AD725 input crystal frequency options

Looking at the spec.  4FSC Vhi_min = 2V so a 3.3V oscillator should be able to drive it.  To be extra safe you can always buffer the oscillator output with something like a ADG3301 and match the power domains to the levels you are interested in.

can you kindly suggest crystal part numbers suitable?

I cannot find ad725 eval board bom and schematic too.

what about the circuit used? is the interlock between them ok?

I have found these crystal oscillators, please check:

NTSC:

https://www.digikey.com/en/products/detail/ecs-inc/ECS-5032MV-143-CN-TR/9953235

https://www.digikey.com/en/products/detail/ilsi/ISM97-3251AH-14-31818MHZ/12608883

PAL:

https://www.digikey.com/en/products/detail/sitime/SIT8008AC-23-33E-17-734475/15222498

I could not find crystal based oscillator for PAL, only MEMS based one.

please check and advise.

also, please check the image in first post for enabling and disabling them... will it work?

PDF

The 14.318MHz oscillators are fine.

I would not use a MEMs oscillator for 17.73MHz oscillator.  These oscillators are used to generate the color burst and MEMs devices in general have too much jitter.  They might work for limited temp ranges.  You can order 17.73MHz oscillators direct from the manufacturers, it's just that they aren't stocked by digikey.  You can also search findparts.com and trustedparts.com for stock.

FYI, Since the AD725 was released 25 years ago, crystal manufacturers have stop making 5V oscillators, the best we can source is 3.3V oscillators now, therefore we need level translators.

I've attached the user guide with schematic for the AD725.  You'll notice it uses a mechanical switch to select oscillators but you can use a silicon switch also.

hello

the problem is when designing boards and making them with pcb makers, it is better to have digikey\mouser stock for sure.

do we really need level translators?? I think someone years ago mentioned it does not, and it can run 3.3v just fine. please check.

assuming I use that mems oscillator in good conditions, like just room temperature. what could go wrong?

According to the AD725 spec Vih_min = 2V with Vcc=+5V.  As stated above, as long as the oscillator output exceeds 2V then it should work.

I understand the issues with part availability, I face them all the time.  

It is recommended to use oscillators for the 4FSC input for accuracy and stability issues.  Of course you can design the system anyway you want to.  Personally I'd will not use MEMs because of potential problems.

hello

can you post a suitable crystal number from a good supplier rather that digikey and mouser?

is there also another option? like frequency dividers and so on?

Digikey/Mouser are just distributors, they don't make oscillators or crystals.  I've spec'd Abracon, ECS, Epson to name a few.  Start by selecting the target frequecy, then ppm, then voltage, then package.  Filter again those and you will find a series you like.  For example if you need a 17.73MHz oscillator and it's not stocked by digikey and you like the Abracon ASFL1  voltage/package then you can order the exact part from from Abracon.  It may take a bit of time but it is doable.  There are lots of options taht need to be sorted through depending on you system

If you find a oscillator that's an integer multiple of the target frequency then you can use dividers.  Fractional divider can introduce jitter via the PLL.  There are some fractional divider that should work but they be more expensive that just a plan oscillator.

hello

i found this one for PAL: https://www.digikey.com/en/products/detail/epson/SG-8018CG-17-734475M-TJHPA0/13661534

it is non-stocked in digikey but mouser allows to buy it: https://eu.mouser.com/ProductDetail/Epson-Timing/SG-8018CG-17.734475M-TJHPA0?qs=zW32dvEIR3sW4Eud%252Bsosow%3D%3D

so based on datasheet, the crystal is outputting its frequency when OE pin is High and it doesn't when OE pin is Low. right? and similar functionality for the NTSC crystal linked above,,, please check if my circuit can work or need some tweaks.

The part should work.  Note I see 20 week lead time  on the part.  You might contact the manufacturer direct.

Pulling OE high will enable the output.  Note that the OE pin has a weak pull down so you'll need a pull up to 3.3V.  10k will work fine.

hello

what does "weak pulldown" even mean? I will pull it up to 3.3v using 10k or similar for sure, that is when I want it to be active. but when not active I pull it down to 0v using 10k as seen in my picture above. please check.

lead time is ok for now since first step is making a test board.

I am still waiting answers to other questions I posted in this section though.

To me a weak pull resistor is >= 100k,  medium pull is ~10-22k and strong pull is <=2K.. Some devices have weak pull resistors on their inputs so the inputs are not floating which can cause problems.  This OE has a weak pull down to turn off the output when OE is not connected to anything.  If you add a 10K pullup to the pin then the pin will have 3.3V * (100/(100+10) = 3V which is greater then Vih_min turning on the output.

To me a weak pull resistor is >= 100k,  medium pull is ~10-22k and strong pull is <=2K.. Some devices have weak pull resistors on their inputs so the inputs are not floating which can cause problems.  This OE has a weak pull down to turn off the output when OE is not connected to anything.  If you add a 10K pullup to the pin then the pin will have 3.3V * (100/(100+10) = 3V which is greater then Vih_min turning on the output.

hello

so you recommend putting 10k or similar? in my circuit above, I will feed 3.3v directly via analog switch without any pull-up resistors... but without the switch being active, I will put 10k as pull-down resistor. what do you think?

You can pull directly to 3.3V via the switch.  The 10k pull-down will just go in parallel with the internal 100k pull down.  Not a problem.