| Author |
Message |
mike_bihan
Joined: 21 Mar 2002
Posts: 259
|
 28 Oct 2004 7:42 Pulse width with negative temperature coefficient |
|
|
|
I was designning a pulse generator ( ring oscillator);
The spec need the period to have negative temperature coefficient.
The hotter, the shorter the period.
If I used a RC delay to form the pulse width, the coefficient will be positive because the resistor will be more resistive.
MOS working close to sub-threshhold region have negative coefficient however, I have no idea to implement.
Any idea? Any paper?
thanks,
|
|
| Back to top |
|
 |
Humungus
Joined: 10 Jul 2001
Posts: 417
Helped: 25
|
| 28 Oct 2004 8:03 Re: Pulse width with negative temperature coefficient |
|
|
|
| If your design is in IC, some technologies offer a hiresistive poly which has negative temperature coefficient. You could also use a MOS resistor biased under its ZTC (Vt decreases with temperature). You could use a PTAT current source instead of a resistor.
|
|
| Back to top |
|
|
rtpq
Joined: 12 Oct 2004
Posts: 15
|
| 28 Oct 2004 8:22 Re: Pulse width with negative temperature coefficient |
|
|
|
| If you want NTC pusle width instead of NTC frequency, why generate a ramp and NTC/PTC voltage? You can use a comparator to generate the pusle.
|
|
| Back to top |
|
|
mike_bihan
Joined: 21 Mar 2002
Posts: 259
|
| 29 Oct 2004 3:59 Pulse width with negative temperature coefficient |
|
|
|
Humungus:
Yeah, it is the direction. However, in detail implementation plan will be more complex. The bias voltage will be very sensitive below ZTC point.
PTAT might be a nice idea.
Any good article on that?
thanks,
|
|
| Back to top |
|
|
