Hi

Could you please help me with the queries enclosed in the attachment? It would be very kind of you. For high resolution copy of the attachment please follow this link: http://img521.imageshack.us/img521/2513/floydisolation10.jpg

Thank you.

Regards
PG

 

Re Q1:

That's a reasonable question concerning AC isolation.

I have a few isolation amps in my old spare parts collection. The capacitor isolation type (i.e. with a differential capacitor isolation barrier) I have is an ISO122P. While not specifically recommended for medical applications they still have some interesting specifications to allow an appreciation of their isolation capability.

The ISO122P has a very small capacitive value between the isolated parts - something of the order of 1 picofarad in either leg of the AC differential path. This means only very small currents can flow across the isolation barrier up to the rated barrier isolation limit which is typically 1500V rms. Such currents would not be hazardous either to people or sensitive electronic equipment.

I'll attach a copy of the data sheet in case you are interested.

 

Thank you, t_n_k.

Are my other queries so unreasonable that they were ignored?!

I still don't get one thing that why only small AC currents would flow across the isolation barrier in case of AC? Thanks a lot.

Regards
PG

 

With 1500V across a 2pF capacitor @ 60 Hz you'd have just over 1uA of current. Do you think that amount of current would hurt you?

 

Re: Q2

The author gets a bit cryptic (careless) with the phrase "lower frequency modulated signal or DC voltage".

I would recast it perhaps as "lower frequency AC or DC modulating [modulation] signal".

The low frequency AC or DC input signal is being used to modulate the higher frequency carrier signal.
You ask (in effect) - how could a DC input could modulate the carrier signal? No a problem as long as the DC signal is within the same dynamic range as a possible AC input signal. Consider the case of a frequency modulation [FM] technique. The FM carrier signal is shifted in frequency about a center frequency f0 [Hz]. With zero modulating input the FM signal transferred across the isolation barrier would have a frequency of f0 [Hz]. Depending on the modulation index the instantaneous modulating signal voltage value will offset the FM frequency by a particular deviation. Suppose f0=100kHz. Suppose also the FM frequency is shifted by ±5kHz for every ±1V of modulating input signal. With a fixed DC modulation input of +5V the FM transfer signal frequency would be f0+5*5kHz=125kHz. If the DC modulating input was -5V the FM transfer frequency would be f0-5*5kHz=75kHz. As long as the internal electronics on the isolated side of the amplifier can correctly demodulate the transferred FM signal back to an equivalent value corresponding to the modulating input signal then the system will work. If it works for DC then it will also work for AC up to some limiting modulating signal frequency.

If you revisit the data sheet I posted for the ISO122 you'll note on page 6 that the modulation technique is "duty cycle" modulation at 500kHz. This modulation method is probably easier to de-modulate than say an FM method.

Re: Q3

The lower the frequency used the harder it is to pass the signal via such a small (1pF) isolation capacitor - since capacitive reactance increases with decreasing frequency. The capacitive isolation reactance at 1kHz (for the input signal) would be 500 times higher than that encountered at 500kHz. So one would have to increase the capacitance proportionately (depending upon the lowest frequency AC input signal to be isolated) to allow effective transmission across the isolation barrier. As discussed earlier [post #4], if one had (say) a 1uF isolation barrier capacitance instead of a 1pF isolation capacitance then the effective dangerous AC isolation safety factor would be compromised significantly.

Also without modulation methods it would impossible to transfer a DC signal across the isolation barrier.

Re: Q4

You are correct - one doesn't intentionally combine common mode noise with the signal of interest. The fetal heartbeat signal is generated by the same physiological processes as the mother's heartbeat. That's the problem. Presumably there are measurement techniques that require the optimal placement of skin probes on the mother to produce the best measurement outcome. I recollect when one of my children was being born (in a somewhat distressed state due to his abnormal presentation position) a heart beat monitor probe was inserted through the birth canal and attached (somewhat crudely) to the baby's head - welcome to the world. My attention was very much focused on the beeping monitor and that little person's racing heartbeat. There was a whole team of doctors & nurses standing by at the delivery. He arrived safely - to his parent's great relief.