Not a source but…

[J!m]

I’m starting to work over my Aragon D2A.

I had someone smart go over all the chips in the digital board. They are all socketed so I figured after 30 years or so maybe there are better chips? Guess not.

Converters are the Burr-Brown PCM 58P chips and they sound good. Not much to do here but I’ll replace the one electrolytic on the board.

The analogue board is fully discrete dual-mono. An entire gaggle of Dale resistors and Sprague caps. Some moderate size 20% tolerance coupling caps too. I got some tighter tolerance metalizer film coupling caps and new power supply caps to replace the Spragues.

Then the external power supply gets replaced. Supposedly this is the largest upgrade or rather, the upgrade with the largest impact on sound. Perry suggested an instrumentation power supply and that will be the next project. I plan to build it in a second 19” wide chassis. I may be picking brains when I get there.

I was even considering listening after replacing parts in one channel only- the ultimate A-B test? Probably not…

 

[MarkWComer]

Chips galore! Got some onion dip?

 

[J!m]

No dip- switches… but we got trim pots!

 

[mr_rye89]

I think that has the same DAC chips as my late 80s Yamaha CD Player. Is it pronounced air-a-gone or ah-ra-gon (rolled r)?

 

[J!m]

I pronounce it air-a-gahn. I guess that's how you do it.

These old Burr-Brown chips are liked by many. I thought the support chips might have lower noise/faster/nicer colors, or replace single op amps with dual op amps after 30 years, but apparently not. I think the entire board would have to be redesigned for any changes to be made. I did find the mix of Sony and Yamaha chips to be interesting. I guess (no idea) the Sony chip supports the optical input? I had a Sony CD player with that same "TosLink" connector, which, coincidentally, was developed by Toshiba (hence, the "Tos"). None of their chips in here.

They socketed everything to have an upgrade path, but never did anything with it. The Mk. II version uses different DAC chips I believe and doesn't sound as sweet (totally different board). They also made changes to the analogue board (as far as I know) which probably also contributed to the sound signature. But I don't know why. Only the bass can be a bit lacking, and the coupling caps will help there. I got higher voltage and tighter tolerance as replacements.

 

[WOPL Sniffer]

Don't forget the little things. All that funky flux they left on the boards.

 

[J!m]

They call that "patina" now. Removing it destroys the value.

 

[mr_rye89]

Nevermind, my CDP (Yamaha CDX-900) has a PCM56P in it

 

[J!m]

Well, these boards are absolute garbage.

Single layer thin tin trace. And they used eutectic or (more likely) silver bearing solder.

Bottom line? Pulling components is like pulling alligator teeth. (If you look close you can see the OEM lifted traces)

So, all I did was the single cap on the digital board and the two mystery coupling caps. I have the 12 caps for the analogue board but I WILL destroy the board getting those out.

I also cleaned the selector switch contacts and the gold RCA connectors which were surprisingly nasty. Hopefully I get a chance to test it this weekend. If it sounds ok I’ll go for the power supply upgrade.

 

[George S.]

If it's lead free solder, try mixing a little leaded solder into it before desoldering. That's about the only way to desolder capacitors and other components on computer motherboards without hurting the traces or pulling the plated through holes out of the board.

 

[Gepetto]

Well, these boards are absolute garbage.

Single layer thin tin trace. And they used eutectic or (more likely) silver bearing solder.

Bottom line? Pulling components is like pulling alligator teeth. (If you look close you can see the OEM lifted traces)

So, all I did was the single cap on the digital board and the two mystery coupling caps. I have the 12 caps for the analogue board but I WILL destroy the board getting those out.

I also cleaned the selector switch contacts and the gold RCA connectors which were surprisingly nasty. Hopefully I get a chance to test it this weekend. If it sounds ok I’ll go for the power supply upgrade.

A real shame that folks cheap out on the boards when the price tag they charge is so high. Bean counters ruling, not quality.

 

[laatsch55]

Whole lot different when the chief designer is also the bean counter..

 

[George S.]

Even the masking is separating. Look at the second photo, at the center, by C45. Several other places also. Bet it was wave soldered with lead free for all the IC sockets, then further hand soldered where flux is showing. I hate lead free, takes 3 times the work of unsoldering leaded solder.
Going to get some Chipquick for future projects and try mixing it in with lead free.

 

[J!m]

I tried wetting it with 60-40 but that didn’t help much. It did help get the heat into the solder but it was cooling and solidifying really fast.

If it was plated through, it probably would have held up better.

 

[J!m]

I found the old data sheet for the burr-brown converters and solved one mystery.

I noticed the trim pot next to each D to A chip. Turns out that is a trim for minimizing distortion to zero for the first bit.

Reading further I learned that the option exists for the second, third and fourth bit as well (and left floating if unused). I’m guessing the improvement does not justify the investment of six resistors and six trim pots to a manufacturer, but in my diabolical mind i’m thinking I could make up a daughter board and make use of this dormant feature.

Trouble is, you need gear and skill to make the adjustments.

And by the way, there is no drop in replacement chip. The architecture is different on the newer chips, despite being the same size and pin count. (Dumb).

 

[Gepetto]

I found the old data sheet for the burr-brown converters and solved one mystery.

I noticed the trim pot next to each D to A chip. Turns out that is a trim for minimizing distortion to zero for the first bit.

Reading further I learned that the option exists for the second, third and fourth bit as well (and left floating if unused). I’m guessing the improvement does not justify the investment of six resistors and six trim pots to a manufacturer, but in my diabolical mind i’m thinking I could make up a daughter board and make use of this dormant feature.

Trouble is, you need gear and skill to make the adjustments.

And by the way, there is no drop in replacement chip. The architecture is different on the newer chips, despite being the same size and pin count. (Dumb).

No one makes drop in replacements unless they are jellybean chips. No money in that.

 

[J!m]

It’s interesting that every chip is socketed- why would a company cutting corners on board quality and lack of additional trim pots pony up for sockets, unless the OEMs told them there would be an upgrade path?

Something doesn’t compute.

These converters may be old but they still sound good. I bet maximizing them by reducing the noise further would be worth while (and possibly what the “big boys” were doing in their more expensive black boxes).