Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Connection of an old Electroencephalogram to a A/D converter

Status
Not open for further replies.

Gustavo Zubieta

New Member
Hello experts electronic gurus:

I am a 55 year old physician with passion for electronic knowledge. I have an old 8 channel Electroencephalogram Mingograf EEG Junior from Siemens-Elema, that previously used ink to register on paper. I want to digitize it to a PCL-711S PC Labcard. I have been able to connect Ekg and other equipment prefectly, before.

But now I decided to connect the EEG from the point of connection to the ink dispensers that injected ink to paper. The voltage reads around 3.5 VDC. But when I connect to the PCL-711, the voltage goes up to 7 VDC. So I used a voltage divider with two 1.5 Mohm resistors in series but when I again connected it, there is still no reading and eventually the main fuse of the EEG blows out.

1) Can someone please tell me what is wrong?

2) I figured I may have to use an optocoupler or optoisolator so I was thinking of using something like a 4N35. I would feed the voltage from the EEG to the led and then I would have variable voltage output at the phototransistor. Is this a good idea and a feasable one?

Thank you so much.

Gustavo
 
Hi Ya Doctor

Well a possible problem may be found in here:

Analog Input
• Channels: 8 single-ended• A/D converter: 12 bit, 25 msec. conversion time
• Input range (V): ±5, ±2.5, ±1.25, ±0.625, ±0.3125
• Trigger mode: Software, pacer or external trigger
• Data transfer: Program control, interrupt (IRQs 2 ~ 7)
• Accuracy: 0.01% of reading ±1 bit
• Common mode rejection: 60 dB typical
• Input impedance: >10 MW
• Input overvoltage: ±30 VDC max.

The analog input channels are single ended which I take to mean they share the same common. Your system likely has the pens driven with a differential signal. That could be the problem and maybe explain what you are seeing.

Pretty cool though in that I would venture a guess that the channels could be fed into a analog input card with 8 or so differential input channels, plotted and recorded.

Just My Take
Ron
 
I agree with Ron that you appear to have a differential signal from the pen driver which, when you are ground them, cause high currents that eventually blow the fuse. The is also indicated by your reading of 3.5 VDC with a voltmeter but it jumps to 7 VDC when you connect the ground.

Basic opto isolators such as the 4N35 are non-linear so would probably not be a good choice to provide the coupling.

The best would be if you could get a Labcard with differential inputs. Alternately TI and others sell differential/instrumentation amps that should also do the job such as this.
 
You could consider this opto coupler device.
 

Attachments

  • HCNR200.pdf
    236.8 KB · Views: 206
Interesting and critical observations

Dear Ron:

Thank you and Carl for your interesting comments!! I am very impressed. I had not thought of that differential input.

I was thinking of trying another A/D serial port from DATAQ Instruments DI-194RS. But I just realized it is also single ended. I have another A/D converter that I have been using over many years. A Burr-Brown PCI-20000. It is very solid and functional. Wish I could get some more in case the one I have fails. But I believe they no longer produce them.

That may be the solution. And it would be cool!!!

Can anyone suggest what differential 8 channel A/D board is best for a laptop? I guess with a USB port.


Oh and thanks also Ericgibbs. That is likewise a good idea. Unfortunately that would take longer as I would have to order it and then set it up. But thanks. could use it in some other fun project.

BW,

Gustavo
 
Last edited:
Hi Ya Doctor

Actually as long as you mentioned Dataq they do make a few that offer USB interface and a choice of for example 8 Differential or 16 single ended inputs, the Dataq DI 710 comes to mind. Nice thing about Dataq is I really like their software. Good stuff for openers and they offer some free SDK (Software Developer Kits) for those who like to roll their own. I have also been looking at Measurement Computing stuff lately. They have some nice looking stuff also but I have not tried their stuff. A good part of this is a mix of budget and how many bit A/D you go with. That with the sample rates tossed in the mix. Going from what you mentioned to 14 biut with 8 differential input channels would be nice. This all goes back to cost. Personally anymore I prefer stand alone modules with a USB interface over PCI Card types where I need to devote a PC to the thing.

Beyond Dataq, and Measurement Computing, there is NI (National Instruments) and a host of other manufacturers. It's about budget and bang for the buck.

Ron
 
Thanks Ron,

Very useful information. I will look into those.

I have prepared the connections from the EEG to the differential A/D converter Burr-Brown (12 bit).

But when I make the measurements:

Each pen driver has 3 poles. One (white) common to all and one Green and one yellow.

When I test the A/D ground with each of these I get the following voltages:

A/D ground to White (common) -7.66 VDC
A/D ground to Green - 7.00 VDC
A/D ground to Yellow - 7.15

When I check EEG ground(from Chassis) to A/D Ground Voltage 0.00 VDC

Is it safe to go ahead and connect the EEG to the A/D ?

Can you please suggest if I should go ahead and connect the

A/D high yellow
A/D low green
A/D Ground to the white (common cable)

This may be a difficult question to answer. I am sending a photo of the pens of the EEG for clarity.
**broken link removed**
 
Last edited:
You can only connect the A/D ground to the white wire if the A/D ground is floating from the EEG chassis ground.

Measure the resistance between the A/D ground and the EEG chassis. It must be very high if you want to make the above common connection. If not, then you will need differential amps to make the connection.
 
Hi Carl:

Thank you for your comments. I have tested the voltage between the Chassis fo the EEg and the A/D ground and there is O VDC.

1) Can you please tell me what you mean by "if the A/D ground is floating from the EEG ground"? does that mean connected?

2) One other question please. Do I measure the resistance with the equipments turned on?. I tested with them off and there is no resistance, in other word open circuit.

Please forgive my ignorance in some of these aspects.

BW,

Gustavo
 
Last edited:
I have a question doc. Do all pens always write or can you select individual pens? Do the pens use a pen lift? Just curious after looking at the pen head assembly.

Ron
 
Hi Ron:

They all write at the same time. When they worked they actually injected the ink through a very fine capillary tube. Beautiful for its time. But they clogged up form lack of use. So when we tried to fix them, two broke. so now I want to give it a try with a A/D converter.

Gustavo
 
Thank you for your comments. I have tested the voltage between the Chassis fo the EEg and the A/D ground and there is O VDC.

1) Can you please tell me what you mean by "if the A/D ground is floating from the EEG ground"? does that mean connected?

2) One other question please. Do I measure the resistance with the equipments turned on?. I tested with them off and there is no resistance, in other word open circuit.
1) By floating I mean there is no connection between the two grounds.

2) It should be the same either way, but you want to measure the resistance with everything plugged into the wall power. Power on is probably better.
 
Leave original load of end stage in place!

Please do keep in mind that the output amplifier is designed for high power to a low impedance inductive load specifically designed for it. So it must be kept in place in order to measure --- and thus to digitize --- correct voltage levels with any A/D adapters. They usually have a very high internal resistance. Removing the original load will make you measure an exorbitant constant 'ceiling voltage' no longer controlled by the signal you want to measure. That is because the output actually is a so-called 'current loop'. If you still want to remove the pens, the problem could be resolved to some extent by shunting the output with a resistor of the same value as the resistance of the pens. But that will not mimic the pens' inductive resistance and will alter frequency response.
A better idea is to connect your A/D adapters to the input, rather than output, of the amplifier end stages. You will probably need schematics of your system in order to locate them. Good luck!
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top