I am trying to use transistors as a switch that can be used to remove a load from the circuit at currents greater than 1 amp and temperatures greater than 40 degree celsius. My original idea was to use the transistors to create an and gate by connecting two transistors by joining the emitter of an NPN transistor to the collector of another NPN transistor. Next, the load would be placed closer to the high voltage rail. Basically like this diagram attached but not exactly. Each one of those transistors would receive a signal from the base to turn it on provided that the temperature is less than 40 degree and 1 amp. The voltage rail would be a regulated 9V instead of a 6V. The problem being formed is that i want the pd across the load to be 9V, however the transistors would inhibit this.

I am asking for possible solutions or a direction where i can read to obtain knowledge to solve this.

Thank You

~ yes i am new to using transistors as switching circuits.

 

What has a 40C temperature? The load? The transistors? Surroundings?
If you switch the load off, so that 1A no longer flows, what happens then?

 

The load has the 40 degree celsius. I would like it such that the circuits used to create the voltage regulation, the current sensing and the temperature sensing remains undamaged (ie, the circuit allows the current to flow but the load is cut off). You dont have to let the circuit look exactly like the one in the diagram.

~Anymore queries?. My apologies if i am not using the correct terminology.

 

There will be voltage drop across any solid state switch, which then means it's just a matter of how much resistance is acceptable. If you absolutely require a full 9V drop, and you cannot increase your regulated power, then you will have to use a relay, or similar mechanical switch that will not cause a significant voltage drop.

However, if you are drawing low currents, there are transistors that have low on resistances. For example: http://www.digikey.com/product-detail/en/NTD4906N-35G/NTD4906N-35GOS-ND/2194521. This transistor has an on resistance of 5.5 mOhm at 30A, which equates to a 0.165 V drop. At 15A, you can halve that drop, which is miniscule for many applications.

 

remove a load from the circuit at currents greater than 1 amp and temperatures greater than 40 degree celsius

Must either (independently) or both (simultaneously) conditions apply to cause load removal?
If the load current reaches 1A do you want the load to switch off and stay off, or do you simply want the load current to be limited to a steady 1A?

 

Thank you b1u3sf4n09. I will look into it.

Must either (independently) or both (simultaneously) conditions apply to cause load removal?

Alec_t, i mean either. I want it such that it works like an "and" gate whereby, for the load to be connected, the temperature must be less than 40C and the current < 1A. If any of these conditions are not met, then the load is disconnected.

If the load current reaches 1A do you want the load to switch off and stay off, or do you simply want the load current to be limited to a steady 1A?

If the load current reaches 1A, i want the load to switch off and stay off until the current falls below 1A. Not necessarily limiting the 1A but it is a good suggestion thank you. (it would make sense because the circuit would try to limit the current and if it reaches greater than 1A the load can switch off)

 

Would a relay be better or transistors? What are the advantages and disadvantages? When would you choose a relay over a transistor for switching?

 

If the load current reaches 1A, i want the load to switch off and stay off until the current falls below 1A.

From which I infer that you want it to switch on again when the current is below 1A? A problem with that is that as soon as you switch the load off the current will immediately below 1A, so the circuit will switch on again, then off, then on, ....... Is that acceptable?

Would a relay be better or transistors?

Depends what the load is.

What are the advantages and disadvantages? When would you choose a relay over a transistor for switching?

Relays are bulky and usually more expensive than transistors, but can easily provide isolation between the controlling circuit and the load circuit. Transistors are more compact, extremely fast to respond to control signals, don't suffer from contacts wearing out, generallly draw less control current than relays.

 

-From which I infer that you want it to switch on again when the current is below 1A? A problem with that is that as soon as you switch the load off the current will immediately below 1A, so the circuit will switch on again, then off, then on, ....... Is that acceptable?

No, i dont think that it is acceptable that the load would turn on and off, then on then off again. I was hoping for it to be such that the circuit would receive a signal such that if the current is too high, then it would isolate the load and remain off until the current drops below 1A. Is there a way to ensure that the load remains off intead of on and off again, and again?...or is it too complex to do with a transistor compared to a relay?

-Depends what the load is.

The load would be 2, 5 ohm 5 W resistors in series.

Thanks

 

if the current is too high, then it would isolate the load and remain off until the current drops below 1A

It is easy to arrange to switch off the load when the current reaches 1A, then keep it off indefinitely. By removing the load the current will then be <1A; in fact it will be zero. But you said in post #3 "the circuit allows the current to flow but the load is cut off". So what current do you envisage will still flow? And what should happen once the load temperature has dropped below 40C? I'm still not clear what you want the circuit to do. In a conventional thermostatted circuit the load would be maintained at 40C plus or minus a small amount by cyclically switching it on and off.

Edit: If your load is 10 Ohms and the supply is 9V the current cannot exceed 1A.

 

These are the transistor i am looking at for Tr3: http://www.alldatasheet.com/datasheet-pdf/pdf/21681/STMICROELECTRONICS/2N3439.html? , since it seems good for switching and has a low Vce
For Tr2 and Tr3 : https://www.fairchildsemi.com/datasheets/TI/TIP31C.pdf , Since it could handle large voltages,

any other suggestions? Criticism with reason is very much welcome

Thanks again

 

It is easy to arrange to switch off the load when the current reaches 1A, then keep it off indefinitely. By removing the load the current will then be <1A; in fact it will be zero. But you said in post #3 "the circuit allows the current to flow but the load is cut off". So what current do you envisage will still flow? And what should happen once the load temperature has dropped below 40C? I'm still not clear what you want the circuit to do. In a conventional thermostatted circuit the load would be maintained at 40C plus or minus a small amount by cyclically switching it on and off.

Sorry about that, I did not show you the entire circuit.
Ask more questions if you like.
attached is a pdf showing you my regulator and how the temperature sensing and current sensing comes into play.

Your comments are much appreciated

 

This is the entire circuit. Forgive my handwriting. I shall clarify whatever you want.

If this is correct, or you agree, can you give me an idea on what kinda of transistor i should choose and why.

Is there a way to do it better?

When a transistor blows, is it open circuited or short circuited?

 

There are a number of problems with your post #13 circuit.
1) The inputs to the comparator which drives the base of Tr1 should be swapped over.
2) The trimmer in series with the 5.1V zener serves no obvious purpose. Voltage adjustment should be done with the voltage-divider resistors.
3) If Tr1 and Tr2 both turn on there is nothing to limit the current through Tr1/Tr2/Tr3 base.
4) There is no hysteresis around the opamps and the circuit may oscillate.
5) The load current will not switch off completely (which is what I understand you to want) if 1A is exceeded or 40C is exceeded, because of the lack of hysteresis.

Transistors can fail either short-circuit or open circuit.
Personally I would choose a low Rds(on) logic-level N-MOSFET as the load-switching transistor, rather than a junction transistor.

 

The supply is a DC by the way.

1) The inputs to the comparator which drives the base of Tr1 should be swapped over.

My mistake, instead i shall swap where the wires connect to the current sensing resistor.

2) The trimmer in series with the 5.1V zener serves no obvious purpose. Voltage adjustment should be done with the voltage-divider resistors.

I added in a potentiometer so that i can vary the current through the zener diode so that i get exact 9V as the regulation. I used it to account for the tolerances of the other 3 resistors as well and the temperature that can affect the circuit. Something like a last minute tweaking.

3) If Tr1 and Tr2 both turn on there is nothing to limit the current through Tr1/Tr2/Tr3 base.

??

4) There is no hysteresis around the opamps and the circuit may oscillate.

What do you mean?

5) The load current will not switch off completely (which is what I understand you to want) if 1A is exceeded or 40C is exceeded, because of the lack of hysteresis.

I shall look more into this term "hysteresis" , thank you.

Transistors can fail either short-circuit or open circuit.

so i guess...it is unpredictable?

Personally I would choose a low Rds(on) logic-level N-MOSFET as the load-switching transistor, rather than a junction transistor.

Thank you for your suggestion, i will research this some more now.

 

The suggestion for the " low Rds(on) logic-level N-MOSFET" as the load-switching transistor, rather than a junction transistor, was an excellent suggestion thank you!!

Just a thought, maybe Tr3 could be a low Rds(on) logic-level N-MOSFET, and to minimise the Rds, you could use an op- amp which saturates the voltage signal, going into the gate of the mosfet, to increase Vgs.

 

you could use an op- amp which saturates the voltage signal, going into the gate of the mosfet, to increase Vgs

A rail-to-rail output type opamp would be much better than an antiquated 741! The Tr1/Tr2 arrangement would need modifying to drive the FET gate, otherwise if just Tr2 were on the FET would turn on, due to the Tr2 base current.

 

This is the diagram that i am using to turn the load on and off. However i am encountering some problems. Basically, the load would not turn off, even with the two mosfests off. suggestions?...are these choices in components sufficient?

I am going to try and use a lower potential than ground for Vee for the comparator to see if the problem can be fixed

for this circuit, i am using the BUZ330 n-channel mosfet and the MC33174P comparator ic.
datasheet for :
BUZ330 - http://pdf1.alldatasheet.com/datasheet-pdf/view/44990/SIEMENS/BUZ330.html
MC33174P - http://pdf1.alldatasheet.com/datasheet-pdf/view/12027/ONSEMI/MC33174P.html

 

Opamp offset is probably causing the opamp output to be well above 0V. Instead of grounding the inverting input of the opamp, connect it instead to a potential ~ half of the supply voltage (by using a potential divider formed of two equal-value resistors ~ 10k-100k).

 

It worked!!. Thank you very much for your help. I actually kept the inverting terminal grounded, used the biasing transistor as 1Kohm and i operated the comparator at +ve and -ve 20V instead of ground and Vsupply. At this voltage, it was sufficient enough to switch off the load (but not totally since there was a 50mV drop across is).

Referring to the "switching diagram" above,
-If both transistors were on, then i got an 8V drop across the load (1 V loss due to the resistance of the FET) . I believe about 2micro amperes flow through the FETs with both transistors on.

-If the temperature sensing FET was off while the current sensing FET was on, i got a 0V across the load.

- If the temperature sensing was on while the current sensing was off i got the 50mV drop across the load. Interesting enough, if i operated the comparator at 15V, instead of 20V, i got 3V instead of 50mV.

-With both transistors off, this gave a 0V as well.

I am currently thinking of ways that i can improve it.

Do you think these are good results?
Any suggestion on how to improve it?


By the way, my reasoning for why 50mV was observed could be because, the gate of the current sensing was soo negative that it inhibited the current from flowing through the 2 FETs. Since little current was flowing, then there would be a diminished pd across the biasing resistor. The comparator would not be able to detect the difference in the voltage and therefore output 0V. The 3 V was observed when operating the comparator at 15V because the comparator has a particular gain associated with it (practically not theoretically) and therefore, the difference was too small to amplify, so it did output a +ve voltage, but very small.

What would happen if the difference between the two terminals are 0V?
Is this correct?
What do you think?...Can you account for the 3V when the comparator operates at 15V?