I am working with a AC skin conductance meter. See PDF file. The system will only measure the conductance part of the skin admittance. I am thinking about to include a calibration network R1,R2,R3, and use really(s) S1,S2,S3 to switch between measurement/calibration. As we can see from the schematic the current and voltage is very low. So I can not use any relay. But the that is note quite obvious for me is as follows. What is the consequences then you are close to the specified minimum contact load specification in current and voltage. I can live with somewhat higher contact resistance say ohms as long as the contact is stable and reliable

 

This might be a place to start studying up on switching very low level signals: http://zone.ni.com/devzone/cda/tut/p/id/3762

http://www.meder.com/bt-relay_uk0.html?&tx_jppageteaser_pi1[backId]=2773
http://www.rhopointcomponents.com/products.asp?category=REED+RELAYS&subcategory=Low+thermal+EMF+reed+relays
http://www.cotorelay.com/html/reed_relay_3500_series.htm
http://www.cotorelay.com/html/reed_relay_3600_series.htm

Ken

 

I would suggest either a good analog switch IC or using mercury-wetted reed relays.

Mercury wetted contacts avoid all the problems of 'dry' contact relays and will work fine at microvolts.

 

Hello,

Mercury wetted relays will be hard to find.
The ROHS does not permit the use of mercury anymore.

What voltage and current should be switched?
Should it be isolated from the main circuit?

Greetings,
Bertus

 

Except where there is no practical alternative..
Search 'mercury wetted reed relays' on Google.

One example of current types:
https://www.comus-intl.com/productpages/mercury_wetted_relays_us.asp

 

Hello,

Perhaps the US is not aware of the RoHS statement.

In europe some hazardous substances are not allowed anymore.

Greetings,
Bertus

 

RoHS applies to very specific categories of (mainly consumer) mass-produced Electrical & Electronic Equipment, NOT to everything electrical or electronic. Quote:

The categories of EEE covered by the EU RoHS are [5]:
1. Large Household Appliances
2. Small Household Appliances
3. IT and Telecommunications Equipment
4. Consumer Equipment
5. Lighting Equipment
6. Electrical and Electronic Tools (with the exception of large-scale stationary
industrial tools)
7. Toys, Leisure and Sports Equipment
8. Medical Devices (with the exception of all implanted and infected products) [Not covered by RoHS]
9. Monitoring and Control Instruments [Not covered by RoHS]
10. Automatic Dispensers​

It should be pointed out that, as of today, categories 8 and 9 are not covered by the RoHS.
Many people misunderstood and thought these products are exempted. Actually, they are not even covered by the RoHS which means they don’t have to obey the RoHS at all.​

--- End Quote.

And it is also completely irrelevent to anything home-made / not for mass production.
Even if this is a prototype for a production design, 'Skin conductance' measurements would appear to be Medical and/or Monitoring, so the equipment is not covered by RoHS...

This is why mercury-wetted relays are still in production and available world-wide.

Can we get back to the original subject now please?

 

Hello,

If you want to go solid state, have a look at the AQV- series from panasonic.

See the attached PDF's for more info.

Greetings,
Bertus

 

If you are switching into high input impedance circuits, one thing to look as in solid state switches/relays is off-state leakage. This is not an issue with electromechanical relays.

Ken

 

Hello,

KMoffet, I agree that hardware relays do not have a leak current.
I have pointed to the AQV series for the rather low leak current in the order of some nA.
Also the ON resistance is low, some ohms.

The image is taken from the datasheet from the former post.

Greetings,
Bertus

 

First off all I can agree that a mercury wetted relay perhaps is the best. But I have to depend on my local electronic part shop. And mercury wetted relays is not in the catalog. Besides I also prefer to take environmental considerations.
Well anyway I have redrawn my circuit so the magnitude of voltage and current are more easy to understand. The simplest model for the skin is a capacitor and resistor in parallel. But I will only measure the conductance part. I did not say that in the PDF file but I will use a frequency in range 40 to 200 Hz. Then I do a calibration, 10K fixed resistors will be used instead of the skin (Z-Stratum corneum) . The skin conductance be in range of say 2-200 uSiemens. So this will give the current floating through the M and C electrode switch. The input impedance of the LT1115 will limit the current floating through R electrode. My goal with this setup is to help out some psychologist students in a project. Skin conductance equipment is expensive. But by using a soundcard and some other parts I have from some older projets I am able to make a good solution. I was hoping to automate the calibration process. But perhaps it is better to have a calibration box that are plugged in manually instated of skin electrodes. But please keep posting your suggestions.