LClock XO PSU

Status
Not open for further replies.
Jump to Latest
C

carlosfm

Member
Joined 2003
#1
Hi,

I'm going to make a new PSU (including trafo) for my old LClock XO.
Although the clock must receive the cleanest PSU possible, I was always puzzled why LC-Audio was selling an unregulated PSU for this (at least at the time I bought this clock, around 1999).
So, what's the best option?

1. Unregulated: trafo + schottky diodes + caps.

2. Regulated: trafo + schottky diodes + caps + reg.

What reg?
Would a LM317 do the job fine?
I have some TL431s around here too.
Some better option?

For the diodes I have some here that I may use:
- MBR1100
- MUR1100
- BA159

BTW you can see the schematic of this clock posted by Lars here, posts #39 and 40:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=24059&perpage=10&highlight=&pagenumber=4

Thanks
 
#3
Yes, Peter.
If you look at the XO schematic posted by Lars (link on my first post), it also has regulation on board.
The question is: dedicated PSU: regulated or unregulated?
I ask this because in some cases double regulation is not better.
 
#5
"I made up my mind: gonna make a regulated PSU, with two outputs, reg. and unreg. and compare. "

That´s my man 🙂

Hehe, Carlos, yesterday I happened to scan thru the old "audible differences in amp" thread, you know the one which never wanted to end? haha so much fun I miss it!

/Peter
 
#8
According to the clock-monks, you

especially need a low noise supply for a clock-circuit to perform its best.
A TL431 nor LM317 are not the regs I (a novice) would use in such application. An LT1086CT would be more my choice. You could also have alook at the power supply used in Elso's clock design
Would be nice to to know the strenghts/weaknesses of an offspring of the LC-Audio and KWAK clock (Possible name could be The LC-KWAK' clock 😀 😀 :clown: )

Henk
 
#9
Hi Henk,

The Kwak clock has regs on board.
The Tent clocks too (XO2 and XO3).
The LClock too.

They are all different.
I just wonder why LC-Audio sells a dedicated unregulated PSU for their clocks?
Maby they found out that the regulation they have on-board doesn't need pre-regulation?
 
#11
carlosfm said:
Hi Henk,

The Kwak clock has regs on board.
The Tent clocks too (XO2 and XO3).
The LClock too.

They are all different.
I just wonder why LC-Audio sells a dedicated unregulated PSU for their clocks?
Maby they found out that the regulation they have on-board doesn't need pre-regulation?
Click to expand...


Hi

Mine is quite quiet (5 nV/SqrrtHz), I still wonder what noise the others exhibit - any measurements ?

LOW noise is key to any oscillator circuit I know of

this

http://www.premier-electric.com/files/STM/pdf/KFxx.pdf

circuit exhibits 150 nV/SqrrtHz (that is +30dB compared to above), so it is a nogo.

Seperate (and quiet) mains transformer / rectifier combo's are advised because of "full" galvanic isolation.

cheers
 
#12
the lowest noise "off-the-shelf" regulators seem to be the LT1761ES5-BYP and LT1964ES5-BYP from Linear Tech -- the company has a couple application notes showing how to use their low noise (Slew controlled) switcher chips with the above linear regulators to make a really quiet supply.

I don't know if they are available in Europe, but DigiKey has them for a couple bucks apiece. They are SOT-23-5 devices.
 
#13
With the regulation present on the LClock XO, I thought that I would not need more than an LM317.
Anyway, that's what I did, last night:

Small trafo + MBR1100 diodes + 2x2,200uf + LM317.
On the LM317 I have a 47uf cap from Adj. to ground.:devilr:
0.1uf poly on the output of the reg.

I haven't installed this PSU on the player yet, maby tonight.

I always use a separate trafo/PSU for any clock I install.
I just find it curious that LC-Audio sells an unregulated PSU for their clocks.

Guido, I'm a fan or your clocks.😉
MY Marantz CD52 + Tent clock (XO, for diyers😀, with your recommended regulation onboard ), with dedicated trafo/PSU/LM317 pre-reg, internal TDA1543 dac with trafo/PSU is a killer .
Deam, the player weights a ton now.:tons:
I've built all this player for myself, but now I have a guy that wants to buy it.😀
 
#14
jackinnj said:
the lowest noise "off-the-shelf" regulators seem to be the LT1761ES5-BYP and LT1964ES5-BYP from Linear Tech -- the company has a couple application notes showing how to use their low noise (Slew controlled) switcher chips with the above linear regulators to make a really quiet supply.

I don't know if they are available in Europe, but DigiKey has them for a couple bucks apiece. They are SOT-23-5 devices.
Click to expand...


what is the output noise level ?

do you realy consider switchers as an alternative ?

best regards
 
#15
Discrete regulators are better than "IC regulators" for sure, but KF's are quite well as they have a better than average noise performance and at the same time exhibit a 82dB ripple/noise rejection from the input.

Some IC regulators have better noise spec than the KFs, e.g. LT1964, but ripple rejection of LT1964 rated at 54dB only.
 
#16
jameshin said:
Discrete regulators are better than "IC regulators" for sure, but KF's are quite well as they have a better than average noise performance and at the same time exhibit a 82dB ripple/noise rejection from the input.
Click to expand...

You can get that value of ripple rejection with an LM317, with a big value for Cadj.
 
#17
carlosfm said:


You can get that value of ripple rejection with an LM317, with a big value for Cadj.
Click to expand...

yes, but the noise is about 10X greater. You can knock down the noise with the Wenzel circuit, but if going to this trouble and expense it would seem just as well to build the Super-Regulators.

I went to the ST site -- are the KF series only available in POS versions? What may make the KF desirable is that it is available in DPAK -- the darn SO23-5 devices are easily lost among the cigar ashes on my workstation.

The ripple rejection of the LT1761 with the BYP option is quite good --

As for low noise and switchers -- I have a quad of ancient Sorensen telecom supplies -- the noise was just above the threshold of my 5A22N scope plug-in. Of course, some Raytheon engineer probably worked for 18 months getting the noise out.

I think that the DIYr could get better noise results than a 7815 or 317 with the slew controlled chips and a linear post regulator, but probably not as good as the Super-Regulator with the AD797 AD811 AD825 (why not LT1028?) .
 
#19
jackinnj said:



The ripple rejection of the LT1761 with the BYP option is quite good --

As for low noise and switchers -- I have a quad of ancient Sorensen telecom supplies -- the noise was just above the threshold of my 5A22N scope plug-in. Of course, some Raytheon engineer probably worked for 18 months getting the noise out.

I think that the DIYr could get better noise results than a 7815 or 317 with the slew controlled chips and a linear post regulator, but probably not as good as the Super-Regulator with the AD797 AD811 AD825 (why not LT1028?) .
Click to expand...


Hi All

The ripple rejection is one item, but there will always remain output noise, even when you fee the input with zero moise.

I measure noise with 60dB preamp and a spectrum analyser. That way i can look at least 80dB deeper down than with a scope

I am talking LOW noise - be aware

To all:

Please quantify otherwise the discussion is useless

Ciao
 
#20
Guido Tent said:


Please quantify otherwise the discussion is useless

Ciao
Click to expand...

i don't know if measuring noise below 120 dB is possible in my environment -- the Jim Williams apnote on characterization of LDO regulators shows the DUT in a cookie tin powered by batteries -- I will have to fill up on Danish butter cookies!

I am in the process of rebuilding my HP 465A preamp -- it is 20dB and 40dB acurate from 10Hz to 1 MHz, but the distortion is 0.01 to 0.03% --
 
Status
Not open for further replies.
Top Bottom