Pic interface with sensor

[Harald Kapp]

I have no idea if i should use a p or n mosfet for the switching.

It depends on your circuit. A NMOS requires a positive Vgs greater than thr threshold voltage Vgsth. As the source is the output, i.e. at 0.65 V when on, the Gate voltage needs to be > (Vgsth + 0.65V). A typical logic level MOSFET (e.g. BSS138, Vgsth_max = 1.5V) will work when Vgate >= 2.15 V. A processor operating at 3.3 V will easily turn on that MOSFET.

 

[Rajinder]

The gate is driven by the micro, drain is at 0.65V and source is at 0V for a n channel mosfet? Is that wrong?

 

[Harald Kapp]

W

The gate is driven by the micro, drain is at 0.65V and source is at 0V for a n channel mosfet? Is that wrong?

When turned on, the source is at 0.65 V, too. This is disregarding a small loss for Vds due to Ids*Rds. You'll have to use a low-Rds MOSFET for high currents..

Again: do not describe circuits by words, that is what schematic diagrams are for. A diagram is much more unambiguous than a verbal description.

 

[Rajinder]

Hi all,
I have managed to find a dc-dc converter that with feedback resistors, i can achieve 0.65V DC pulse. It is the LTC3549.
The design has changed, where i need to test 5 to 10 sensors for comparison.
I thought i could use a high p fet switch and drive it with a pnp transistor.
I am not sure this will work. See my sketch below. I need to find a way to switch tge 0.65V pulse.
Thanks in advance

 

[Rajinder]

Hi
How can i switch a 0.65V pulse?
I am testing some sensors and 0.65V pulse is required for the heater voltage.
I was thinking of using a pfet high sude switch. Drain as output, source at 0.65V pulse. Gate to drain with a pull up to keep the fet off.
The other end of the pull up connected to a npn transistor. When the npn is on, the gate is driven low. The pulse should appear at the o/p.
The npn is driven from a PIC I/0.
Thanks in advance.

 

[AnalogKid]

Your plan would work if the output were 5 V or more, but not with 0.65 V. IF I understand your description correctly, the gate-source voltage (Vgs) will be only 0.65 V when the FET is on. This is not enough to enhance the FET. You would have to pull the gate down to at least -2 V (depending on the FET's Threshold Voltage) to get consistent conduction.

Another way to go is to use a bipolar transistor as an emitter follower. When you pull the base up to approx. 1.3 V, you should get your 0.65 V at the emitter.

How accurate/precise does the output voltage need to be?
What is the output current?
What is the load the pulse is driving?
How fast do the pulse edges have to be?
What are the minimum and maximum pulse widths?
What are the min/max output pulse repetition rates?

Discussing circuit details with nothing but text is fraught with peril. Please post a schematic of your proposed circuit. A hand-drawn sketch is fine. Be sure to have a reference designator for each component.

ak

 

[Rajinder]

Hi
I have tried to post a scematic in different formats all less than 3mb but get errors.
Output needs to be 0.65V (tolerance of 1%).
The width has to be 100ms.
I have generated the pulse using a dc dc converter Ltc 3549, which gives exactly the output i require. The sensor heater is at 150mA.
So we pulse the heater first, then apply 3V and read the output V sense.
From 3V and Vsense we can obtain RS resistance of sensor.
I need to test 5 to 10 sensors and need to pulse each sensor in turn. Hence i am trying to switch the 0.65V DC pulse. Sensor 1 is pulsed, each additional sensor is pulsed 6 seconds after each other. After 10 sensors we return back to sensor 1.
The load the pulse is driving is 510R.


I have a pfet. Source at 0.65V, output will be the drain. I have a pull up 47K between gate and source to keep the fet off.
The gate of the fet is fed from collector of npn via a 10K. The npn has a 10k at the base connected to the PIC I/O. A logic 1 turns on the npn, pulling the gate low. I thought the pulse will now be at the drain and load?
If i pulse the emitter follower with 1.3V pulse, then i will get a 0.65V at the emitter.
So can i get the dc dc converter o/p into an emitter follower?
Thanks in advance.

I will try the upload again.

 

[Rajinder]

This is the info for the gas sensor.
Any help would be aporeciated.
Thanks.

 

[Harald Kapp]

@Rajinder : Why do you create yet another thread on the same topic?
Stay in one thread so our members can see what you are about and where youre issues are.

Again I merged it with your existing one.