Can someone send to me the Elso's latest clock schematics (version6)? :)

[bitrate]

Do You think that thermal and contact voltage noise of this trimmer will not be converted by comparator into phase noise(jitter -in time domain)?

In my opinion resistance of this trimming resistor is far too big and i prefer two constant walue resistors acting as attenuator .

[6h5c]

Originally posted by kenev
Ray,

The trimmer on the inverting input does not adjust the offset but adjusts the symmetry of the waveform.

BTW, I think that the inclusion of the BC550 transistor in Elso's regulator is not at all necessary, as current draw is ca. 12 mA for the positive and ca. 5 mA for the negative rail (I have breadboarded the clock and measured currents) and TL431 can comfortably supply those currents without any need for current boost. At least, this is how I designed mine. Just my opinion.

Evangelos

Hi kenev,

As I use the oscillator without an output cap, since I have only a single supply available, there's about 2,5V offset present on the non-inv input. By applying the same offset to the inv. input the switching point of the comparator should move to the center of the wavform and make a 50% square wave. At least that's my idea, and it's the same as your's I think, only with a different explanation

The 431 circuit uses feedback to regulate the output voltage. The divider is at the output side and provides a feedback to the 431. Any noise produced by the 431 will be present at the output because it's a closed loop, correct me if i'm wrong.

The other circuit has the divider at the input side and doesn't use feedback. The voltage at the base of the BC550 is well filtered by the divider R1/R2 and C3. Nothing more than a passive RC filter. The transistor is a simple voltage follower that provides extra current. My guess is that this transistor produces less noise than a 431 + transistor combination. But of course it may sound worse, i'm hoping to find out

Greetings,

Ray.

[Elso Kwak]

Originally posted by bitrate
Do You think that thermal and contact voltage noise of this trimmer will not be converted by comparator into phase noise(jitter -in time domain)?

In my opinion resistance of this trimming resistor is far too big and i prefer two constant walue resistors acting as attenuator .

My goodness,
Do you really thinks so?
I am am using a Cermet Vishay trimmer.
Earlier versions had no trimmer but sounded less good.....

[6h5c]

Originally posted by Elso Kwak


For single supply operation I suggest R6= 1k5 and the pot 2k. The pot adjusts the tripping point of the comparator, hence the duty cycle of the wave form.
I tried it with and without pre-regulator!
I hope this answers your question. I tried many, many other things!!

O.k., thanks, I'll run a bit more current through that part, and I'll try it with the 431 supply also.

Ray.

[6h5c]

Originally posted by bitrate
Do You think that thermal and contact voltage noise of this trimmer will not be converted by comparator into phase noise(jitter -in time domain)?

In my opinion resistance of this trimming resistor is far too big and i prefer two constant walue resistors acting as attenuator .

Hi Bitrate,

Won't the 100n cap take care of that?
Two fixed resistors sounds good, at least it rules that part out.

Ray.

[bitrate]

I suggest to use low noise Vref rather than voltage divider directly from supply to adjusting duty cycle .

[Elso Kwak]

Originally posted by bitrate
I suggest to use low noise Vref rather than voltage divider directly from supply to adjusting duty cycle .

Did you try it?
Schematic??

[bitrate]

Originally posted by Elso Kwak
My goodness,
Do you really thinks so?
I am am using a Cermet Vishay trimmer.
Earlier versions had no trimmer but sounded less good.....

Thermal noise is Vendor independent so no matter if You use Vishay ,Bourns or other , value of resistor is important.

Originally posted by 6h5c
Hi Bitrate,

Won't the 100n cap take care of that?

In higher frequencies of thermal noise yes , but contact noise is 1/f noise so this cap do nothing with them.

[Extreme_Boky]

Hi bitrate,

Beautiful cat!!!

[6h5c]

Originally posted by bitrate


Thermal noise is Vendor independent so no matter if You use Vishay ,Bourns or other , value of resistor is important.

In higher frequencies of thermal noise yes , but contact noise is 1/f noise so this cap do nothing with them.

Hm, good point indeed. Maybe I will throw in an extra regulator and use a fixed offset voltage divider to see how that works/sounds.

How about using a LM317 with adjust pot to be able to trim the offset voltage? Does it have enough lf-rejection on it's adjust-pin?

Thanks for the tip!

Ray.