Hi All.

I found a temperature circuit online and have modified it to suit my project.
I had to modify some of the values to make it reach temperatures of 120deg Cel.

I have 2 questions please if anyone can help?

1.) Does the circuit and values look ok to use? I need it to be reliable.

2.) I cant seem to find a base resistor that stops the transistor from getting very hot. Could anyone help with that please?
I have been using a BC557 pnp transistor simply as I have some. From my meter measurements the output from the LM339 goes from 12v+ to almost 0volts when triggered.

Thanks for any help.
Darren.

 

The two pnp transistors are being turned on as the LM339 output terminal's open drain structure activates and conducts the base current to ground. Trying to proportionally control the primary transistor current by limiting the base current may not be the best approach. The relay only requires 150 ma current. Try adding a series resistor at the node between the relay and the transistor output, at the anode (top) of the diodes, labeled D1 and D2. A 47 ohm, 5 watt, would be a good load resistor to start with. Just measure the voltage drop across the resistor in order to verify the magnitude of the current. If you don't happen to have a 47 ohm, 5 watt resistor, then two 100 ohm, 5 watt resistors twisted together at the ends would work just as well.
... As to a reasonable value for the resistor at the base of the pnp transistors, the BC557 datasheet specifies a maximum base current of 200 ma. Therefore, try using a resistance of 200 ohms in series with the transistor base. The resulting 60 ma base current would result in approximately 0.72 watts power, so two 1 watt, 100 ohm resistors in series would be adequate at the transistor base.
... just trial values suggested here, so please report any unexpected observations.

 

@drc_567
Thank you for your your help.

I tried with a 220ohm on the base and a 56ohm on the relay (sorry they the closest I had.)
The transistor became hot just powering the circuit with triggering the lm399.
Are these values close enough or do I need to be exact?

 

@drc_567
Thank you for your your help.

I tried with a 220ohm on the base and a 56ohm on the relay (sorry they the closest I had.)
The transistor became hot just powering the circuit with triggering the lm399.
Are these values close enough or do I need to be exact?

What is collector to emitter voltage of the transistor when it has switched on the relay?

 

... something is not right ... follow up on AlbertHall's inquiry.

... thinking that the current gain is too great ... Try 10 kohm on the transistor base.

 

... one more thing to check out ... sometimes happens .... Get the transistor datasheet and verify that the transistor emitter, collector, and base pins are actually where they are supposed to be.

 

If I use a 10k the trans stays cool - the relay switches on ok but it doesn't turn off again untill the temp goes does to like 80deg.
I need it to go on at like 110 and of at 105.

I have 12v across the em/col with the 10k. and around 6.5 when it drops the relay.

 

but it doesn't turn off again untill the temp goes does to like 80deg.
I need it to go on at like 110 and of at 105.

Even with the 500k pot set to maximum?

 

... just to see if there is a difference, try adjusting the hysteresis potentiometer to the minimum resistance.

 

The external power source ( + and - terminals) is connected backwards to most of the circuit. In a circuit like this, PNP transistors can function as a saturated switch, pulling the load *up* to the + terminal.

If your goal is to have proportional fan speed control with a pot, the circuit has several errors. Linear speed control (as opposed to pulse-width switching) is entirely possible, but the output transistors will need heatsinks.

drc - an LM339 cannot sink 60 mA reliably.

ak

Edit - removed LM339 connection comment. Thanks, AH.

 

There are some serious deficiencies in the provided diagram. I have attached a marked up version of part of it with the positive supply marked in red and the negative in green. The '339 is connected correctly to these supplies, but the PNP transistors have their emitters connected to the negative supply which cannot be correct. Further the transistors are connected in parallel with the relay coils and there is no power connection for the coils.

 

As the relays apparently do switch, the diagram in post #1 does not correctly represent the actual circuit.

 

The BC557 is rated for only 100 mA of continuous collector current, and your relays draw 146 mA each. Obviously this is not a reliable situation. While a better small transistor such as a 2N4403 has the right ratings for this application, you will get much better long-term reliability with a small power transistor such as a TIP30 or TIP32. Select R9 and R10 for 10-15 mA of base current.

The schematic is more of a wiring diagram, making it difficult to determine the intent of the designer. It appears to be an on-off controller rather than proportional. When the other issues get corrected, the transistors no longer should overheat.

The unused LM339 input pins must be connected to something, and connecting all of them to GND is not recommended. Better to connect them to other circuit points and leave the outputs open. For example:
Pin 8 > pin 7
Pin 9 > pin 6
Pin 10 > pin 5
Pin 11 > pin 4
Yes, the + inputs are connected to other - inputs and vice versa. Irrelevant.
ak

 

First pass, this is my interpretation of one channel of what the circuit is supposed to be. I have not done any calcs on the bridge values to see if they are correct. I converted the bridge resistors to the nearest 1% tolerance equivalent for better stability.

Note - U1 needs decoupling capacitors. I recommend a 10 uF electrolytic and 0.1 uF ceramic in parallel, as close as possible to pins 3 and 12.

ak

 

With VR2 at 50%, the inverting input reference voltage is 4.692 V.

When the ambient temperature is below the trip point, the comparator output is open circuit. Any hysteresis current is coming through Q2, not a good design. As VR3 is decreased for more hysteresis, it is possible for the current path Q2 > R10 > VR3 > RT1 > R5 > GND to cause Q2 to turn on even though the comparator has not tripped.

With no hysteresis (VR3 at its max value), 4.7 V at the sensor input equates to a sensor current of 3.31 mA. Side note - too high - causes self-heating errors.

This yields a thermistor resistance of 855 ohms. For a typical 10K thermistor, the equivalent temperature is around +90 C. Seems very high for an adjustment range mid-point. Commercial axial fans are not rated for operation above +70 C, and at that temp their operating life is greatly reduced.

Razzle - what temperature adjustment range do you need for your application?

ak

 

Razzle - what temperature adjustment range do you want?

From post #7:

I need it to go on at like 110 and of at 105.

 

If I use a 10k the trans stays cool - the relay switches on ok but it doesn't turn off again untill the temp goes does to like 80deg.
I need it to go on at like 110 and of at 105.

To confirm post #1, degrees C?

If the control circuit is in that environment, you need to change the comparator to an LM139.

Also, do you have a part number for the thermistor?

ak

 

Hi all.
Thanks so much everyone for your interest and help.
Been a couple of busy days with doctors and stuff. I picked up a tip30 trans today and I'll implement these suggestions soon as I can and let you know how I go.

Thanks again.
Darren.

 

P.S The circuit is to run thermatic fans in the car. So I want one to come on @ 110 and off @ 105ish and one on @ 115 and off @ 110 ish Celsius.
@AnalogKid whats the difference with the LM139?
The thermistor is a tkdc17c24.

 

Datasheet page 2:

LM339 - Commercial temperature range: 0C to +70C
LM239 - Industrial temperature range: -40C to +85C
LM139 - Military temperature range: -55C to +125 C

Please post the thermistor datasheet.

ak