Hello fellow engineers!
I have a question for you guys.
I'm building a little project for a friend, it involves operating a latching solenoid from a low voltage signal. I'm sensing a voltage signal which ranges from about 0.5v to 1.3v, when the voltage gets up to about 0.7v i need to send a short pulse of about 1.5v 0.44A to the solenoid. When the signal voltage reaches about 0.8v i need to send another pulse (same specs) to the solenoid.

What i'm doing now. I am amplifying the signal from .7v/.8v to about 3.5v/4v using ina122 (inst. amp) [this step is still in theory since i haven't bought the amp yet hopefully it will work]. I then use quad comparator (2 of them) lm324 to compare the voltage which will send one signal when voltage is 3.5v and another when voltage is 4v.

This is where my question comes in, based on those two signals i need a rising edge trigger circuit for 2 signals which will send two separate pulses to the H-bridge which connects to the solenoid. I tried using the attached circuit but it doesn't work (yes i know it uses falling edge, just testing it) , when Vin changes from high to low, trigger voltage X changes only about 200mV, not enough to trigger 555. I also just read that the 555 timer cannot give me enough current (about .44A) to get the solenoid going. How do i get a pulse on the rising edge of my signal juicy enough to power my solenoid?

Any suggestions or comments are very appreciated!!

 

Hello, heryrg. Welcome to AAC.

I have a question for you guys.
I'm building a little project for a friend, it involves operating a latching solenoid from a low voltage signal. I'm sensing a voltage signal which ranges from about 0.5v to 1.3v, when the voltage gets up to about 0.7v i need to send a short pulse of about 1.5v 0.44A to the solenoid. When the signal voltage reaches about 0.8v i need to send another pulse (same specs) to the solenoid.

You just need a single momentary pulse once it hits 0.7, and then again at 0.8v? Those just happen once, until some event re-sets them?

What is the desired duration of the pulse? 100mS, 300mS?

You specified 1.5v 0.44A; about 3.41 Ohms. Is that the specification of the solenoid, or is that the limitations of the supply, or do you need to limit current?

Will the 1.5v come from a 1.5v source that you already have available?

What are other supply voltages that you have available?

What i'm doing now. I am amplifying the signal from .7v/.8v to about 3.5v/4v using ina122 (inst. amp) [this step is still in theory since i haven't bought the amp yet hopefully it will work]. I then use quad comparator (2 of them) lm324 to compare the voltage which will send one signal when voltage is 3.5v and another when voltage is 4v.

OK, 1st thing: I'm wondering why you feel that you need to use an instrumentation amplifier, when a comparator can sense right down to ground quite precisely - unless there is something else I'm not aware of.

2nd thing; an LM324 is a very old and slow quad opamp, not a comparator. It does sense down to ground and can output to the negative rail, but in this case a comparator would be a better choice.

This is where my question comes in, based on those two signals i need a rising edge trigger circuit for 2 signals which will send two separate pulses to the H-bridge which connects to the solenoid.

OK, now you're throwing an H-bridge in the mix - can you be a bit more specific about the H-bridge?

I tried using the attached circuit but it doesn't work (yes i know it uses falling edge, just testing it) , when Vin changes from high to low, trigger voltage X changes only about 200mV, not enough to trigger 555. I also just read that the 555 timer cannot give me enough current (about .44A) to get the solenoid going. How do i get a pulse on the rising edge of my signal juicy enough to power my solenoid?

A bjt 555 timer requires at least 6v to operate.
You can run a CMOS 555 down to 3v, but it has puny current source/sink ability at so low a voltage.

You can use a transistor, a Darlington transistor or a power MOSFET.

Since your load is 440mA, a single bjt (bipolar junction transistor) will require a base current of about 44mA (going by the standard rule of thumb). That will be a bit much for the output of a comparator.

A Darlington would take far less current to drive, but has a Vbe of perhaps 1v or so.

A power MOSFET requires little current; it's basically a voltage-controlled switch.

More details about your existing circuit would be most helpful.

 

Hello, heryrg. Welcome to AAC.



You just need a single momentary pulse once it hits 0.7, and then again at 0.8v? Those just happen once, until some event re-sets them?

Yes i need a momentary pulse once the voltage rises to 0.7v then again at 0.8. After that the voltage will keep rising for a while and then start to decline after 8hrs or so. Once it declines to 0.8v i want to send another pulse. Once at 0.7v i want to send another pulse.

What is the desired duration of the pulse? 100mS, 300mS?

I think 300mS to 500mS will do.

You specified 1.5v 0.44A; about 3.41 Ohms. Is that the specification of the solenoid, or is that the limitations of the supply, or do you need to limit current?

Thats not the specs, that is tested value. The higher the voltage gets the more pressure the plunger applies. 1.5v and 0.44A will be enough in my case, although more voltage will not hurt. The H-bridge supply must be at least 6v 0.44A and the inputs (A&B) to it needs to be 1.5v

Will the 1.5v come from a 1.5v source that you already have available?

No i do not have the 1.5v source. I have one 2.4v or 3.6v source depending on how many batteries i will decide to use. If i can i would like to use 2.4v.

What are other supply voltages that you have available?

2.4v or 3.6 and i will have to amplify that to 6v or higher for H-bridge

OK, 1st thing: I'm wondering why you feel that you need to use an instrumentation amplifier, when a comparator can sense right down to ground quite precisely - unless there is something else I'm not aware of.

At first i was going to use the lm339, but for some odd reason i couldn't get it. I will try again. I won't amplify, just use the comparator then.

2nd thing; an LM324 is a very old and slow quad opamp, not a comparator. It does sense down to ground and can output to the negative rail, but in this case a comparator would be a better choice.

I'll try

OK, now you're throwing an H-bridge in the mix - can you be a bit more specific about the H-bridge?

See attached circuit. The H-bridge works. The supply needs to be 6v or higher. The inputs A and B need to be around 1.5v min.

A bjt 555 timer requires at least 6v to operate.
You can run a CMOS 555 down to 3v, but it has puny current source/sink ability at so low a voltage.

At the A B inputs, the H-bridge doesn't suck any current so maybe the LMC555N can be used to control the H-bridge.

You can use a transistor, a Darlington transistor or a power MOSFET.

I'll search around, any links you can provide would be appreciated??

Since your load is 440mA, a single bjt (bipolar junction transistor) will require a base current of about 44mA (going by the standard rule of thumb). That will be a bit much for the output of a comparator.

A Darlington would take far less current to drive, but has a Vbe of perhaps 1v or so.

A power MOSFET requires little current; it's basically a voltage-controlled switch.

More details about your existing circuit would be most helpful.

 

Is Motor in H br. really your solenoid?? On rising V .7 pulse comes first= F [forward] followed by .8 pulse R [ reverse] ; on falling V, .8 comes next, giving two Rs in sucession, followed by a F, is this right or should falling V sequence be inverted??

 

Is Motor in H br. really your solenoid?? On rising V .7 pulse comes first= F [forward] followed by .8 pulse R [ reverse] ; on falling V, .8 comes next, giving two Rs in sucession, followed by a F, is this right or should falling V sequence be inverted??

Yes the motor is the latching solenoid, the bridge does work.
On the rising 0.7v, a pulse will go to the B terminal thus reversing the solenoid or pulling it back, the solenoid will be pulled back so this is a repetitive and needles motion but i dont care if its easier to design. On the rising 0.8v, a pulse will go to the A terminal thus releasing the solenoid, push action. On the falling 0.8v, the pulse will repeatedly go to the A terminal releasing the solenoid again, this is a repetitive motion which i really don't care about. On the falling 0.7v, a pulse will be sent to the B terminal pulling the solenoid back.

 

If you want to run this thing from 2.4v to 3.6v, forget the MOSFETs and forget the H-bridge design that you have.

It could be used, but you would need to add a switching boost power supply, which just makes the whole thing start to look like something Rube Goldberg dreamed up.

The N-ch MOFETs in the H-bridge schematic are logic level, but not the P-ch on the upper side; they require Vgs=-10v in order to fully turn on - and at 140nC, they have a very large gate charge requirement.

While normally I'm a fan of doubling the ratings of a component for an application, in this case using a 30A rated MOSFET for an 0.44A load is like swatting a fly using a battleship.

So, tell us more about these batteries that you're considering using. I don't want to hear "AAA" or "AA" cells, either - unless you like to keep the battery companies in business.

 

SgtWookie thanks for the info.

I don't get why i can't use the h-bridge, i understand that the mosfets i have are overrated, i could defiantly replace them with lower rating ones.

When i plugged the circuit up to a power supply it drains about 20ma constant and .4a when solenoid under load which will be 4 times in 24hrs. To get the higher supply voltage for the solenoid i was planning to use an instrumentation amp INA122, a single power supply amp. Will the INA122 be able to output voltage higher then source voltage? I'm under the impression it could? Am i wrong?

The batteries i plan to use are two NiCad AA 600mAh. The reason for this being i need a battery which is ok with overcharging. This whole circuit will run off solar power. The solenoid will turn on and off in the morning and night. The batteries will be hooked up to the charging circuit 24/7 so NiCads are my best choice, i think. Correct me if i'm wrong.

If anybody thinks this is not the right way to do this, please let me know i'm open to any suggestions or comments. Thank you guys!!

 

I don't get why I can't use the h-bridge

You could use an H-bridge, but not with those MOSFETs at the voltage you want to operate them with.

MOSFETs are basically voltage controlled switches. The lower N-channel MOSFETs specified in the schematic would work OK with a 5v supply, as they are rated for operation at logic level voltages (5V DC).

However, the P-channel MOSFETs that are specified on the high-side of the bridge are standard level MOSFETs; they require -10v on the gate relative to the source terminal in order to turn the fully ON.

However, you want to run this thing on 3.6v or even lower 3.6v is not even enough to fully turn on the logic level MOSFET, so that whole idea goes out the window - unless you want to try to build a switching boost-type power supply, which will have about 85% efficiency, and add a lot of complexity unnecessarily.

Or, you could change over to a transistorized design that will not require such high voltage.

I understand that the MOSFETs I have are overrated, I could definitely replace them with lower rating ones.

Generally, you want a MOSFET that's rated for twice to three times the needed current, and at least 1-1/2 times the voltage that might be experienced in the circuit.

When i plugged the circuit up to a power supply it drains about 20ma constant and .4a when solenoid under load which will be 4 times in 24hrs.

Well, you might look at using CR123 batteries. I don't know how long you expect your batteries to last. What are you expecting in this regards? How many hours of operation until it ceases to function?

To get the higher supply voltage for the solenoid i was planning to use an instrumentation amp INA122, a single power supply amp. Will the INA122 be able to output voltage higher then source voltage? I'm under the impression it could? Am i wrong?

Instrumentation amplifiers, Operational amplifiers, comparators, etc - cannot output MORE voltage than they are supplied with. Only more modern designs can approach the positive and negative power "rails".

The batteries i plan to use are two NiCad AA 600mAh. The reason for this being i need a battery which is ok with overcharging. This whole circuit will run off solar power. The solenoid will turn on and off in the morning and night. The batteries will be hooked up to the charging circuit 24/7 so NiCads are my best choice, i think. Correct me if i'm wrong.

Let's see - 600mAh, discharged at a 20mA rate gives you a maximum of 30 hours of operation. That's if the solenoid is never engaged. The solenoid represents a very heavy load; practically a dead short across the batteries.

NiCD's are sensitive to overcharging. It's best if they are charged at a slow rate.

What is the output rating of your solar charger? It will have to both power the system and charge the batteries while it's able to produce power.

This whole thing is just getting more complex by the minute.

 

Well, you might look at using CR123 batteries. I don't know how long you expect your batteries to last. What are you expecting in this regards? How many hours of operation until it ceases to function?

The whole circuit will drain about 30ma maybe 50ma, when not loaded and loaded only 4 times during the day. I need the batteries to last about 10hrs maybe less.

Instrumentation amplifiers, Operational amplifiers, comparators, etc - cannot output MORE voltage than they are supplied with. Only more modern designs can approach the positive and negative power "rails".

Thanks for the info. So a switching boost ps is the only way to up the voltage higher then source voltage?

Let's see - 600mAh, discharged at a 20mA rate gives you a maximum of 30 hours of operation. That's if the solenoid is never engaged. The solenoid represents a very heavy load; practically a dead short across the batteries.

NiCD's are sensitive to overcharging. It's best if they are charged at a slow rate.

What is the output rating of your solar charger? It will have to both power the system and charge the batteries while it's able to produce power.

The output of the solar cell is 3.0v 40mA. I may use two of them but less is better in my case so if i can get away with one would be great. I dont think the batteries will drain in one day thus the solar cell will keep them charged, maybe not fully charged since the low current but at least ~70-80% hopefully.

Come to think of it, I might be able to get away with using a 9v NiCad, they're a bit low in mAh's the best one i found, quickly, was 8.4v 300mAh. I think the charging circuit i have is able to charge that with some minor changes. Then i'll only have two comparators, some sort of timing trigger which will send a small pulse to A or B, and the h-bridge for the solenoid. Maybe i'll have to use two batteries giving me 600mAh. Do you think this would be easier then having to up the voltage??

This whole thing is just getting more complex by the minute.

Lol BTW thanks for your help, i really appreciate it.

 

The whole circuit will drain about 30ma maybe 50ma, when not loaded and loaded only 4 times during the day. I need the batteries to last about 10hrs maybe less.

Don't forget about those intermittent 440mA bursts of raw blazing power that will be needed to actuate the solenoids. They might be short in duration, but bursts of current like that is sort of like whacking a battery with a hammer.

So a switching boost ps is the only way to up the voltage higher then source voltage?

There are many types of switching power supplies; some are remarkably efficient - up to 98%. However, if you need 20mA current at 6v, you will need over 40mA current from a 3v source. As the supply voltage goes down, the current demand increases. There is no "free lunch".

The output of the solar cell is 3.0v 40mA. I may use two of them but less is better in my case so if i can get away with one would be great. I don't think the batteries will drain in one day thus the solar cell will keep them charged, maybe not fully charged since the low current but at least ~70-80% hopefully.

Sounds to me like you definitely have a power management problem.

Do you need for this thing to work in the dark, too? How about if you don't get sun for a few days?

I think that the only way you'll be able to meet your meager power supply requirements is to use a microcontroller that has a "sleep" function; it could be programmed to snooze most of the time away while drawing very low power, and just wake up occasionally to check the voltage, and activate the solenoid if necessary. With a bit of clever programming, and if the circumstances permit, the program could heuristically decide what the optimal durations for sleeping are, to only wake up near the times when the voltage levels dictate the solenoid should be triggered.

Come to think of it, I might be able to get away with using a 9v NiCad, they're a bit low in mAh's the best one i found, quickly, was 8.4v 300mAh. I think the charging circuit i have is able to charge that with some minor changes. Then i'll only have two comparators, some sort of timing trigger which will send a small pulse to A or B, and the h-bridge for the solenoid. Maybe i'll have to use two batteries giving me 600mAh. Do you think this would be easier then having to up the voltage??

Bad idea. 9v batteries have a very high internal resistance. This would make activating the solenoid problematic.

 

Don't forget about those intermittent 440mA bursts of raw blazing power that will be needed to actuate the solenoids. They might be short in duration, but bursts of current like that is sort of like whacking a battery with a hammer.

If these bursts are only for about 500ms you think it will really drain the battery that bad?

There are many types of switching power supplies; some are remarkably efficient - up to 98%. However, if you need 20mA current at 6v, you will need over 40mA current from a 3v source. As the supply voltage goes down, the current demand increases. There is no "free lunch".

Sounds to me like you definitely have a power management problem.

Do you need for this thing to work in the dark, too? How about if you don't get sun for a few days?

I agree It will not need to operate the solenoid during the night only in the beginning of night and in early morning when the sun has been out for hour or two.

I think that the only way you'll be able to meet your meager power supply requirements is to use a microcontroller that has a "sleep" function; it could be programmed to snooze most of the time away while drawing very low power, and just wake up occasionally to check the voltage, and activate the solenoid if necessary. With a bit of clever programming, and if the circumstances permit, the program could heuristically decide what the optimal durations for sleeping are, to only wake up near the times when the voltage levels dictate the solenoid should be triggered.

Its a good idea but the circuit needs to run 24/7 and compare the output voltage which will trigger the solenoid because during the day the solenoid might have to close or open.

Bad idea. 9v batteries have a very high internal resistance. This would make activating the solenoid problematic.

You think?? i couldn't just plug that 8.4v to the h-bridge and use it. Will the mosfets i have be a problem with the 8.4 battery??

I'm thinking about using 5 AAA NiCads for now (6v) until i can find better mosfets for the bridge. Can you help me out in re-modeling the h-bridge so that it can work with less voltage source. I know you said that the upper fets need to be changes. Can you give me some hints as to what ratings i should be looking for.

 

If these bursts are only for about 500ms you think it will really drain the battery that bad?

Much of the power required to operate the solenoid will be dissipated in the battery/batteries themselves as heat. When batteries are subjected to very high loads, it drains them very quickly.

re: power management; dark operation

I agree It will not need to operate the solenoid during the night only in the beginning of night and in early morning when the sun has been out for hour or two.

So, if it turned off 2 hours after the solar cell stopped producing power and then woke up in the morning, that would work?

Its a good idea but the circuit needs to run 24/7 and compare the output voltage which will trigger the solenoid because during the day the solenoid might have to close or open.

How often might this voltage wander across the 0.7v threshold and the 0.8v threshold? Is this like a water level control or something?

Re: 9v "transistor" batteries

You think?? i couldn't just plug that 8.4v to the h-bridge and use it. Will the mosfets i have be a problem with the 8.4 battery??

The MOSFETs would start to turn on. However, the current sourcing ability of a 9v cell is so puny that it would dissipate most of the power needed for the solenoid internally.[/QUOTE]

I'm thinking about using 5 AAA NiCads for now (6v) until i can find better mosfets for the bridge. Can you help me out in re-modeling the h-bridge so that it can work with less voltage source. I know you said that the upper fets need to be changes. Can you give me some hints as to what ratings i should be looking for.

Good luck with using the AAA cells.

If you won't get realistic, there is really no point in continuing this.

 

Your 2:24 post makes better sense, ON-Off in AM, On-Off PM. 'D'nt know why-but not my problem, but here is start. First ,,if you have a large capacitor, 1000 to 10,000 uF, charge to 6V and see if it will actuate solenoid; if you can charge a "9" v battery-try one also. H-br. does not need to be powered up ,excopt on demand, 1 sec twice a day, remainder of logic draws a few mA[?]. As shown, boost is powered up only on a pulse, with time to charge a large cap. A 9V @ 300mA hr has about 1000 coulombs one relay pulse around .01 coul.

 

The whole project is for a temperature controlled solenoid using battery power .

re: power management; dark operation

So, if it turned off 2 hours after the solar cell stopped producing power and then woke up in the morning, that would work?

Yes this would work but it would have to turn off after the solar cell stopped working and the temperature related voltage would drop below 0.7v.

How often might this voltage wander across the 0.7v threshold and the 0.8v threshold? Is this like a water level control or something?

I'm measuring a temperature of a system ( i can't really go into too much detail about the system and what the system is for because my friend is trying to develop something he might patent, please understand) How many times the temperature might wander across the 0.7v and 0.8 depends on the sun exposure onto the system during the whole day, so if its sunny on off it might wander through the thresholds may be up to 3 times, not too sure yet though. I think a safe assumption would be 3 times, so turn on in the mourning(goes above 0.8v) close at 1(goes below 0.7v) or so maybe open again at 4(goes above 0.8v) and probably close at the end of the day(goes below 0.7v).

Re: 9v "transistor" batteries

The MOSFETs would start to turn on. However, the current sourcing ability of a 9v cell is so puny that it would dissipate most of the power needed for the solenoid internally.

Do you think i can design an h-bridge with a min source of about 3.6? I also think i might get a smaller solenoid with a current needed of hopefully 2 times less then current about 200ma

Good luck with using the AAA cells.

If you won't get realistic, there is really no point in continuing this.

[/QUOTE]

Honestly i am really serious about this. I might not know much about somethings but i need to complete this project, thats why im asking all the questions. BTW i'm not ignoring you bernard, i just have to get off computer and don't have time to ask you some questions. Please stand by! LOL

 

The whole project is for a temperature controlled solenoid using battery power.

I see. So, if the temperature was just checked every minute, or maybe every 10 minutes, that would probably be OK?

Yes this would work but it would have to turn off after the solar cell stopped working and the temperature related voltage would drop below 0.7v.

That wouldn't be difficult at all.

I'm measuring a temperature of a system ( i can't really go into too much detail about the system and what the system is for because my friend is trying to develop something he might patent, please understand) How many times the temperature might wander across the 0.7v and 0.8 depends on the sun exposure onto the system during the whole day, so if its sunny on off it might wander through the thresholds may be up to 3 times, not too sure yet though. I think a safe assumption would be 3 times, so turn on in the mourning(goes above 0.8v) close at 1(goes below 0.7v) or so maybe open again at 4(goes above 0.8v) and probably close at the end of the day(goes below 0.7v).

So, it really doesn't need absolutely constant monitoring; just intermittent monitoring to ensure that the general thresholds aren't exceeded.

Do you think i can design an h-bridge with a min source of about 3.6?

I know that I can. Can you?

I also think i might get a smaller solenoid with a current needed of hopefully 2 times less then current about 200ma.

Well, it's going to depend on what the solenoid is having to actuate. For example, if you're turning off a water valve that may vary in diameter for different flow rates, it'll likely take more force to actuate the larger valve.

Honestly i am really serious about this. I might not know much about somethings but i need to complete this project, thats why im asking all the questions.

Larger size batteries, such as "C" or "D" cells, will be much better able to handle such heavy loads as solenoid actuation. The problem that you'll get into trying to drive those kinds of loads with little batteries is that the load will short out the voltage supply, and the electronics won't function properly while the solenoid is engaged. Keeping a large capacitor charged up is one idea, but large capacitors have a substantial leakage rate, which causes a continual current draw that grows worse as the capacitor ages. Capacitors do not discharge in the same way as batteries do.

NiCD batteries do not last very long if they are not fully cycled (aka; completely discharged and then recharged). They develop a "memory", and as time goes on, they are able to accept/release less and less of a charge.

NiMH batteries do not exhibit this "memory" problem, although it is important that they not be overcharged. It is best to charge them at a 14 hour rate, but some types can be charged much faster if the temperature rise is monitored.

There are large areas of the planet that don't receive sun every day. Even though I live in usually-sunny Florida, we might have clouds for a week straight or more during the rainy season. If your system does not have the capability of managing its' power consumption, it will soon be non-functional.

Would you want the system to be able to "limp along" and at least maintain some control during "lean power" times? Or just run continually until the batteries are dead, and have a complete system failure?

 

I see. So, if the temperature was just checked every minute, or maybe every 10 minutes, that would probably be OK?

Yes it would be ok. I don't want to invole a uP yet, because i think i can deal without it for for the Version 1.0 of this project. For the Version 2.0 i think i will get a uP and use intermittent checking to save power and other small changes to make this more efficient. Don't get me wrong, i wana get this project done, it's just i'm close (except for new components to re-design the h-bridge for lower Vs and designing the trigger system) to completing this project.

So, it really doesn't need absolutely constant monitoring; just intermittent monitoring to ensure that the general thresholds aren't exceeded.

Yes, but like i said earlier i don't wana involve uP, mainly because i'm not great at programming. I would be able to do it but it would take me some time.

I know that I can. Can you?

I hope i can!! lol. As i understand first i'll need to change the upper P-channel mosfets to accommodate the lower Vsupply. I do this by picking the mosfets with a lower Threshold voltage (Vgs) correct?? Right now the Vgs is from -2 to -4, this means that Vg - Vs should be between -2 & -4 correct? For lower Vsupply i'll need to lower the threshold voltage correct?

Well, it's going to depend on what the solenoid is having to actuate. For example, if you're turning off a water valve that may vary in diameter for different flow rates, it'll likely take more force to actuate the larger valve.

It's moving a small plastic opener, which requires very little force. I think i will buy a new solenoid with lower power requirements, hopefully around 150mA to maybe 200mA to activate.

NiCD batteries do not last very long if they are not fully cycled (aka; completely discharged and then recharged). They develop a "memory", and as time goes on, they are able to accept/release less and less of a charge.

NiMH batteries do not exhibit this "memory" problem, although it is important that they not be overcharged. It is best to charge them at a 14 hour rate, but some types can be charged much faster if the temperature rise is monitored.

How long would you say the NiCads would last, i know its hard to say but what do you think? I would use the NiMH but the batteries will be always connected to the solar cells. Yes sometimes they won't be charging for whole day, but sometimes they will be if there are a couple of really sunny days in a row.

There are large areas of the planet that don't receive sun every day. Even though I live in usually-sunny Florida, we might have clouds for a week straight or more during the rainy season. If your system does not have the capability of managing its' power consumption, it will soon be non-functional.

Would you want the system to be able to "limp along" and at least maintain some control during "lean power" times? Or just run continually until the batteries are dead, and have a complete system failure?

Good point. If the weather will be sunless then the temperature related voltage will not go above 0.7v which means that the solenoid will not have to be activated. I think even if there is no sun ( BTW this project will be for places with a higher then average sun exposure, not places like seatle, lol) the batteries will be charging a little bit everyday.

If the solenoid does not have to be activated then the system only consumes max 50ma so my three AA NiCads will last about 36hrs, but considering that they'll charge a bit even with clouds i think it will keep limping, maybe not work at night. You made a good point i will have to discuss this with my friend for whom im designing. For now i just wana get a prototype going.

I'm still having problems with designing a circuit for rising edge trigger detection. When the temp goes above 0.7 threshold i will get about 2.5v (or higher depending on the voltage source i use) I've been looking at the LMC555 circuits but can't get it to work. Any advice or links you can give me?
Thanks for all the help!!

 

BTW What do you think of the attached circuit for creating a pulse?
I know you said the caps will leak current, in the attached circuit case do you think it will be large or does it depend on the quality of caps i use?

Also, a stupid question, the triangle in the picture is that a diode or op-amp??

 

It will produce an attenuated positive pulse. Triangle is a non-inverting buffer, a better choice might be a schmitt trigger; non inverting ??, or use two like 74C14 to re invert the inverted sig.
Is there a problem if signal V rises to .7V and holds because of a cloud cover, leaving solenoid ON ?
Caps under 500 uF +/-, generally do not pose a leakage problem. ' have a 10,000 uF used as timing with a 2.2M resistor with no leakage problem; just have to measure each one.

 

BTW What do you think of the attached circuit for creating a pulse?

It should work OK.

I know you said the caps will leak current, in the attached circuit case do you think it will be large or does it depend on the quality of caps i use?

In the case of a short-duration pulse, it won't matter much.

Also, a stupid question, the triangle in the picture is that a diode or op-amp??

Neither, really.
It's a non-inverting buffer/amplifier.

 

Can you guys help me out with picking the new mosfets. As i wrote earlier:
As i understand first i'll need to change the upper P-channel mosfets to accommodate the lower Vsupply. I do this by picking the mosfets with a lower Threshold voltage (Vgs) correct?? Right now the Vgs is from -2 to -4, this means that Vg - Vs should be between -2 & -4 correct? For lower Vsupply i'll need to lower the threshold voltage correct?

I'm not asking for you to design a new circuit, that would be nice though lol
Just for some help about mosfets.

I tested that design, for a pulse, instead of the logic buffer i just put a diode, and the circuit works well.