IEC EN 60601-2-27 subclause 50.102.5 ECG Multichannel Crosstalk

I am having difficulty trying to meet part of the 60601-2-27 (2nd edition) subclause 50.102.5 Multichannel crosstalk.

The subclause states:
"When an input signal limited in amplitude and rate as per 50.102.2 is applied to any one LEAD WIRE of the multi-channel EQUIPMENT, with all other LEAD WIRES connected to the N (RL) LEAD WIRE, the unwanted output in the unused LEAD WIRES shall not be greater than 5 % of the applied input signal.

[...]

Close switches S, S2 and S4. Set the EQUIPMENT SENSITIVITY to 10 mm/mV. Connect LEAD WIRES F(LL), V1, and if provided, the Frank E to the R (RA) LEAD. All other LEAD WIRES are connected to the N (RL) LEAD WIRE. [...]. From the signal generator, apply 2,5 mV peak-to-valley 30 Hz triangular wave between the F (LL), V1, and if provided, the Frank E and the N (RL) LEAD WIRE. Set the EQUIPMENT SENSITIVITY to 10 mm/mV using any display device. The output in the other channels shall be ≤ 5 % peak-to-valley of the full scale output."

The problem I'm having is that some of my inputs V1-V6 are referenced to a Wilson Centre Terminal voltage derived from an average of RA,LL,and LA signals. If I apply a voltage as stipulated in this clause to a subset of RA,LL, or LA, I will have a significant voltage that shows up on leads V1-6, which is the way it's supposed to work in the field.

However the way the clause is written, I will never be able to pass this test. Am I misinterpreting this clause or is the way my circuit is constructed improper to satisfy this clause? Is this going to be changed in the upcoming 3rd-edition release?

Thanks!
Jesse

Re: IEC60601-2-27 ECG Multichannel crosstalk

Hello Thread Poster,
I could not pull out the 60601-2-27 so can only comment from behind a fence.
The Wilson terminal configuration on V1-to-Vn is likely not the problem, because another ESG product is sold in Europe.
I base this on knowing another ESG product with the CE mark and your describe Wilson configuration.
The elivated level due to your description and design is of a floating negative side. If the other leads are all connected to the neutral leg, then there can be no input signal present on the other legs, (LL, RL, RA, LA, and all the Vs). The outut of your ESG driver should not have any either, except for the 2.5mV driven channel. Because of the very small level of the signal, the noise present and ground currects can effect a measured display. Measuring a 0.5mVp-p signal at 60-150 hz is no easy effort via oscilloscope. I last used the 7A22 differencial plug-in on a Tektronix scope, which could effectively view the signal. Today's scopes have 1mVp-p per div. minimum gain displays. High input Z and 100uV/div is a rare display product.
In Short: This can't be the final word, but you likely have a method of test problem.
I hope you get a better answer to your description.

Re: IEC60601-2-27 ECG Multichannel crosstalk

Hi george,

The issue isn't a signal measurement issue -- I have measurement capability down to less than 0.1 uV.

Basically the test, as I read it, wants me to apply a 2.5mV signal to only one of RA,LL,or LA with the others connected to neutral (RL). When I do this, I get a 2.5mV signal on, say, RA electrode and a 0mV signal on LL, LA, and the V1-6 electrodes. So my circuit setup seems good.

But... the V1-V6 leads go into instrumentation amplifiers with the + input connected to the leads and the - input connected to an analog signal that is the average of RA,LA,and LL (the Wilson Centre Terminal). So when do the test, I will get 0mV on the + inputs into the inamps and 2.5mV/3 into the - inputs. That's exactly what should happen but the test wants less than 5% signal on V1-V6, which is impossible.

There must be something I'm missing. Hope it's not too confusing.

Thanks for the reply.

I re-read your comment and again the instructions.
The description in the first post indicates connecting LL to V1, but I would understand that to mean all V inputs.
Connecting all V inputs and the LL to RA, and connecting others to neutral, (RL).
You would then have 2.5mV on all V and the RL input. You would then look for <5% on the remaining leads,
(RL=neutral,LL=2.5mv,RA=2.5mv,LA<0.5mv) This would work.
I count RL, LL, RA, LA, V1, V2, V, V4, V5, V6, There must be a few others. 12 is common.
Just a thought. You would not be looking at the V input leads in this setup.
Another deviation would be to connect the 2.5mv signal to the V input(s) without the RA.
This would avoid the average negative input signal going to the V inputs.
I would bet there is a dispute to this condition, and a compliance work-around has been documented.
I am sorry I don't have/know of it. As you mentioned. What about the next rev. is a good comment.
Good luck on this. It is a condition common more than people talk about. Another has PASS'ed, so likely you can.

I have been working with both IEC 60601-2-27 and IEC 60601-2-51, and it seems there are several clauses which have been written with some object in mind but the test methods have not been "tested" for practical implementation. This is one of the tests that need to be improved to specify exactly which voltages should be applied to which lead electrodes in a table format, taking into account the networks in a typical ECG.

For chest lead testing of 12 lead ECGs, I normally test by applying a voltage to one chest lead and confirm that the rest do not show any indication above the limit, and repeat the test with at least 2 more chest leads (in total V1, V3, V5). If cross talk exists, it will normally show up in adjacent channels.

For a 5 lead ECG (only V1 available) I would put the same a voltage on RA/LA/LL and confirm no indication on V1. However, this method assumes the sensitivity of RA, LA, LL is the same and the Wilson center terminal network is accurate, both of which are reasonable assumptions from experience and typical hardware/software implementations in ECG systems.

Interestingly, the requirement and the test methods in clause 50.102.5 (Multichannel crosstalk) seem to be misleading.

The requirement describes such that the input signal is applied to any one lead of the multichannel equipment, all other leads are connected to the N (RL) lead.

However, the test methods state that connection lead F (LL), V1, and if provided, the Frank E to the R (RA) lead. All other lead are connected to the N (RL) lead.

Regarding above mentioned setup, which one shall be followed?

As mentioned, this is a mild problem with several clauses in IEC 60601-2-27 where the test methods are not fully fitting the requirement or are reasonable in practice. To make matters worse, I found the IECEE test protocol does not match the standard either in many places.

Anyway, for multichannel cross talk, I recommend to understand the principle or intention of the test and adapt the test method to suit the design of the equipment under test.

For example, if you are a manufacturer, you could inspect the raw input channels before any lead networks are applied. This would allow you to apply the signal to one lead electrode at a time and verify there is no cross talk to other input channels.

Modern electronics has extremely low levels of cross talk so this test is really a formality anyway.

Hi Folks,
We are developing 5 lead ECG device wiith WCT configaration.

Leads we have in our configaration:- RA,RL,LL,V4,V6

We are doing multichannel crosstalk as per 2-47

LL has to connect to +ve lead
RA,V4,V6 has to connect to -Ve lead
RL to regerence point

This means that when i am giving an input signal to Lead 2, i should not get any signal on my V4 and V6 as per the standard

But because of the drive i am getting a clear signal on V4 and V6 theese channels also. Can you please help me in resolving this isue.

The result is correct.

For testing multi-channel cross talk, actually it is not possible to test completely because of the Lead relationship matrix (e.g. Lead I = LA - RA) including the Wilson Central Terminal. Ideally, we should test cross talk prior to the matrix being applied, e.g. in special software if the matrix is applied in software, or in hardware by inspecting the raw analogue signals.

By relying on the leads we can only apply limited tests, as follows:

Signal on LA: check no signal on display Lead II
Signal on LL: check no signal on display Lead I
Signal on RA: check no signal on display Lead III
Signal on Cn/Vn: check no signal on displayed leads except Vn

The latest version of IEC 60601-2-25 has been corrected for this point to follow the above pattern.