7220PB with OSCON and sometimes I reroute the +/- wires. For example a separate ground wire makes a big difference. When modified like this the SAA7220PB is a chip I can live with. But it is not the best marriage in the world I have to admit 
It is a garbage transmitter of the best quality. Very hard to make it quiet. A separate supply is the best solution but an expensive one for a paying customer.
When I think of it: IMO the distance of decoupling caps to the regulators is too much as well. If I would have a CD80 myself I would redesign the power supply board as it is not of the same league as the other boards are. There is some space for placing the caps next to the regulators when you drill some 1 mm holes.
BTW The bracket in the middle is an excellent spot for mounting a low jitter clock. Short path to the SAA7220PB and reasonably nearby the power supply. Only some care is needed when drilling holes in the bracket as the upper cover has a fin that can be at the same spot as where the clock will be mounted. So before drilling please check with the upper cover closed if you choose the right spot for the clock.
Are you talking about the CD931, Guido ?
It is a garbage transmitter of the best quality. Very hard to make it quiet. A separate supply is the best solution but an expensive one for a paying customer.
When I think of it: IMO the distance of decoupling caps to the regulators is too much as well. If I would have a CD80 myself I would redesign the power supply board as it is not of the same league as the other boards are. There is some space for placing the caps next to the regulators when you drill some 1 mm holes.
BTW The bracket in the middle is an excellent spot for mounting a low jitter clock. Short path to the SAA7220PB and reasonably nearby the power supply. Only some care is needed when drilling holes in the bracket as the upper cover has a fin that can be at the same spot as where the clock will be mounted. So before drilling please check with the upper cover closed if you choose the right spot for the clock.
Are you talking about the CD931, Guido ?
CD931
Jean-Paul, Guido,
has either of you been able to lay hands on a 931 and see for yourself or are you referring to my old ravings somewhere in this forum? Just discovering what chips had shared ground paths made me reel...
Still, it got good reviews in its day, and you still see many operative ones on ebay...
Regards,
Eric
Jean-Paul, Guido,
has either of you been able to lay hands on a 931 and see for yourself or are you referring to my old ravings somewhere in this forum? Just discovering what chips had shared ground paths made me reel...
Still, it got good reviews in its day, and you still see many operative ones on ebay...
Regards,
Eric
Re: CD931
Eric
I saw one when I helped somebody out, i do not own one (don't need one, have avery decent sounding TEAC VRDS 10)
Ofcourse all chips should sit on the same groundplane. By design however one should prevent the currents through the planes mxing up with each other.....
See also the short tutorial that I wrote a while ago on layout and decoupling:
http://members.chello.nl/~m.heijligers/DAChtml/Supply_decoupling.pdf
all the best
capslock said:
Eric
I saw one when I helped somebody out, i do not own one (don't need one, have avery decent sounding TEAC VRDS 10)
Ofcourse all chips should sit on the same groundplane. By design however one should prevent the currents through the planes mxing up with each other.....
See also the short tutorial that I wrote a while ago on layout and decoupling:
http://members.chello.nl/~m.heijligers/DAChtml/Supply_decoupling.pdf
all the best
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.