I need to design ADF4350 synthesizer that can use either one of two TCXOs. Do you have any switch that can do this job?
The connection is as follows:
TCXO1 ----->|Switch |--------->ADF4350
I assume that the switch shares the supply voltage with the ADF4350(3.3V +/-10%) and the TCXO signals are single ended. If that is the case, you can use any of the following switches: ADG619/719/749/779.
For diferential TCXO signals the ADG636/736/787/772 can be used.
Thank you for your help. My switch shares the supply voltage with the ADF4350(3.3V +/-10%) and the TCXO signals are single ended.
Since ADG619/719/749/779 are analog switch, they should not add jitter for the TCXO reference signal, right?
Thank you in advance.
ADG619/719/749/779 analog switch not add jitter for the TCXO reference signal.
PSB will be carefully created only.
Thank you for your reply. What do you mean PSB? FSB suggested ADG619/719/749/779 analog switch does introduce jitter, I am confused a little bit. Maybe the jitter is due to the ON resistance uncertainty.
PSB is printed circuit board only .
Jitter is very-very low, as at passive component (as resistor with equal value ) Don't worry!
The analog switch is not jitter free, but if you choose a device with the lowest ON resistance, it will minimize the random jitter due to the noise introduced by the switch resistance. On the other hand, the bandwidth of the switch has to be sufficiently high to pass the signal with minimum distortion to reduce the amount of deterministic jitter introduced by the switch.
Do you have any recommendation which switch to use (with lowest ON resistance) and least jitter. I have looked at ADG839 with 0.6ohm ON resistance. But most of parts do not provide jitter information.
Do you have specs for jitter performance of those switches?
I'm not familiar with TCXOs, but I assume the signal they provide is either a sine or square wave. Would it be possible to share with us some more details about the TCXO signals such as frequency and rise/fall time(square wave only)?
The TCXO we use is 26M Hz TCXO with chipper sine wave output. The part is from NDK. Please see the attached PDF file. The chipped sine wave is perfered since it will feed to the ADF4350 synthesize as reference clock. Squrewave has a lot of harmonics.
The reason to switch between two TCXOs is that sometimes we need to use external 10M sinewave input instead of this TCXO for synchronization purpose.
So in this case, we need to switch between this TCXO and external 10M refrence input (typicall Sinewave, from Test equipment).
Thank you very much.
Since you want to select between a 10MHz size and the 26MHz TCXO, I would suggest to start by estimating the bandwidth of the 26MHz clipped sine wave from the TXCO and choose a switch with the lowest On resistance that meets the bandwidth criteria.
The ADG839 has a bandwidth of 25MHz which is less than the fundamental frequency of the TCXO. I would recommend instead the ADG849 which has 38MHz bandwidth with a similar ON resistance or the ADG852 for 100MHz with higher ON resistance of about 1.5 Ohms.
ADG852 is a good choice for my application. Do you know what kind of jitter performance for this analog switch?
The footprint of ADG852 is not common. After some more research, I found popular ISL84714 (Intersil) and MAX4714 (Maxim) with footprint SC70. ISL84714 have 0.4 Ron and bandwidth is >100MHz.
Sorry to bring other company's part. I am trying to find the best solution for ADF4350 using two reference clocks.
What was the criteria you have used when deciding which part to use?
Was it the package? If so, is SC70 preferred to the smaller LFCSP? Or was the pinout the deciding factor? Why is the pinout of the ADG852 not suitable for your application?
Have you checked the bandwidth of the Intersil and Maxim parts?
The criteria if of course the performance. Which means least jitter will be added in the clock chain. Let us just discuss the performance. Did Analog Device test the ADG852 to see how much jitter is added on top of the reference clock?
An example circuit is:
TCXO->ADG852->Agilent clock Jitter analysis Instrument
For my case, the decision factor is the jitter performance. I have checked Intersil and Maxim parts and they are about 100M bandwidth too.
In general, analog switches are not tested for jitter performance and the ADG852 is not an exception from this rule.
I understand that performance is the key criteria in choosing the switch, but you have mentioned in your previous post that the ADG852's footprint is not common.I was just wondering if the pinout was contributing to the decision to remove the ADG852 from the parts considered for this application. Also, I would appreciate if you could explain what did you mean by "The footprint of ADG852 is not common."
I'm glad I found this thread! I'm looking to do a very similar switching approach to Lotus.
Is there a single supply switch that can handle a 0.8Vpp clipped sine? 20 to 60MHz?
It looks like the ADG619/620 can do it....but that is with dual rails (+/-5.5V).
Retrieving data ...