A Simple Circuit concept to study biasing transistors

Thread Starter

fredric58

Joined Nov 28, 2014
203
ok! i need to back up just for a moment. i have maybe 30 hours invested in researching using transistors as switches since my last post. i have garnered quite a bit of positive information. HOWEVER, in watching youtube videos and carefully listening and practicing the authors implementation. i found they all don't seem to agree with one another. "or" they get to a point and say something completely opposite of what they had just been trying to demonstrate. i won't bore you with details.

i have looked through arduino, sparkfun and other source for information on "data sheets". i understand how helpful they can be when available. but it is the way it is explained that is...............uhm............they pick a product and tell you about stuff that doesn't pertain to what you are working on, though there is a lot of information. so if anyone has a few minutes i would like to get a more concise understanding of how to interpret a "data sheet" and i hope to be able to be "specific". knowledge of these parameters could be priceless in my estimation.

i would like to start with INPUT. specifically these items:

i have a date sheet. it tells me this:

input LOW voltage- sym Vil (0.8v) max
input HIGH voltage-sim Vih (2.0) min
output LOW voltage
output HIGH voltage
Vcc current (operating)
Vcc Current (standby)
input LEAKAGE current
input current HIGH
input current HIGH (not a typo)
output load Impedance
preamp input resistance
MIC sp +/- gAIN.

That's a lot and I don't expect anyone to have time for all of it. so would anyone like to take the first (2)?

OR.....point me to a page that can explain it since you have already been there. thanks
 

hobbyist

Joined Aug 10, 2008
887
The data sheet about inputs and outputs, right now is really not needed, just any info. on voltage and current consumption for each device, is needed if that can't be attained, then you still have enough information on what current the soundboard needs so you can build and test your circuit, to see if the transistors are able to switch each device on with the inputs from the other device.

Looking at what you shared in your post about current measurements,
29-32mA for the soundboard, gives you a good springboard to work off of.

T2 needs to be biased to sink around 29-32mA of current.
To saturate T2 you would need to use a base resistor that would supply 1/10 of that, around 2.9-3.2mA. of base current.

You determined the output voltage of the driving source (PIR), so now you can take the output voltage of the PIR minus Vbe of the T2, then divide into that the base current you calculated above to give you a ballpark figure for resistor R2 that will be your base resistor for T2.
Build and tweak this resistor value to see if you can get it to turn on the load (soundboard).

Now you could just hook up the PIR at this point and see if it all works together, at this part of the design.
 

hobbyist

Joined Aug 10, 2008
887
Now if your talking about the PIR outputs then we will have to work with that, to find what voltage is needed to drive T2 base current. But your original schematic, showed 3.3V out of the PIR. So I went with that to show this example.

transistor switch.jpg
 
Last edited:

Thread Starter

fredric58

Joined Nov 28, 2014
203
this is very helpful and my numbers were only off by a few ohms. i found a really good vid on transistors and did some more searching on understanding data sheets this morning and i believe it won't be long before i can do the whole chain of components. correct me if i am wrong, but, i believe, that as more components are added (you mentioned having to tweak) that it may be similar to my hydroponics irrigation system. when you have (1) valve it's relatively easy to set the water flow, (2) valves not too bad. but by the time you have (20) valves on (1) supply pump you really have to start at the very end, and gradually add resistance (close) the ones that are at the beginning or else nothing makes it to the end. till each area has an equal flow of water. a rather simple assimilation but i believe electronics may be quite similar (?) COMMENTS?

i have another redundant question in regards to "the load" you mentioned a HIGH SIDE and a LOW SIDE. (this was before i had a data sheet and had tested the low side with a stack of resistors.)

OK, that data sheet says the "operational LOAD is 30mA". I would think (not sure) but I would think that you might want some cush, for lack of a better word. by that I mean, maybe I would want a range of say 25mA to 35 mA's? I would refer to it as a buffer or maybe slop, maybe wiggle room. I've noticed that it isn't set in stone. meaning resistors can be with in 1% to 5% tolerance. beta, from the same manufacture, same batch can have a substantial difference between identical components. so i would "assume" which i know is bad. that there is a HAPPY medium? IF.... you were going to MASS produce something, you really couldn't affordably check each individual component. RIGHT? so you have to have this window. it goes from "here" to "here". and as long as you're in that window you can be pretty confident it is going to work.

please explain the purpose of putting a 1 - 10 ohm resistor between the Vcc and component. and also putting a 1 - 10 ohm resistor between the COMPONENT/LOAD and the GND. I think this correlates with the above question. but i am not sure.

thank you again for participating in this thread!
 

hobbyist

Joined Aug 10, 2008
887
Ok you have 3 individual devices, you want to drive the last two with transistor switches.

In order to use transistors as switches they need to be turned on via base current.
The base current is then multiplied by ten to give collector current that will drive the load.

These devices are unknown as far as circuit impedances. A data sheet for each device would be first used to see if working voltage and current is given for each device. If it is not then you need to do preliminary testing to determine the current each device needs to work optimal.

To do this test it is best to use a volt meter not an ammeter, because an ammeter could blow a fuse.

step 1) Begin placing one resistor at a time, do not stack resistors in this procedure in the place as shown as "test R3"

read the voltage across R3 then calculate current through it. (this case around 31mA), ok write down the current and the resistance R3.

1.jpg

now replace that resistor with higher resistors for R3.

2.jpg

keep doing this until the device stops working properly, or you have determined its properties are going bad.

Here I have determined that performace is just starting to drop.
Ok this would be the limit of current loss for this device. This case its around 23mA.
It has dropped from 30mA to 23mA, so this is the current range I will design the collector current for.

3.jpg

Now in order to know what current to choose for the collector current I need to see what base current I have available. The base current to turn on the transistor will be determined by the source thats driving it.

So now I move onto the source output. (device 2)
A voltage reading at its output shows around 3.3V.

4.jpg

Alright So now I put a resistor of some large value at its output to ground to see if it pulls the output down.

5.jpg

next I keep lowering that resistance for test R2

6.jpg

Ok at this resistor the source (PIR) just now starts to become unstable.

I measure the voltage across the resistor R2 and calculate the current to be around 2.3mA
Now I have determined that my base current cannot go any higher than 2.3mA or so.

Now I have some current values to work with to work into the transistor calcualtions.

Lets take the max. current allowed by the source (PIR) of 2.3mA
And lets determine saturation current for the transistor, so we take (2.3mA X 10) = 23mA

OK now looking at the range of collector currents we found in the last steps we see that 23mA is in its collector current range.

Now we can start working out a value for the base resistor for T3 to work at into saturation.

Very important take the voltage at the output of the source (PIR) which is 3.3V then subtract Vbe of 0.7V to give around 2.6V.
This is the voltage drop across your R2 bias resistor. So now take the 2.6V and divide it by the base current calculated above of 2.3mA, (the allowable (PIR) output current) to give you the approximate value for the bias resistor R2.

equation {(Vout source - Vbe) / Iout source} = R2,,,,,,,

{(3.3V - 0.7V) / 2.3mA] ~= 1.1K ohms.

7.jpg

Now very important again, check the output voltage of the source (PIR), and you see that your in the voltage range of proper working for the source, the source allowed down to 2.3V drop before it would act up. Your around 2.6V your driving your source down towards its maximum allowable, so from there you could tweak some values, you can begin to raise your R2 value up to get your source output higher in voltage while maintaining the transistor to be lower than 300mV drop acroass it.

So lets tweak R2, lets go to 10K ohms.
Ok we brought the input voltage back up, BUT the transistor is no longer saturated, a reading across there shows around 800mV.

8.jpg


so lets drop R2 value until we find the limit to where the transistor just comes out of saturation.
So as experimenting takes place we see that around 8.2K ohm resistor seems to bring the transistor border line into saturation, and the source voltage has dropped very little.

9.jpg

So now a final tweaking for this stage, shows that we learn how to compromise current inputs for current outputs, how much voltage drop are we willing to allow the source to drop to on its output, yet perform as we expect it too, while at the same time allow the maximum current possible to drive the output load with the collector current.

In this case we can go anywhere between 1.1K ohms to 8.2K ohms, so lets go somewhere above the middle so as to keep voltage drop at the input low, try 6.2K ohms.

10.jpg

Now we have very little voltage drop at the PIR output, and we have our transistor saturated, and were getting good collector current that the load was looking for under optimal conditions.

Then the whole process is repeated for the next stages, and so on.

Hope this helps make it more clear.
 
Last edited:

Thread Starter

fredric58

Joined Nov 28, 2014
203
WOW, what an incredible amount of tutorial material. I am sure I will have at least a few more questions but that won't be till after I work through and understand each of the steps provided several times on paper and then apply them to an actual working circuit this weekend. Hopefully I will come to a point in the process that I will be able to answer those questions on my own. Many Thanks for your mentorship. I will keep you posted on my results. Thanks again!
 

MrAl

Joined Jun 17, 2014
7,659
Hi,

I dont know if anyone mentioned this yet, but there are three basic configurations for a transistor:
1. Common emitter.
2. Common collector.
3. Common base.

Each one has something specific that makes it better to use for one purpose or another.
It is worth looking into each of these configurations.

Also another point not talked about too much is that the base emitter reverse bias can not go to high or else the transistor blows out or becomes seriously damaged. Many transistors only have a max reverse Vbe voltage rating of 5 volts even though their Vce rating might be 100 volts.
 

Thread Starter

fredric58

Joined Nov 28, 2014
203
thank you, points taken and i'll look into them after I have a better understanding of the instruction so far received.
 

Thread Starter

fredric58

Joined Nov 28, 2014
203
being a forward thinker I have a question which is not relevant to my current quest. and it boggles me. I have a game camera. this question is about power consumption. my game camera has (6) 1.5V batteries in it. this thing will run for a year +. I know it has to be powered to do what it is supposed to do. but! BIG EXCLAIMATION POINT! it "seems" to sleep. until something comes along to trigger it. then it seems to go back to sleep when nothings going on. (really the only way I can explain it.) after it takes pictures. it amazes me that 6 AA batteries can last for such a long period of time.

to me it is like magic. I build stuff and it may last a week, 10 days at best. NOTE: all my stuff is battery operated. it amazes me! I can leave my camera out for a year. watch a raccoon hit the feeder, then have a mate, then see all it's "kids" come back to the feeder months later. I can see the whole life cycle. I'm talking about a YEAR! with (6) 1.5V batteries. how is this even possible?

how do you get to be THAT smart? most wouldn't recognize little things like this. but I do! it's fascinates me.
 

MrAl

Joined Jun 17, 2014
7,659
being a forward thinker I have a question which is not relevant to my current quest. and it boggles me. I have a game camera. this question is about power consumption. my game camera has (6) 1.5V batteries in it. this thing will run for a year +. I know it has to be powered to do what it is supposed to do. but! BIG EXCLAIMATION POINT! it "seems" to sleep. until something comes along to trigger it. then it seems to go back to sleep when nothings going on. (really the only way I can explain it.) after it takes pictures. it amazes me that 6 AA batteries can last for such a long period of time.

to me it is like magic. I build stuff and it may last a week, 10 days at best. NOTE: all my stuff is battery operated. it amazes me! I can leave my camera out for a year. watch a raccoon hit the feeder, then have a mate, then see all it's "kids" come back to the feeder months later. I can see the whole life cycle. I'm talking about a YEAR! with (6) 1.5V batteries. how is this even possible?

how do you get to be THAT smart? most wouldn't recognize little things like this. but I do! it's fascinates me.
Hi,

Not sure what kind of camera you have, but microcontrollers can go to sleep and when in sleep mode they use nano power, which means they take very little power from the batteries.

In sleep mode, they often wake up for a tiny fraction of a second so they can monitor input conditions and trigger the appropriate response if needed. For example, i built a refridgerator monitor some years back using a PIC micro controller and it wakes up very often but only then runs for a fraction of a second to look at the temperature inside, then goes back to sleep if nothing is wrong. Every 10 minutes or so it displays the temperature then goes back to sleep mode. Two alkaline AA batteries last for two years in it.

Is your camera motion activated? If so that could be it. It only wakes up when something moves. Well, it wakes up for a short time to check motion conditions then goes back to sleep if noth8ing is detected. It takes such a short time to check the inputs that it takes very little power from the batteries. So the duty cycle is very very low and thus you get a long battery life. Not sure what kind of cam you have but that is one possibility.

If you want to build projects like that you need a microcontroller chip like the PIC or perhaps the Arduino.
 

Thread Starter

fredric58

Joined Nov 28, 2014
203
Hi,

Not sure what kind of camera you have, but microcontrollers can go to sleep and when in sleep mode they use nano power, which means they take very little power from the batteries.

In sleep mode, they often wake up for a tiny fraction of a second so they can monitor input conditions and trigger the appropriate response if needed. For example, i built a refridgerator monitor some years back using a PIC micro controller and it wakes up very often but only then runs for a fraction of a second to look at the temperature inside, then goes back to sleep if nothing is wrong. Every 10 minutes or so it displays the temperature then goes back to sleep mode. Two alkaline AA batteries last for two years in it.

Is your camera motion activated? If so that could be it. It only wakes up when something moves. Well, it wakes up for a short time to check motion conditions then goes back to sleep if noth8ing is detected. It takes such a short time to check the inputs that it takes very little power from the batteries. So the duty cycle is very very low and thus you get a long battery life. Not sure what kind of cam you have but that is one possibility.

If you want to build projects like that you need a microcontroller chip like the PIC or perhaps the Arduino.
Yes it is a pretty standard game camera with a motion sensor on it. I have Arduino boards already. I have used them in projects but they drain the battery too quickly. They're great for telling the project what to do and a lot of fun to work with. I have also read that you can put them in sleep mode. But I get what you're saying, that it only takes a millisecond to check for a signal or maybe a microsecond. And what you're talking about is very similar to what I'm trying to achieve with transistors as switches. ( turning on stuff when it's needed so it doesn't have to run all the time) Not that It would be anywhere near is a efficient but perhaps more efficient than my previous designs.
 

MrAl

Joined Jun 17, 2014
7,659
Yes it is a pretty standard game camera with a motion sensor on it. I have Arduino boards already. I have used them in projects but they drain the battery too quickly. They're great for telling the project what to do and a lot of fun to work with. I have also read that you can put them in sleep mode. But I get what you're saying, that it only takes a millisecond to check for a signal or maybe a microsecond. And what you're talking about is very similar to what I'm trying to achieve with transistors as switches. ( turning on stuff when it's needed so it doesn't have to run all the time) Not that It would be anywhere near is a efficient but perhaps more efficient than my previous designs.
Hi,

Yes i agree that turning stuff off definitely helps, that's how microcontrollers do it too they turn periph's off when not needed as well as power things down like the ADC when not needed.

It's a little tricky getting Arduino to get to go to sleep, but you can read about that on the Arduino site. One fo the problems is the voltage regulator. The 5v and 3.3v regulators will draw power even when not being used. I had this problem too when i made a voltage monitor for my car some years back. The chip i used was a PIC chip and it did not draw much power, but the darn 5v regulator would draw 5ma even when the uC chip was in sleep mode. To get rid of that i switched to a different low quiescent current regulator.

Also, i thought that we could make a circuit that turns EVERYTHING off for some time, and then cycles itself back on every so often. It does not have to precision timing, just maybe once per second or half second. A low power transistor circuit or maybe even a CMOS 555 perhaps could do the cycling, and that would turn everything off completely except for itself but it would consume very little power itself.
It could drive a single transistor that turns on the uC chip and related, and the uC chip could respond by holding it on while it did it stuff, then signal the cyclic circuit to turn off again, and the cyclic circuit woudl then time out and turn everything back on again after the required time. That should work and not even require using the sleep mode so the Arduino would be easy to do this way too. The cyclic circuit would have to be designed to use very very little current though when the rest of the circuit is turned off, so we dont use the battery up too fast.
 

Thread Starter

fredric58

Joined Nov 28, 2014
203
sounds very interesting. my current project version (about 10) works like this. I use an LDR circuit with hysteresis that determines whether it is day or night. the arduino nano samples the LDR circuit every half hour. it isn't precise but close enough. once it gets a reading that it is night it begins to perform several different functions. some programmed at calculated times and others when the PIR senses movement. there isn't any city power so it is set up in a manner that it can determine whether a crowd of people are walking by so it doesn't use up the battery or video recording time but it knows when there is an individual and then it takes a little 15 second video clip. my 1st project ran for 12 hours, my second for a couple days, my latest will go 8-10 days before it dies.

things I have learned and ideas to correct them are. (1) the arduino sucks up power. the LDR has to run all day and night=more power. so in this new attempt I am planning on hacking a 12v digital timer but running it on 5v because I won't need the 12v required for the mechanical relay. I only need a signal. then I can just set the time for it to go off and be done. (2) I'm going to lose some features not using the nano doing it this way because I will only have (1) trigger, the PIR. it's signal time is only from a few seconds, to a few minutes. however, when the PIR signal goes high it will intern, turn on the camera. I use a sound board for test so I can hear the noise and tell it is operating. (tired of blinking LED's) I have also though of dumping the arduino, but that would fill the sim card with video too quickly.

I have incorporated a 3" x 3" solar panel with hopes it will charge the battery during the day. however, I use 3.7v lithium ion batteries (3) in parallel, which has a charging and circuit protection AND it puts out 5v. so there is another loss. most of the components I use are off the shelf and seem to run fine on 3.5 to 5v. my first project was 12v with a 5v regulator. that wasted power too.

this actually started (the project) as a bet in front of a fire in the middle of the desert one night 2014. I have been learning electronics now for almost (3) years when time is available and apparently it has become an obsession. a good one though. every time I save a milla amp its a good thing.

so here is a list of things I have accomplished but it comes back too battery power

1. device turns on at a given time, (late dusk) turns off at a given time. (sunrise) // lm339 and LDR // going to try a digital timer
2. during it's ON time, every half hour, function (1) actuates. to be precise, for 30 seconds // 555 // that was scrapped for the nano, to operate both functions.
3. when ON, and when the PIR senses an intrusion, function (2) activates. in this example consider both functions identical.
function being defined as: to "RECORD" an image, OR to "PLAY" a sound. either one or the other. the biggest challenge I have faced is "the FUNCTION" has to be trigger. meaning items 2 and 3 the have to be triggered to TURN ON, but once they are triggered on, they have to be triggered again to record or play. at this point I use the PIR signal to do both simultaneously.

if you look at the original schematic. the PNP at the end was what I was thinking could ground/trigger the device. what I have found to be an incorrect assumption. but still think it could be done with an NPN now that I have a much better understanding of using transistors as switches.

but the biggest of all challenges is POWER. that's why I "think the digital timer would be great. the clock runs on a separate coin battery for one. up to 3 years. I will only need a small amount of current for switching the transistor and actuating the PIR since the mechanical relay will have been removed. the PIR operates in "standby" on 1-2 micro amps, has enough output to switch a transistor to turn on the video or audio boards and they have end of run markers, so......30mAs for say 15 - 30 seconds, then they go off. and we're back to the PIR and micro Amps. of course I do compromise and lose my record once every 30 minutes. sorry to just ramble on but it helps me think things through.

time to do some math!
 

Thread Starter

fredric58

Joined Nov 28, 2014
203
letting the circuit make the decision. I have 2 different signals that "could" control the same component. In my head after a day of reading about them, I went to "op-amps". there are many versions, or should I say, applications. in layman's terms, basically, it states, if this is TRUE and this is TRUE, DO THIS. if this is TRUE and this is FALSE, don't do anything, or......... do THAT. you know what I mean. A truth table. remember I am simply being "conceptual" at this time. looking for guidance in which avenue I should apply my time with study. it's limited now that I am back in the work force. I may seem to bounce from here to there on topics. but when I get an idea I have to investigate. it goes in a notebook, gets evaluated, tested and proved to be right or wrong. I have annoyed the most brilliant minds with "WHY" since I was ten. that was 50 years ago, and I have no intentions of quitting now. your patience and guidance is greatly appreciated.
 

hobbyist

Joined Aug 10, 2008
887
but still think it could be done with an NPN now that I have a much better understanding of using transistors as switches.
My last example tutorial, was pretty much designed to be used that way, I was thinking you wanted to sink the current from the soundoard, that's why I showed that example, but now as I look back at your original schematic I see that the NPN was feeding an input signal to the soundboard, and the PNP is used to sink the current.

If you use the method I showed in that last tuorial, you should be able to determine the amount of current needed to sink for the soundboard through a NPN transistor.
 

MrAl

Joined Jun 17, 2014
7,659
sounds very interesting. my current project version (about 10) works like this. I use an LDR circuit with hysteresis that determines whether it is day or night. the arduino nano samples the LDR circuit every half hour. it isn't precise but close enough. once it gets a reading that it is night it begins to perform several different functions. some programmed at calculated times and others when the PIR senses movement. there isn't any city power so it is set up in a manner that it can determine whether a crowd of people are walking by so it doesn't use up the battery or video recording time but it knows when there is an individual and then it takes a little 15 second video clip. my 1st project ran for 12 hours, my second for a couple days, my latest will go 8-10 days before it dies.

things I have learned and ideas to correct them are. (1) the arduino sucks up power. the LDR has to run all day and night=more power. so in this new attempt I am planning on hacking a 12v digital timer but running it on 5v because I won't need the 12v required for the mechanical relay. I only need a signal. then I can just set the time for it to go off and be done. (2) I'm going to lose some features not using the nano doing it this way because I will only have (1) trigger, the PIR. it's signal time is only from a few seconds, to a few minutes. however, when the PIR signal goes high it will intern, turn on the camera. I use a sound board for test so I can hear the noise and tell it is operating. (tired of blinking LED's) I have also though of dumping the arduino, but that would fill the sim card with video too quickly.

I have incorporated a 3" x 3" solar panel with hopes it will charge the battery during the day. however, I use 3.7v lithium ion batteries (3) in parallel, which has a charging and circuit protection AND it puts out 5v. so there is another loss. most of the components I use are off the shelf and seem to run fine on 3.5 to 5v. my first project was 12v with a 5v regulator. that wasted power too.

this actually started (the project) as a bet in front of a fire in the middle of the desert one night 2014. I have been learning electronics now for almost (3) years when time is available and apparently it has become an obsession. a good one though. every time I save a milla amp its a good thing.

so here is a list of things I have accomplished but it comes back too battery power

1. device turns on at a given time, (late dusk) turns off at a given time. (sunrise) // lm339 and LDR // going to try a digital timer
2. during it's ON time, every half hour, function (1) actuates. to be precise, for 30 seconds // 555 // that was scrapped for the nano, to operate both functions.
3. when ON, and when the PIR senses an intrusion, function (2) activates. in this example consider both functions identical.
function being defined as: to "RECORD" an image, OR to "PLAY" a sound. either one or the other. the biggest challenge I have faced is "the FUNCTION" has to be trigger. meaning items 2 and 3 the have to be triggered to TURN ON, but once they are triggered on, they have to be triggered again to record or play. at this point I use the PIR signal to do both simultaneously.

if you look at the original schematic. the PNP at the end was what I was thinking could ground/trigger the device. what I have found to be an incorrect assumption. but still think it could be done with an NPN now that I have a much better understanding of using transistors as switches.

but the biggest of all challenges is POWER. that's why I "think the digital timer would be great. the clock runs on a separate coin battery for one. up to 3 years. I will only need a small amount of current for switching the transistor and actuating the PIR since the mechanical relay will have been removed. the PIR operates in "standby" on 1-2 micro amps, has enough output to switch a transistor to turn on the video or audio boards and they have end of run markers, so......30mAs for say 15 - 30 seconds, then they go off. and we're back to the PIR and micro Amps. of course I do compromise and lose my record once every 30 minutes. sorry to just ramble on but it helps me think things through.

time to do some math!
Hi again,

Not sure if you are aware of this or not but, many sensors will detect whatever it is they have to detect under varying operating conditions, and sometimes they can detect very quickly which means they may not have to be on for very long and/or operate at a lower current.
For example, back when i was designing my fridge monitor i started with a 10k thermistor and 10k series resistor, and that was because the input resistance for the ADC on my microcontroller demanded 10k max resistance for 1/2 bit accuracy, so i wanted to make sure i had less than 10k. After rethinking this though, i realized that the 1/2 bit accuracy was not really needed so i moved to a 50k thermistor and 50k series resistor. That saved current right there.
But there is also the time factor. If the thermistor is connected for all time (24/7) it draws about 30ua current from the battery, but, if it is switched on for only 1 second every 30 seconds the average current draw from the battery is only that divided by 30, which comes out to an average of only 1ua. Now if the circuit is fast enough, i might even be able to go with 0.1 seconds every 3 seconds and get the same average current draw.

So the rule then is, if the sensor can sense fast enough then the duty cycle can be reduced so that it draws much less average current from the battery and thus a longer run time per battery change.

The second rule is that the sensor might draw less current on less voltage so the voltage might be able to be reduce just to the sensor. In my case, using a 50k resistor and 50k thermistor, i wanted pseudo linear operation, but what else i could have done was use a 150k resistor and 50k thermistor, and that would reduce the current draw down to 1/2 of what it was before. That would mean instead of 30ua it would only draw 15ua (without the power down option above).
The catch here is that the ADC still has to be able to read the thermistor voltage with the required resolution, and the resulting non linearity has to be made up for in code. The extra code in turn means longer run time which will have an impact if the 'time' option is also being used, but it will be insignificant using a short formula for calculating the more linear results from the measurements.

So the two main rules are:
1. Reduce the time the sensor is on if possible.
2. Reduce the operating current if possible.
 

Thread Starter

fredric58

Joined Nov 28, 2014
203
1. sensor on time. that sounds like a good idea. but the sensor needs time to stabilize once it goes on. also, during the stabilization period it squeaks out random out put signals.

2. currently it runs on 5v reg voltage. however it can run on 3.3 to 3.7. I use 3.7v lith batteries and it steps the out put to 5v. I use one of those portable back up phone charges only there are 3 and sometimes 4 batteries in parallel.

that's about to change and I will just power with the 3.7v parallel battery packs and figure out an undercharge circuit protection/ charging system later.

today I was met with a challenge. the 12v digital timer circuit with the mechanical relay "appears" (can't find a schematic) to always have 12v power going to it. and it seems that it uses the ground (and a transistor) to switch the electromagnet on/complete the circuit. I based that on the diode that protects from the collapse of the magnetic field.

it has 17 different ON/OFF capabilities and lets you choose what ever day or days you want it to operate, darn!

I am progressing nicely on my transistor studies so it hasn't been a wash out of a day.

I have learned something.
 
Top