advertisement


Philips CD160 not working

It's known to be rather value specific, it must be 33uF, don't be tempted to uprate it.
 
Oh, rather interesting. I used 47uF caps of circa 1 ohm ESR and have had no real issues except skips / jumps etc on discs which are damaged, although I admit both CD160s are more "sensitive" to poorer discs than my Micromega stage 3. I have bough two 33uF, 40V BC axials which were NOS but late '82 so circa 37 years old. I have reformed them and they now measure circa 38-40uF, 0.5 ESR and have about 1-2µA leakage current (they "loose" about 1V in 30secs by self discharge i.e. only 10 megaohm voltmeter connected across them). What do you think, try them or leave the others in place as I suspect that the much newer 47uFs will not be that much different in value?
Cheers,
AP
 
Just get fresh new BC/Vishay 33µF axials. New old stock caps are a no-no in my opinion. That some people might sell them beats me.
 
I get them next week so will fit, run in and trial versus #2 machine with 47uF 1ohm ESR in that position, versus a Micromega Stage 3 (which is less sensitive to surface damage) and report back. Could be a few weeks tho'.
Best,
AP
 
Caps turned up Sat which was good because I was assessing the damaged discs to see how they behaved in the three machines and making some notes. Took a new 33uF at random and measured it: 46uF and 1.5 ohm ESR so well within spec (-10/+50%). Took out the 47uF that was fitted which measured 47uF and 0.8 ohm ESR. Was sceptical that it would make any difference, but fitted one and ran it for a short while before working through the damaged tracks I had assessed earlier in the day. Made no notable difference TBH. Got me thinking though as my #1 machine was clearly as bad (on discs with small pits / scratches) or worse (on a disc with a 3mm wide circumferential damaged zone) than my #2 machine whilst the Drive 3 sailed through the lot with no problems. I had removed the swing arm from the #1 machine a few months ago to clean / grease the bearings and despite trying to accurately align this with visual markers, it is clear that I can't have got it in the original position. I made a 5" square disc out of perspex and put a 16mm hole in the middle so I could check the optical alignment. The reflection in the lens (compared to that from the disc) was found to be about 2mm off-centre (outwards WRT the radial arm and at 90 degrees it was radially inwards WRT the disc) so I adjusted to get this down to about 1mm in both. Massive improvement, especially on the circumferential defect. Readjusted to get both offset values as close to zero as possible and this was clearly worse so have gone back to about the mid point. Not quite as good as the first adjustment but still better than it was. This also seems to have improved upper mid-range detail retrieval and image focus / depth.
Is there a method for optimising the optical alignment using the four diodes used for tracking / focusing as the service manual says to get the lens reflection within 4mm of the disc reflection, but it is clear from my results above that this would clearly not be optimal at +/-4mm in both planes when a much smaller adjustment has such a profound effect?
 
Two weeks ago I tried soldering four wires to the back of the ribbon socket (#2 CD160) but made a hash of it. Damaged one of the traces to the chip, so removed the wires, replaced the trace with a wire and put it back as is; thankfully all OK. Last weekend, used a piece of strip-board which I thinned the backing down and placed over the chip so I could tack-solder to the legs of the chip and scoped the four diodes. D1 had a lower output than the other 3 (about 2/3) near the inner of the disc but this evened out with D2 (both about 80% of D3 and D4) towards the outside of the disc and making an adjustment to the lens alignment made no visible difference on the scope.
Had a good listen to both the CD160s and compared to my Arcam Delta 170 transport (as this has the SAA7220 same as the CD160s, but the B variant rather than the A) with Bushmaster ii DAC and my conclusion was that there was little difference between the three. In fact I confirmed my previous suspicions that there was more difference from changes of seating position than from the three players and suspect my comments from above are due to this rather than alignement. Confirmed this with headphones (DT990 driven by a Ion Obelisk 3). I must admit though that my hearing starts to tail off about 6kHz and I can't hear test tones of 10kHz so I suppose it is possible that there are HF differences that I can't hear.
 
Managed to get a Tentlabs XO from Colin @ Chevron Audio and am awaiting receipt of a Flea kit from Ray's Audio in the Netherlands.
Looking at acoustic.org I notice that Martin Clark indicates that this is better powered with a separate transformer rather that via an on-board supply because of separating the return currents from the clock signal earth to the SAA7220. So can I get away with using an existing supply to feed the Flea or do you only get the full benefits with a separate TX/psu? Would hate to fit it an be disappointed, especially as its not that much extra work/cost.
Thoughts please?
 
Well, I suppose the answer is that you should get an improvement if using an existing supply from the player but you won't get the full benefit without an additional transformer.

It's well worth supplying separate clock feeds to the SAA7220 and decoder chip.

Give the new XO a bit of time to settle in before making any judgements too.
 
Thanks Mike,
Yes, I'd seen the comments (by Martin I think on Dowser's CDi thread) about feeding the decoder chip as well. Not prepared to cut the leg of the SAA chip so will have to cut a track to enable this to be done as this would be easy to return to original if required. Going to use a 15V twin secondary of about 1.5VA (i.e. 50mA/winding) and will thus have a spare winding available.......... (-15 volt supply to TDA chip?)
 
Mike P said:
Well, I suppose the answer is that you should get an improvement if using an existing supply from the player but you won't get the full benefit without an additional transformer.

It's well worth supplying separate clock feeds to the SAA7220 and decoder chip.

Give the new XO a bit of time to settle in before making any judgements too.
Click to expand...

I tried an additional transformer, a super-clock but couldn’t hear any differences.
Truth is, I thought I could, but the effect faded quickly.
I didn’t measure anything back then, but I guess I could have seen something there.
But what’s the point?
 
Sorry chartz, not sure what you mean about your 'what's the point?' question. In making measurements at the time or in changing from a crystal to a Flea / XO?
Certainly according to page 9 of the Flea manual on acoustica.org, adding the extra transformer to provide a separate supply to the Flea / XO was 'the audible benefits are not subtle' is how Martin phrased it.
Best,
AP
 
The point is the sonic difference is zero, just the result of imagination – and the pride of doing it – and so measuring anything is rather pointless.
One old man’s opinion of course.
 
I have to admit that I've had pretty variable results with fitting aftermarket clocks. Sometimes I felt it was an improvement, sometimes I felt there was no difference at all and sometimes I felt it made things worse.

My favourite one-box player is my Sony CDP-555esd and I do have an Audiocom Superclock 4 fitted to that. I found a way to fit one very neatly and in a way that didn't damage the player at all, so it can be put back to original any time I choose.
 
Chartz,
OK understood. Thanks. ('One old man's opinion of course' made me laugh).
Mike,
Any pointers as to where you found gains or losses? (Eg Sony: works well / Philips not so good, or some such)
Cheers,
AP
 
a.palfreyman said:
Chartz,
OK understood. Thanks. ('One old man's opinion of course' made me laugh).
Mike,
Any pointers as to where you found gains or losses? (Eg Sony: works well / Philips not so good, or some such)
Cheers,
AP
Click to expand...

In terms of general advice, I'd just advise you to do a temporary installation first, that way if you don't like the results you can more easily put it back to how it was before.

In terms of technical advice, I doubt I can tell you anything you don't already know and Martin's your man for technical questions. However, a few pointers that may be helpful:

  • Keep the connections from the clock output as short as practically possible. Locate the clock to keep these connections short.
  • Use a twisted pair or coaxial for the clock output connections.
  • With the clock all connected up observe the output on a scope. You'll have something approximating a square wave. If you have any significant overshoot or ringing damp the output with series resistance until the overshoot and ringing are minimal. You don't' want to be injecting unnecessary HF noise into the IC receiving the clock.
 
OK Mike, thanks.
Only 20MHz scope here so I'm sure it won't give the same sort of images as those posted by Martin Clark, nor show much in the way of ringing either. When I've looked at the existing clock signal I see a sort of 'shark-fin' or rounded triangle-wave shape. The Flea manual states 47-100 series resistance from XO output for decoder / digital filter but I'll see what the kit comes with. Initially I'll just do the SAA7220 input and leave the decoder as this is will be transmitted from the SAA anyway.
 
20MHz scope showing a 'shark fin' / rounded triangle' is about what you;d expect with a scope like that - enough for proof of function :)
47-100ohm will always work on the output, with a twisted pair or coax - so don't worry about that; and Mike P has made some very good points I'd agree with just above.
 
I'll be working on a re-clock'd Marantz CD17 soon. I will revert to the original clock and get rid of the silly second transformer and PSU.

Funny, eh?
 
You must log in or register to reply here.


advertisement


Back
Top