EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: PyroFire on January 27, 2024, 08:10:50 am
-
Hi, I am new here with a power electronics background :)
I've been reading topics related to Isovu and other isolated probes and became quite curious as to why the isolation probes have taken a route for analog isolation.
Sounded like such optical isolation is not new, but calibration across temperature and offset is quite challenging.
What are the pros and cons of digital isolation? To clarify, I've added two block diagrams below.
[attachimg=1]
[attachimg=2]
Would bandwidth limitation be the main limiter for having an isolated amplifier + ADC + some form of fast communication back to FPGA?
-
It's a question of power/bandwidth. Do some calculations as to what power it would take to digitise an analog signal at lets say 10..12bits and 2 or 3 GHz sampling rate. At these speeds ADCs are parallel output, so you need to serialize the stream (even more speed) to be transmitted over 1 fiber, also the clock needs to be transmitted to the receiver too for sync purposes.
Eventually you will realize that many Watts of power are required for this kind of digital transmission, which is incompatible with a fully isolated probe head as needed in these applications.
So it's better so go the analog way and solve the offset/amplitude calibration by other means, when these probes are stable in temperature the calibration is fine and in some of these products calibration takes only 3..5 seconds and can do so even with an input signal live.
The digital approach would make sense for low (<10MHz) bandwidth probes
-
https://cleverscope.com/products/CS1200
I was reading the spec for clever scope's probe (which I think doing analog to digital conversion in the probe head) and 14 bit / 500 MSPS gives 4W. 2 x 5.5AHr cell is needed, which gives 10 hr. May be I am making a wrong assumption somewhere but 5W seems manageable through power-over-fiber?
-
PoF cannot do 5W unless you transmit multiple laser beams.
That probe uses 2 huge 21700 5.5Ah batteries thus making my point about power consumption also considering its "only" doing 500Msps for 200MHz which is not that much, I'd like to see the transient response of that probe.
-
0.6W is what Broadcom currently offers in their AFBR-POC series, and laser+receiver is like >$1200