generating a weird signal ...

[TucoRamirez]

Hi,  may i ask you something?

How can you generate a 3 states ( +Vo, gnd -V) signal (kinda RZ signal)  (of course the 200Hz pwm'ed signal is not a problem wth a tiny 3V pic10 ...)  with the following restrictions... only a 3V batt allowed and output must feed 0.5mA max ...

I was thinking about current sources and a 3state switch  array (to draw current from +V to the output load, then  the 'hiz' state with the load connected by itself to ground and then the -V state with the current flowing towards -V ... (i mean 3    0  -3V)

But i have some doubts about that solution and in the choice of the correct 3V regulator

(i'll upload the schematic of my idiotic idea as soon as i can )


best regards

Tuco!

[Ichan]

Not a very clear explanation.

You can generate negative supply (-3V) from positive supply (+3V) using ICL7660 or similar ic, then use an op-amp (supplied by +3V/-3V) configured as a comparator with +1.5V reference voltage - so the output is about -3V for input below +1.5V and about +3V for input above +1.5V.

-ichan

[Gallymimu]

Maybe a charge pump is what you need (like Ichan's ICL7660).  It is a very simple way to create boosted or negative voltages.  You can probably control the charge pump with your small PIC

[solutions]

An H-bridge driver will probably work. 3V and GND is all you need if I am correctly GUESSING what you are trying to do.


---------------------------------
A                  B                        I
I                   I                     ----- 3V
---LOAD-------                       --
I                   I                        I
C                  D                       I
---------------------------------

AD +3V
BC -3V
CD  GND/LOADSHORT

[TucoRamirez]

yop !! yes you were so right i forgot that ^^...

:p now i'll write the pic10 testbench code ^^

btw, what could have if i connect this electrical outs  to a frog's leg ^^ !!! :p  (Evil Tuco moment ) (anyway , how much ohms does it have a load like that ??)

[solutions]

yop !! yes you were so right i forgot that ^^...

:p now i'll write the pic10 testbench code ^^

btw, what could have if i connect this electrical outs  to a frog's leg ^^ !!! :p

I don't know. A Frenchman with electrical burns?
 

You can have an extremely low source impedance with it...all depends on your switches. Just watch out for "shoot-through" (AC or BD on, even momentarily)

p.s. don't abuse live animals

[TucoRamirez]

so the "shoot through" is what i see as the spikes on the output?? (sorry but i 'm a krappy english and french speak/writing ape )

hehe yes i want to cause pain to french people ^^^i m jokin'

anyway the load is not inductive so i think i can modify it yo insert a "flipping"  current source to limit the I_max

[solutions]

I don't know why you want "current flipping" when current is only in one direction if you pick the right place in the circuit

One depletion mode device, with appropriate Idss, might do what you want without "flipping" anything. Or you can put a current source in its place, etc


---------------------------------
A                  B                        I
I                   I                     ----- 3V
---LOAD-------                       --
I                   I                        I
C                  D                       I
---------------                        I
         I                                   I
         I                                   I
      - -                                   I
     I                                       I
     I                                       I
  --I     D-mode FET                  I
  I  I                                       I
  I  I                                       I
  I   --                                    I
  I     I                                    I
  I     I                                    I
  -------------------------------
  

[TucoRamirez]

In fact i want to have the +3V 0  -3V signal driven by a pwm and tricky logic as in your h bridge, but the current drawn to the charge is kinda weird to my understand , in fact is a ... ok lets face it is not muscle but living tissue ...  and as long as the load is a freaking dynamic moving changing load, i want to guarantee enough current from the givem impulsions to cause enough pain to the rat (i wanted to say the undergraduated students ^^)

and no its not an orgasm usb powered smart gadget  ^^

[dotm]

well you did not say how long you you want that 0.5mA max to be held (in seconds)
what percentage can the desired +/- 3V change when loaded with the maximum current?
can the reference ground be virtual?

[TucoRamirez]

oki,  the current peak must be held for tenths to hundred of milliseconds...  the current can range from  100uA to 500uA  and no matter about ground reference to the output.

[dotm]

I must admit that H-bridge solution seems to be brilliantly elegant, very smart. As i understand it, the current given (+/-) and feed trough (gnd), is only limited by the Fet used, resulting to a small output impedance. I'd propably use some sort of a software-pwm to ensure that the rising edge of the susequent switch state wont fall together with the falling edge of the precedent state.

[TucoRamirez]

what about this modification?

[Gallymimu]

Are you really stimulating live tissue?  We used to do this a lot in our lab.  We did transcutaneous (through the skin) which required much higher voltage).  We also did intramuscular (stimulating using needles in the muscle tissue).

Normally for stimulating tissue a current source is used.  What exactly are you trying to do?

[TucoRamirez]

send negative and possitive current/voltage peaks to brain tissue .

[Gallymimu]

send negative and possitive current/voltage peaks to brain tissue .

Yeah you should ONLY do that with a current source.  The results will be inconsistent otherwise.  Even after a short amount of stimulation the impedance of the tissue will change as the potassium and calcium channels (and some other channels) open/close with longer term adaptation to the stimulus.  Longer term stimulation will be inconsistent too as the electrodes are attacked and partially encapsulated by the immune system.

Is this for real research or just for fun?  What region of the brain are you stimulating?  Are you stimulating deep brain tissue or cortical layers?

I find this stuff very interesting

[TucoRamirez]

I'm trying to win a place in a new res. Group where i'm studying.  Back to the game...  the h bridge seems to me to be ok but mos switched paths make me lose current...

[Gallymimu]

I'm trying to win a place in a new res. Group where i'm studying.  Back to the game...  the h bridge seems to me to be ok but mos switched paths make me lose current...

are you putting in dead time with your PWM signal?

[TucoRamirez]

i'll start to do that next week, in the meantime i still have doubts about how to ensure enough ( but low anyway )  current while switching on the load ...

[solutions]

I think PWM will be quite useless, unless you are into eating cooked brains (some cuisines do).

I think you need to FM current-impulses.

Those same circuits will work for this.

Like all great scientists have done, refine the method on yourself before exposing another living (or dead) being to your creations

[TucoRamirez]

fm current impulses ?  ?  what are you sayin' ?   you mean sending the pwm-like signal but in FM mode for a given pulse duration??? or attack directly the current source???

[solutions]

Impulses. Change the off-state spacing only. PWM, but fixed on-state time.

[Gallymimu]

speaking of your signals, what frequency, pulse width, burst size, and waveform shape are you planning on using?  The effectiveness of neural stimulation is pretty dependent on these parameters.  There are TONS of publications on this stuff and how to do it effectively.  Most of the good papers require journal access but here is one that has some useful info.
http://students.washington.edu/zanos/literature/Motor%20cortex/Cortical%20electrical%20stimulation/Cogan%202008%20-%20Stimulation%20and%20recording%20electrodes%20%5BR%5D.pdf

We did some good work stimulating the pleasure centers of rodent brains to the extent that you could get a starving rat to prefer pressing the lever to deliver pleasure stimulation over eating food.  Interestingly (and luckily) the same region in the human brain doesn't elicit the same extreme response.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3008353/

[TucoRamirez]

pleasure centers  ... oh yeah !!! 

as i know, i need to supply less than +/-300uA max for a  moving " enemy"   ^^ load of 1k max... frequency to be tested but somewhere between 100Hz and 800Hz..

so burst size  (or pulse width) for each peak to be calibrated too  between 5% and 30% of f ^^

I think i'll see a frog or a mice attached to the jtag port ^^ i already havea an usb memory ^^  hehehe  ... ( i think i'll not test-on-myself like the ecg !!! )

Posted on: February 06, 2013, 07:57:36 07:57 - Automerged
btw first link is veeeeeeeeeeery interesting

[solutions]

Mouse



Mouse



Only one needs a wire up its ass to be happy. I'll let you figure out which one.

[Gallymimu]

Ah yes, but the first one is much more modern, wireless and all....