Gooch said:
Hi Dave,
Until a couple of years ago I ordered by Steve. Unfortunately he's more in money than in supplies these days. $2 for a resistor, aaaah!
The 1/8W are much smaller than those at Angelas. They fit in the housing
Packs of 50 sold for $5 - $10 in the good ol' days.Thanks anyway!
Jaap
When measuring with the ears be aware it's not a flaw you are compensating with another flaw...
You can have trannies instead. In UK you have Sowter, making excellent transformers.
Although sometimes put the other way around, it are the metalfilm resistors adding to sound because of their properties. Carbon composite adds nothing (some say noise, but that's another story).
I like lemon ice and hate banana. How about you?... Do you think I have ruined mine by using 3300uF caps around parts? It did sound better, but maybe it would be better still if I'd used smaller ones
Wait a couple of days and behold my modlist. Simplicity rules!
Regards, Jaap
my idea...
hi fellows,
I'm ordering the components upgrade lit from Ray, and will apply most of the tips : remove mute and speakers. ( output caps already gone... ).
After I plan to fit a clock as Tent labs (or one cheaper if I can build it), and have a try with OPA2134 and AD8620. Should already have a good player then
And who can tell me where to download the workshop manual ? Thx !
hi fellows,
I'm ordering the components upgrade lit from Ray, and will apply most of the tips : remove mute and speakers. ( output caps already gone... ).
After I plan to fit a clock as Tent labs (or one cheaper if I can build it), and have a try with OPA2134 and AD8620. Should already have a good player then
And who can tell me where to download the workshop manual ? Thx !
disco said:
I take your point but those caps are nice for the money, that much is virtually proven fact
You can have trannies instead. In UK you have Sowter, making excellent transformers.
It's just a simple op-amp based pre-amp, which probably isn't worth taking to such extremes. The next pre-amp though...
Yeh, I've heard people say it's noise and colouration.
I like lemon ice and hate banana. How about you?
Well I rather like banana!
Wait a couple of days and behold my modlist. Simplicity rules!
Regards, Jaap
I wait with bated breath.
Cheers,
Simon
SimontY said:Hi guys!
My maths is rather poor and I couldn't quite work this out
Could you help me choose a minimum cap value please? My pre-amps input impedance is 100K.
I hope to be using this discrete output in a couple of weeks or so lol (with FETs)
Thanks!
Simon
disco said:Simon,
Good choice, the fet buffer
Maths come down to this simple eqation.
Take the outcome of the division of 1 by the result of the input impedance of your amp (in ohm) multiplied by the coupling cap (in Farad) multiplied by 6.28
So, f=1/(6.28 x 100000 x 0.0000033) = 0.48 Hz, which is low enough.
And, f=1/(6.28 x 50000 x 0.0000022) = 1.45 Hz, which is also low enough for 50kHz input and 2.2uF cap.
Regards, Jaap
Hi guys,
That's all true, but......
there's also the low frequency group-delay that starts to act up when the signal frequency is at and below the -3dB point. I don't know how much harm that will do, because the low-end is less critical. But, just in case, to be sure, if you have some room to spare (and some bucks $$$) take a cap that's bigger.
How much bigger? I always like to calculate things
. Well, you have to set some limits to determine the cap size. So I looked at the group-delay at 5kHz and though it would be nice to have a point in the low-end where the group delay would be no more than twice that value. Is that sane? I don't know. We're talking microseconds here at low frequencies. Maybe ten times that value? Or 1% of the frequency's period? Anyone?To give you an idea how all this looks:
Both graphs are with 100k input impedance. The top graph is with 3u3 and the bottom one with a 10u output cap. The only problem is, at what frequency do we want this to happen??
Regards,
Ray.
Attachments
6h5c said:
Hi guys,
That's all true, but......
there's also the low frequency group-delay that starts to act up when the signal frequency is at and below the -3dB point. I don't know how much harm that will do, because the low-end is less critical. But, just in case, to be sure, if you have some room to spare (and some bucks $$$) take a cap that's bigger.
How much bigger? I always like to calculate things
To give you an idea how all this looks:
Both graphs are with 100k input impedance. The top graph is with 3u3 and the bottom one with a 10u output cap. The only problem is, at what frequency do we want this to happen??
Regards,
Ray.
Don't know LOL
Its all swings and roundabouts
Brent
6h5c said:
Hi guys,
That's all true, but......
there's also the low frequency group-delay that starts to act up when the signal frequency is at and below the -3dB point. I don't know how much harm that will do, because the low-end is less critical. But, just in case, to be sure, if you have some room to spare (and some bucks $$$) take a cap that's bigger.
How much bigger? I always like to calculate things
To give you an idea how all this looks:
Both graphs are with 100k input impedance. The top graph is with 3u3 and the bottom one with a 10u output cap. The only problem is, at what frequency do we want this to happen??
Regards,
Ray.
Hi Ray,
That's an interesting graph but I for one won't be worrying about that tiny amount of delay. I'm sure my subwoofer is capable of causing problems 1000x more severe and audible. And then there are the crossovers.. Can't wait to hear your filter in action!
Simon
SimontY said:
Hi Simon,
Thanks for sharing this insight on your favourite meal...
My modification is inspired by Diva
After heavy modifications on my first CD63 I tried how far I could come with minimal parts. I got surprised.
Looking at the central 5 volt power supply it's obvious it has to deliver divers currents into divers loads, one affecting the other.
Safest remedy is to build separate supplies but I reasoned otherwise. Why not make use of small capacitances and two reservoirs? One at the beginning as usual near the regulator, and one at the end at the most remote part of the traces. This resulted in two benefits as inter capacitor load currents are kept small and the smaller caps fill up quicker than big ones.
So, what I did was replacing the 5V reservoir cap (C815, was 3300uF) with a Black Gate NX 1000uF and placing another in position C504 (HF LPF, was 220uF). I made a new ground connection to the central PSU ground and cut the original at U256. Furthermore I replaced every electrolyte with standard Black Gate 10uF/50V (unfortunately I had not enough, so I also used some 47uF general purpose Nichicon). Modern caps have very good HF decoupling property, so there's little need for bypass anymore.
As one can imagine a faster supply has impact on dynamics. As a result Music is more natural, vivid. The soundscape has become clearer. Bass has authority without the often heard boom.
Another field of progress lays in stopping resonance interfering with laser tracking. Looking at the flimsy case brings tears to your eyes. The effort put in the circuits is completely negligent in the housing. So, out it goes. I have build a wooden case of 9mm birch multiplex, two layers for the bottom, one layer for the sides. The transport was fitted on four pillars of foam (ca 5x5x5cm).
At the moment I am experimenting with damping. A weight of app 80 gram on top has big impact on the soundstage. Music is heard in 3D, the stage deepens. Voices sound softer yet better distinctive, now lacking the squeaking coloration.
I intended to give a elaborate list on components I have changed, but it comes down to HF-suppression (TNT-audio), coax and shielding of the XTAL. As time goes by my modifications will change because of the experiment, so I don't think it's wise to produce a list now. If someone is interested in the partnumbers for stripping the superfluous circuits, I'll happily supply.
Best regards, Jaap
My hifi is waiting sings Norah...
I have had the discrete pcb running on a standard cd67. I am comparing the original output with the discrete.
The 67 HDAM and caps have been removed!!
The discrete pcb with bc550 transistors gave more air and a little more bass.
The FET version (matched 2SK170GR) sounds less open and has less low end bass than the original opamps.
I have 3.3uF caps
mmmm
At the moment I feel the transistor pcb is better.
Brent
The 67 HDAM and caps have been removed!!
The discrete pcb with bc550 transistors gave more air and a little more bass.
The FET version (matched 2SK170GR) sounds less open and has less low end bass than the original opamps.
I have 3.3uF caps
mmmm
At the moment I feel the transistor pcb is better.
Brent
Just to follow up my last post.
The discrete pcb is tagged onto r21 etc and 12v regs with crocodile clip leads and therefore not ideal.
I am switching between two inputs on amp (cd67 rca and discrete o/p).
Also I found the phase between the output of discrete and cdp was 180 deg out so I just swapped Ro/Ron and Lo/Lon around for the test
Brent
The discrete pcb is tagged onto r21 etc and 12v regs with crocodile clip leads and therefore not ideal.
I am switching between two inputs on amp (cd67 rca and discrete o/p).
Also I found the phase between the output of discrete and cdp was 180 deg out so I just swapped Ro/Ron and Lo/Lon around for the test
Brent
rowemeister said:
Hi brent
very interesting
just finished layout for my fet version, except 2sk389 input
was going to etch pcb in morning, might wait at present
did you use all bc550's, including running as darlington on output?
i was thinking about the 2sk170 as driver on the output by itself!
and wondering if it was suitable.
allan
I had bc550 for first two and 2sk170 ripped from hdam for output.
I must say though I may have posted a little hastly!!!!!
Im just listening to Krall now and not to areas of the sound. The bass is defo less but when i flick over to opamps the bass feels overblown.
The Fets are a more natural sound, nothing over the top. mmmm
Im going to run them in before I fit to my 63
Brent
I must say though I may have posted a little hastly!!!!!
Im just listening to Krall now and not to areas of the sound. The bass is defo less but when i flick over to opamps the bass feels overblown.
The Fets are a more natural sound, nothing over the top. mmmm
Im going to run them in before I fit to my 63
Brent