differential clock

Status
Not open for further replies.
Jump to Latest
#142
EUVL said:
Show me a photo. Maybe I can help you with some ideas. Am mechanical by profession.
Click to expand...

Thanks, this is the motherboard, analog filter boards will be mounted on the back side.
Ground connections will be made via the "Abstandshalter".
The motherboard will be mounted in a frame that fits a case from W&G.

I have a milling machine and could make my own design case but the machine allows only 6 mm max. tools and it is a lot of work...

motherboard1.jpg
 
#144
> Measured: Supply +/- 2,3 V, junction of 1k and FET + 1,1 V.

I asked because I want to just use +5V and Gnd, like Elso Kwak. I think I will, as I want to use an additional buffer, like Jim Hageman's Hagclock.

> Perhaps to have the ability to tune to the exact desired frequency which is not important for audio.

I recalculated everything today based on the balanced circuit, 22MHz and using J310 JFETs. I would end up with either 110p/110p (100p/120p) for feedback plus 220p from gate to Xtal on each half, or feedback caps only 91p (82p/100p). So your values are not far off. My Xtal has 22p CL, and 35R Ri, according to Segor.


Patrick
 
#152
EUVL said:
At 4 euros a piece, you can afford to test.
Click to expand...

Where ?

Did you see the EP series ? Works from 5 V and 3,3 V. Flipflop and buffer are faster than EL series, have jitter specified = 0,2 ps.

I reviewed all my datasheets and found some new coincidences.

EP series: Jitter +/- 20% = propagation delay / 1000 --- EP11: 220 ps = 0,2 ps
E / EL series: Jitter +/- 20% = propagation delay / 500 --- E131: 500 ps = 1 ps
ELT / EL series with large swing: Jitter +/- 20% = propagation delay / 100 --- ELT23: 3500 ps = 35 ps
 
#155
> Where ?

OK 5.50 at Digikey. You can still afford. 😉

> Did you see the EP series ? Works from 5 V and 3,3 V. Flipflop and buffer are faster than EL series, have jitter specified = 0,2 ps.

Yes, and I have ordered EP whereever I can. But they don't have everything that I need in EP.

> From ONSEMI TND301 appl. note

Well found. 😉

> Will the jitter be reduced with parallel devices ?

In theory when the error is random it will reduce with parallel devices. The questions is in this case however not how much jitter, but how well matched are the devices in propagation delay tolerances (usually +/-80ps or so). And since they are all voltage output, you end up smearing the rising edge of the signal over 80ps as signals from different paralleled devices arrive asynchronously. Not really clever, it is ?


Patrick
 
#156
EUVL said:
In theory when the error is random it will reduce with parallel devices. The questions is in this case however not how much jitter, but how well matched are the devices in propagation delay tolerances (usually +/-80ps or so). And since they are all voltage output, you end up smearing the rising edge of the signal over 80ps as signals from different paralleled devices arrive asynchronously. Not really clever, it is ?
Click to expand...

I wonder what happens with my 16 DACs per channel switching at the same time...
 
Status
Not open for further replies.
Top Bottom