EEVblog #909 - Defibrillator Teardown

[Brumby]

Ahhhhh. ..
You want to see the optimised DaveCad.

[filssavi]

The relays only have to interrupt current long enough for the SCR's to turn off, thus they will not have high voltage applied to them, and only have to break a current for a short time.

Problem with that is that the SCR will not turn off, if you open such a small relay it will arc, and without relay internaly designed to quench the arc  at that power levels(which those are not) the voltage drop across the arc will not be enough to turn the SCR off

I think you are underestimating how much an scr loves, to be turned on (or more properly speaking you underestimate the power of positive feedback) to swtch a normal SCR off it must be starved off current, so the voltage has to drop at least to zero (or there-about)




This is the structure of an SCR, the "important" region is the big n- in the middle, the junction associated with her are the only one in the whole device able to substain a significant voltage without breaking down (due to high doping the others will break down at 10V or even less), this is important because to bring one of that two junctions in reverse you will need to deplete the space charge region, form the carriers that are in there due to diffusion current in the active region, to do so you need either a negative anode-cathode voltage for a relatively short time or a very low voltage (lower than a threshold voltage roughly) and a lot of time the n region is quite big and has a very low doping (for high breakdown strenght) and so there is a lot of charge to take out, carrier densities when the device is on are much much higher than the doping atom's concentration, to get low losses the carrier lifetime is quite long and so relying on recombinations woks but is slow.

[Cerebus]

I think you are underestimating how much an scr loves, to be turned on

Dirty, slutty little SCRs. 

[NivagSwerdna]

It would be nice to get the audio samples out of the audio chip... looks like simple... isd4004... spi from a friendly microcontroller?

[rs20]

I wonder if the inductor is part of a clever dance to turn off the SCRs, rather than a PUT protection device. Consider that the circuit is in the first phase (red diagram in the video), and we want to turn the SCRs off and switch into the second phase (green diagram in the video).

1. Now, we turn on the bottom right IGBTs. This is not shoot-through, purely because the inductor is there -- it keeps the current steady, and allows the IGBT to succeed at pulling the Sternum point to ground.
2. With Sternum at ground, the Apex will fall to ground, and with zero voltage across them, the bottom left SCRs will turn off. Victory #1.
3. Now we turn the IGBTs back off. With the bottom left SCRs now off, there's nowhere for the inductor current to go, inductor current collapses to zero, and the top SCRs current consequently drops to zero and they will also turn off. Victory #2.
4. Now we trigger the IXYS and turn the IGBTs back on, and we've successfully transitioned into Phase 2 (green).

The only missing piece I see in my story is the inductive voltage spike that the inductor would produce in step 3, but maybe there's some sort of clamp/snubber/flyback diode somewhere.

[Pras]

I teach equipment for doctors and would be very very grateful if the role of the inductor in this circuit is clarified for me. In monophasic defibs, the inductor was hugely important in protecting the patient from a initial high current. However, I was told by an expert that inductors do not do this in modern biphasic defibs. So I am very surprised to see an inductor here. So, is this inductor there to protect components or patient. Grateful for any clarification. Thank you.


http://www.google.co.uk/patents/US5824017

[Pras]

Please forgive my ignorance, I am a medical doctor and not a trained engineer. I am trying to make sense of the diagram. Is it possible that in the diagram, the IGBT and the lower left TPS have got interchanged ?
 If you change it, then the circuit will make sense. Phase one, TPS after inductor and IGBT in new position come on. Then the IGBT is put off, and since there is now no voltage, TPS after inductor also shuts off. Phase 2: TPS (where current IGBT is located) gets switched on and IXYS gets switched on. Then sequence is terminated with IXYS remaining on, and IGBT (where lower left scr is in diagram) going on . This shorts capacitor terminating sequence. The inductor is there to protect the SCR from inrush of current from capacitor. Does this make sense ?                         

[Andy Watson]

At 22:18 in the video a diagram of the electrode placement is presented. This diagram reappears at 25:10 but it now has the Sternum electrode labelled as positive. I might have missed it, but I did not see where this information came from - I believe this may have been a wrong assumption in Dave's explanation.

This paper says that electrode polarity is not important: https://www.ncbi.nlm.nih.gov/pubmed/11719177 however these people have taken the sternum electrode to be the "cathode" (which I am assuming is relative to the first, main portion of the bi-phasic pulse): http://circ.ahajournals.org/content/91/6/1768.full and http://circ.ahajournals.org/content/91/6/1768?cited-by=yesl91%2F6%2F1768r91%2F6%2F1768

If we take the sternum as the negative then the order of operation that Dave gives must be reversed and the schematic (to my mind) makes much more sense.
I think the correct operation is as follows:
The "shock" cycle begins with the capacitors being charged to the required voltage. The relays will be closed and remain closed until the cycle is complete. The IGBTs will be turned on and the first positive pulse to the apex will be delivered by triggering the large IXYS device. This pulse is terminated by turning off the IGBTs - note these are the easiest devices in the bridge to reliably turn-off - the IXYS device will also turn-off. The negative portion of the cycle is achieved by triggering all the smaller thyristors - the ones labelled TPS. The negative pulse is terminated by either using up all the charge in the main capacitors, or by turning on the IGBTs to by-pass the patient. I think the inductor is there to limit the current under these latter conditions. It may also be possible that the two thyristors in series with the IGBTs can be used to deliberately discharge the main capacitors - again, the inductor would limit the initial current.

I don't not believe that the inductor in this circuit plays a significant part in limiting the current through the patient. I think the patient current is limited simply by the resistance of the body between the electrodes!  In the case of the older, mono-phasic defibrillators I am guessing that the inductor is present to shape the waveform, i.e. it is intended to provide "adequate" current for a longer period of time. To achieve the same adequate current for the same time with an R-C circuit would necessitate a much higher initial current - I think this is the sense in which the inductor is protecting the patient.

[rs20]

Does anyone have any problem with my explanation for the inductor presented in my previous message?

[bktemp]

Does anyone have any problem with my explanation for the inductor presented in my previous message?

Look at the link to the patent. It explains the operation in detail:
It works the way Dave showed in the video, but with both phases swapped: The IGBTs are used to interrupt the current in the first phase.

The patent isn't very clear, but it says the inductor is both inductive and resistive. It probably avoids exploding semiconductors if the electrodes get shorted and it is also used to dump the remaining part of the energy if the second phase is stopped before all the energy is used.