Caiman Mods

[MartinT]

These Dexa regulators from Parts Connexion look interesting. If the "100 times better ripple rejection, 10 times lower noise" claims are true then they should make a substantial difference over the stock 7805/7812 regulators. I'm ordering a 5V and 12V module just for the fun of using something different and will report on my findings when received. I should have the Caiman PSU soon, ready for them.

http://www.partsconnexion.com/regulator_dexa.html

[StanleyB]

The Murata versions are similar in spec, and a fraction of that price.

[MartinT]

Understood Stan - I was feeling adventurous and just thought I'd try them. Think of it as adding to our pool of component knowledge.

[leo]

If spending that sort of money I'd have gone for the Paul Hynes , anyway let us know how you get on with the Dexa's Martin

[Carson]

So I finally had some time to put in the new non-polarised caps this morning, they're still settling in, but I can hear a pleasant improvement.
Here are a couple of photos of the operation.



The capacitors quite literally only just fit into the case. So much so that, I placed a couple of stickers on the underside of the lid over the op-amp board to ensure i dont get any chances of a short.

I also fitted a couple of OP1234 op-amps. Must admit I like the improvements.

[Shanedudddy2]

Stan if you could please let me know the minimum capacitor value to get rid of the DC offset without limiting the bass I would really appreciate it.
Since I don't have a camera I`ll explain what I did, disconnect the 3 wires(labeled L,G,R (for left, right and ground)) on the far right side of the board, on the closest board to you when looking at the DAC front on. Then I disconnected the line out connectors to the board. Then soldered a "high quality" capacitor from electronics store near me to fix the DC offset, between the left and the negative connector and the same for the right and the positive connector).

[here we go again]

Having been quietly following the Beresford DAC phenomena since the 7510 days - I finally gave in and ordered a Caiman from Stan about a month ago to partner my Shigaclone.
I have to agree with almost everything already said about this fantastic DAC - out of the box it was good, but obviously burning in - within a day it was something else - now with a couple of hundred hours under it's belt it is a wonderous thing and like most others, I have been rediscovering my CD collection.

Inevitably the urge to see what could be extracted from this beast had me rolling opamps recently - I'm fortunate to have a good stock of so-called 'audiophile' opamps hanging around from various projects so i got to it. My usual favourite is the THS4032 - but in the Caiman it just wasn't right IMO - I went through most of the usuals - OPA627, AD845,AD846,AD847, OP2132,LT1364 + more - in the end i returned to the stock LM4562's that Stan fitted. Nice one Stan - I reckon you got it bang on with these! - Other opamps can do some things better but in my system none of them had the balance that the stock ones have

So - then I flirted with the resistor & cap in the feedback loop - but didn't get on with that either - again the balance seems better with just the stock resistor and MCL5/6 left empty. After that it was the 4 line stage coupling caps on the analogue board - a quad of 3.3uF Wima polys replaced the lytics and have removed a thin veil - the jury is still out on them and I turned my attention to the power supply.

Because just one 7805 feeds all the +5v in the device - I started to take a look at just how many seperate regs it would take to give every user it's own supply. The truth is - I think there are at least 9 different users points!! - the important ones are obviously the DAC and receiver chip - but in reality the DAC itself has 4 seperate supply points - and the receiver has 2. I began by knocking together a trio of +5V LM2940's on a small piece of vero - then looked at the Caiman board to see where I could seperate the single supply and inject some fresh volts.

the bottom line is - splitting these supply points is not THAT easy - injecting seperate volts for the receiver digital and analogue supplys was pretty straightforward - but finding a suitable place to seperate the digital supply for the DAC chip wasn't easy - mainly because of the size of the device and the fact that pins 8 & 9 (+5V dig & +5V R.Analogue) share the same local decoupler and bypass! - this was super tricky and not for the faint hearted. But - I now have 3 extra +5V regs feeding the dig and analogue parts of the receiver and the dig supply of the DAC chip - all on a neat little piece of vero that fits inside the case still - I'll get a pic or 2 up soon. The existing +5V reg still supplys everything else - Eventually, I may seperate the Wolfson analogue supplys too with an additional 3 regs.

So how does it sound? - quite stunning IMO!! - seperation is the word here - really, the ability of this DAC now to extract micro details in the music is just outstanding. The tonal balance has not been affected too much either - it might have become a little brighter - although I suspect this may actually be because all those little cymbal splashes are now more evident.

Anyway ....................................... the main reason for this post was to say that Farnell UK have a couple of dozen of those Murata +5V switching regs in stock - no. 1703410 - I ordered one to see what it can do - should be here tomorrow.

thanks to Stan for a great product and thanks to all the contributors to this and the other great DIY threads on AoS.

[toad]

what are the BG equivalents....?

the 47uF 6.3V NX Hi-Q or something rated higher...?

anyone...?

hmm...i thought this was the right place to ask this....

[leo]

hmm...i thought this way the right place to ask this....

Sorry Toad, I'd say 6.3v is a bit too low voltage in this application, go for 16v or higher.
If your struggling with the taller caps I guess you could lay them on their sides, the grey caps Stan recommends are pretty good though and fit ( a little tight but still fit)

Edit, if your set on using BG's with their unique colouration the BG N 33uf 16v may fit, not sure, you'd have to measure the space in the case, measurements for these caps are 6.3mm dia x 11mm length

[toad]

thanks leo....

drat...that wasn't the reply i was hoping for as i ordered them a week ago...

alright then....16v caps it is...