See the difference
View attachment 113665

View attachment 113666

I believe @joeyd999 one turns ON harder

You needed a simulator to tell you that? It is obvious upon inspection.

 

You needed a simulator to tell you that? It is obvious upon inspection.

Naaah ! I had sometime off and I wanna show it to @AnalogKid.

So this part is settled...
There is one more (Part II)

 

If I read the drawings correctly, the output saturation voltage is *lower* by 400 uV with R3 in place.

ak

 

Hello,

It is also odd that the simulator shows a negative voltage at the collector in saturation.

Bertus

 

If I read the drawings correctly, the output saturation voltage is *lower* by 400 uV with R3 in place.

ak

Isn't is 11mV lower with R3. Without R3 base drive current is more causing more turn ON.

Hello,

It is also odd that the simulator shows a negative voltage at the collector in saturation.

Bertus

The collector will be connected to PIC Digital IN. So it's no biggie.

View attachment 113686 (insane)

Please explain ur diagram.
I am noticing your posts are more confusing than helpful.

 

Ignoring all the stuff in this thread; Twisted pair shielded wire will go a long way, provided the shield is connected to ground at one end only.

 

you're absolutely right about this one - i made a multi-feature test as current signal - , voltage signal - switch , parallel AC line effect at at 1cm away , switch strobe - to show/include in simulation the relative influence to setup your "clear readout" -- your simulation switches ground through 50 feet to switch 50 feet back to BJT BASE - only the inductance of the cable smoothing the "kick to GND" -- (( "not helpful ... blah blah blaa" - ((indeed)) ))

Please sum it out for me.
Is the result good or bad ?

Ignoring all the stuff in this thread; Twisted pair shielded wire will go a long way, provided the shield is connected to ground at one end only.

I know, but installation is out of my hand. I will recommend that to the customer but I dunno know they will always comply.

 

if your going that distance, do you have to be worried about plenum cable and fire codes?

 

Isn't is 11mV lower with R3. Without R3 base drive current is more causing more turn ON.

Yes, the base drive current is greater without R3, but that does not automatically mean that the transistor is more "on". Which is more important to you, a lower saturation voltage, or lower efficiency plus more heat dissipated in the device plus lower long-term reliability? Removing R3 basically doubles the base current, but that is not always a good thing.

ak

 

Ignoring all the stuff in this thread; Twisted pair shielded wire will go a long way, provided the shield is connected to ground at one end only.

Plenum cables? Nope normal cables
Fire....? No worries.

http://www.allaboutcircuits.com/textbook/alternating-current/chpt-14/characteristic-impedance/
http://www.rfcafe.com/references/electrical/dielectric-constants-strengths.htm

+ my (insane) setup doesn't consider the wires running along metal trays as they do in some installations where the coupling between signal cables and power lines are restricted to 1m length (i guess) and would have some 1k greater effect e.g. 2+ mH instead of 2 µH

Again more confusing.
I asked you to sum it up.
Please don't throw me around.
I asked, Good or Bad ?

Yes, the base drive current is greater without R3, but that does not automatically mean that the transistor is more "on". Which is more important to you, a lower saturation voltage, or lower efficiency plus more heat dissipated in the device plus lower long-term reliability? Removing R3 basically doubles the base current, but that is not always a good thing.

ak

So ....to sum it up, Is your version better, concerning noise immunity and long term reliability ?

 

usually (not always) you "blind hit" with your first most
thus #7 #11 (coz long line × contact strobe × *inter supply (*if they're not inside same unit/GND ... feed from same line ) -- witch you can always filter off and inside the µC programmatically)
you're not telling us the whole story (not that i'm interested) when #8 will also suit your needs (the way off from previous)
((+ don't feed pulses from inductors/transmission lines stright to gate/base -- if you don't know it's what it suppose to be))

 

I need around 500msec time constant to slow down the response during on and off.
I have a counter routine setup in program but once in a blue moon I am getting a glitch during a contact bounce simulation.
I believe I can eliminate this by adding a RC between the PIC digital input and the circuit.

From the calculator I am getting the values of R28 and the 10uf cap. I included R8 in to the calculation.
Are the values correct for a 500msec time constant ?

 

Are the values correct for a 500msec time constant ?

T = R * C, where T is in seconds, R is in ohms and C is in Farads. So for the low-to-high transition on the PIC I/O line, it will be 0.47 seconds, and for the high-to-low transition it will be 1 millisecond.

 

View attachment 115530
I need around 500msec time constant to slow down the response during on and off.
I have a counter routine setup in program but once in a blue moon I am getting a glitch during a contact bounce simulation.
I believe I can eliminate this by adding a RC between the PIC digital input and the circuit.

From the calculator I am getting the values of R28 and the 10uf cap. I included R8 in to the calculation.
Are the values correct for a 500msec time constant ?

Another thought?
http://www.labbookpages.co.uk/electronics/debounce.html

 

500ms is not a "bounce".

 

Regarding "Bounce".
Does the bounce occur more during making contact than breaking the contact.
My Pressure switch is heavy duty one. Customer does not have anything small. The contacts can switch a 2KVA motor easily. The other inputs are,
2 Inputs opto coupled sensing the motor 24V or 5V DC.
4 relay contacts.
They may occur in repetition or just one once in a while.

So making a soft debounce to all inputs takes a lota space and I need memory space to setup the USART msg's. Which is not implemented yet. So I need to use minimum space for the LCD messages which so far is using 40% and I have not yet coded all of them let alone USART.

My setup is based on counters with 1 sec increment. With that I get delayed response to inputs which is fine.
I simulated the input with hand to check if debouncing gives me problem. I know a debounce is fast than doing by hand but I thought I will check.
During the check I am getting a glitch. Not a program glitch as the program is working. It's the LCD display update. It blanks out a part. This happening during a certain count position when the input is triggered the display is not updated.
I know this is due to the change input.
I know slowing down the input change will fix it.
Currently the test circuit does not have any RC filters.

I dunno how to make two debounce to one switch to detect make and break conditions. I can make a IRQ based debounce to a PB switch. Which is just one state.
And that is why I made a counter based detector which works. I believe it will work as I do not think the way I am simulating will occur in installation.
But I have to make sure.