Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Designing a current source for an unknown load

Status
Not open for further replies.

afacanboy

Newbie level 4
Joined
Nov 30, 2009
Messages
6
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Location
turkey
Activity points
1,350
Dear Member,

I am trying to design a current source for a medical device. The load is unknown but in megaohms range (body impedance).
I have 200V voltage source. I generate this 200V with royer topology. The current range is between 0-50mA.

The waveform shape is pulse...
The current can be set via GUI. The body impedance changes according to tissue (in Megaohms range).

I have tried this using current shunt monitors but failed. I want to set this current from high side...

I attached some schematic files...

200V Generator - Royer Topology.PNG

Patient electrode connectors.PNG

Could you please help?

Regards.
 

I think your math is off. If you want 50mA through one megohm you need 50000 volts. Please don’t connect this device to living things.
 

Hi Barry,

I know math. But this is a different case. I connect this device more than 5 times with current design.
But there is low side control to adjust the current (for the current design, there is an OP-AMP to adjust current. I attached the related schematics).

CURRENT SINK.png

Low side control causes some issues. Because low side is not GND. That is why I want to control the current from high side.
 

You said Megohm range load, 200V supply, 50 mA current range. Which of the three specs should be corrected?
E.g. 10 mA can be achieved with up to 20 k load resistance.

The load resistance limitation is basically the same for "low side" or "high side" current source, just a matter of Ohms law.
 

Barry and FvM,

First of all, thanks for your responses.

Ok. I may misundersttod an article. I remember that the article says that the body impedance changes between 100k and 1M ohms.
But I just measured the resistance, its about 13k ohms. I measured this value with a stick electrode. This may affect the impedance.

Now, any idea?
 

The problem is;

If you look at the circuits above, there is no control at the positive channel. It directly supplies the voltage to the body.
And after the negative electrode, there is a current sense resistor, which is 32.4 ohms.
Assume that the user set the current level 10mA from Graphical User Interface. Then the voltage accross the 32.4 ohms should be 324mV.
The non-inverting input of U20A (TLC277) is set to 324mV and TLC277 will arrange 10mA goes through positive channel - human body - negative channel - 32.4R - GND.
I want to apply pulse wave that is 2.71Hz and duration (high) 200us. But when I apply such a stimulus, I see the signals below;
The yellow signal represents positive channel. The green signal represents negative channel. The pink one is (Ch+) - (Ch-). Since the Ch- is not connected directly to GND, I measure (Ch+) - (Ch-).
The pink one's signal shape should be pulse. And remain side of the pulse should be zero. Please refer to attached scope picture of the signals;

View attachment scope_3.bmp

If we take a closer look to scope output above;

View attachment scope_5.bmp

As it is seen, the result signal (pink one) is not a pulse.

On the other way, I applied the pulse signal to body from a signal generator (20Vpp), I see the resulting signal below. There is no math function here, because the reference of the signal generator is GND.

View attachment scope_8.bmp


That is why I want to connect the negative channel to GND directly and control the current from high side.
 

You should be able to drive it from the low side and get a clean 200us pulse.
It's a lot easier to control the low side, then the high side.

Post the complete constant-current circuit including the pulse drive, not just parts of it.
 

Boy, I REALLY don't like the idea of having 200 volts tied to my body. Even if your low-side current source is off, there may be a path for current to flow. Through me.
 
I assume this is your loop performing poorly.

It may have marginal stability, low bandwidth and/or windup (the loop starts trying to push full current).

High-side/low side is irrelevant.

How much accuracy do you need? Medium accuracy with good regulation performance can be had with a simple transistor current source topology (google it)
 

it would be better to use a CT on the o/p of the Royer Tx - this way you can reference the current signal to anywhere you want.

By the way uncontrolled DC of 200V across the chest of people with slightly weakened hearts can easily see them off this mortal coil - you had better have 2 x protection ckts on your device to ensure this doesn't happen - and some time out stop ckts - else you will be liable when it fails and kill some one ...

you should also have some over rated current limiting resistors to define the max possible current out ...
 

Dear asdf44,

I am controlling the current with an op amp (TLC277) and a current sense resistor. This op amp switches the mosfet which is on low side current path.
How can I analyze/decide whether my loop has marginal stability or not?
 

Attachments

  • SEP_STIM.pdf
    444.8 KB · Views: 98
Last edited by a moderator:

the loop is not stable without negative feedback, damping and freq roll off so that the control is "inside" the speed of the power loop...
 

The #3 circuit, low-side MOSFET with source connected current sense could be assumed stable by working of the internal OP compensation a first sight. In practice it's usually not, or at least not for all drain loads due to MOSFET capacitances.

A safe stimulator circuit must be galvanically isolated, in so far there's no convincing reason against a low side steered current source.
 

Dear asdf44,

I am controlling the current with an op amp (TLC277) and a current sense resistor. This op amp switches the mosfet which is on low side current path.
How can I analyze/decide whether my loop has marginal stability or not?

Suggestions (assuming Q25 is the low side fet we're talking about....what is Q23?):
1) Add an integrating cap between your opamp output and inverting input for some compensation. Use calculation/simulation to set this at some reasonable frequency. Maybe 50khz is reasonable bandwidth here.
2) Deal with offset/windup issues: Add a weak pull-up from the opamp inverting input to the positive supply to ensure that when Q26 is on the opamp output saturates to ground (100k-1meg).
 

.........
How can I analyze/decide whether my loop has marginal stability or not?
The easiest way is to simulate it with an analog simulator, such as the free LTspice from Analog Devices.

Try a capacitor between the U20A output and its inverting (-) input to see if that helps (with R40=1k ohm).
Start with about 1nF.

What is Q23 used for?
 
Last edited:

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top