After reading Pete Goudreu article about bypass caps http://groups.google.com/groups?selm=4te25c$hcn@agate.berkeley.edu&output=gplain I decided to do it in my DAC as well. I used 15nF X7R , 330nF X7R chip caps 0612 from DigiKey and Panasonic 120u HFQ. Mounting and stacking those tiny cap is not an easy job.
Attachments
Panasonic HFQ caps are mounted on the opposite side. So what's the improvement? My DAC already had a pretty good sound and it was better than my TT. By adding the caps I achieved another level of refinment (although not something that blows me off). It's just that reproduced music becomes slightly more real, the notes seem to be floating longer in the air and there is a bit more magic in performance. The cost is not much and it took me an hour to do, so I would recommend it. The chips visible here are 2 stacked 1704 DACs. The shielding brought a big improvement.
The cost is not much and it took me an hour to do, so I would recommend it. The chips visible here are 2 stacked 1704 DACs. The shielding brought a big improvement.Attachments
This is the look of main board which came from David Broadhurst. I placed damping material on most caps. However, I noticed that too much damping kills the liveness of music, so eventually I removed some of it. Damping the clock was really important on my board. I tried to put damping pads on top panel, but it sounds better without damping. So my recommendation is to experiment, because one never really know what's good and what's not.
Attachments
Banned
Joined 2002
Banned
Joined 2002
Banned
Joined 2002
Did you get any resonance from stacking two chips of different capacitance in the same form factor? In my experiments this is sometimes the result, so I have taken the practice of using decreasing sizes for decoupling. For example, 4.7µF in 1206 in parallel with 0.1µF in 0805 for local decoupling.
Anyway, how you check for resonances?
1.First get a degree in Electrical Electrical engineering.
2. Go to work for a company with a 500MHz Network Analyzer.
3. Read the directiions for Analyzer in step 2.
4. Measure resonances.
I am pretty sure that is how Pete did it but I think his degree in Physics.
Sin Bin by Sundown!
Ren
1.First get a degree in Electrical Electrical engineering.
2. Go to work for a company with a 500MHz Network Analyzer.
3. Read the directiions for Analyzer in step 2.
4. Measure resonances.
I am pretty sure that is how Pete did it but I think his degree in Physics.
Sin Bin by Sundown!
Ren
Banned
Joined 2002
THAT was mean..
Acually I was having a bit of fun. This kind of thing is pretty difficult to measure and I was just trying to make that point. I believe they probably is some advatange to this approach and I bet money Pete did measurements pretty much as described since he is an RF and signal integrity expert. No harm intended.
Measuring supply noise with a 100MHz scope is a good starting point for most of use mere mortals.
Ren
Maybe some info at www.sigcon.com
Acually I was having a bit of fun. This kind of thing is pretty difficult to measure and I was just trying to make that point. I believe they probably is some advatange to this approach and I bet money Pete did measurements pretty much as described since he is an RF and signal integrity expert. No harm intended.
Measuring supply noise with a 100MHz scope is a good starting point for most of use mere mortals.
Ren
Maybe some info at www.sigcon.com
What about adding a high value ceramic 1206 or 1210 cap (4.7uF or 22uF) to the triplet, if you have room for it. I've made some measurements (no audio results yet), and found that these little chips do a tremendous job removing digital supply noise, when combined with such a triplet.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.