Hazardous current limitation

[Coen]

As soon as the DRL circuit runs into its 40 uA current limit, the power supply to all active electrodes is shut down. This prevents excessive current flow in case of multiple active electrode defects. In other words: the DRL current limit prevents too large currents between a defect active electrode and the DRL lead (Single Fault Condition), while the extra power supply shutdown prevents currents between different defect active electrodes (multiple faults at the same time).

The response time of the power shutdown circuit is in the order of milliseconds. Please note that this response time is only relevant in the unlikely case that two active electrodes break down at exactly the same moment, with the extra condition that one active electrode chip shorts its input to the positive power supply, and the other one shorts its input to the output. The more realistic scenario is as follows:

active electrode breakdown ----> DRL circuit limits current and shuts down the power supply to all active electrodes ---> further defects to other active electrodes cannot influence the safety situation anymore.

The DRL circuit can of course not "know" the origin of excessive currents. So, the safety circuit may indeed also be triggered by currents caused by RFI, ESD etc. This is exactly what we want the circuit to do: every current larger than 40 uA is considered unsafe, and should trigger the current limiter and active electrode power supply shutdown. Subject safety is the primary objective, even if the biopotential measurement is (momentarily) interrupted. However, thanks to the optimal isolation of the ActiveTwo front-end (battery power supply and fiber optic data transfer), leakage currents due to (HF) interference and ESD are typically much smaller than the 40 uA limit, and "false" triggering of the safety circuits is not a problem in practice.

Best regards, Coen (BioSemi)

[J.Contreras]

Dear Coen,

I really appreciate your answer, but I have to post again because the issue is still not clear for me.

Let’s assume the following:

*Your equipment has any kind of input over-voltage protection for every amplifier input.

*The input protection element has a limited capacity in terms of voltage, energy, response time, etc.

*There could be an over-voltage condition due to several reasons (ESD, Defibrillation, EFT, etc) that could put the protection in an over-stress state. This perturbation could damage the protection leading to a reduction of effectiveness in terms of spike limitation.

*The former could lead to damage a pair of amplifiers cable connection due to lack of protection.

*These two leads could generate a current through the patient that wouldn’t return through the DRL electrode.

Questions:
1)How could you detect this condition is there isn’t an increase in the DRL electrode current?
2)In any case, Do you mind that you measure currents in the uA range flowing through the DRL lead? I think that this is quite difficult to do in an accurate way without a very well designed circuit connected to the input stage next to the DRL active electrode.
3)Have you measured the output signal under this circumstance in order to detect it and power down the failing stages?

Thanks again, and have a nice day.

Jorge

[Coen]

Question 1): Scenarios like the above will always also lead to increased current via the DRL lead, since the DRL essentially is a (current limited) low impedance connection to the amplifier Common. In case of a situation with multiple defects the current sinks and sources will not cancel each other completely, and therefore some current will leak via the DRL lead. This increased current is detected, and the power supply to the active electrodes is shut down

Question 2): The measurement circuit is actually quite simple, and very reliable. I do not want to publish the full circuit details on this public forum (competitors might be intersted too), please email me directly. Anyway, the information given in http://www.biosemi.com/publications/pdf ... ection.pdf and http://www.biosemi.com/publications/pdf ... uction.pdf will point you in the right direction.

Question 3): The beauty of the DRL based error detection is that we do not actually have to scan all the electrode outputs individually to look for problems. Also, in case of excessive leakage currents current, we shut down ALL active electrodes, which again makes the safety circuit simple and reliable (also see viewtopic.php?t=32)

Best regard, Coen (BioSemi)