Hi - I am designing a circuit with two solenoids (12 V, 1 Amp each) in parallel in order to get power of two upon activation using an arduino.
I am able to wire one solenoid using 12 V adapter rated 3 Amp, 1 TIP120 darlington transistor, 1N004 diode and Arduino connected to base of the transistor. It works, however I want to add another solenoid in parallel, what else do I need in this circuit? attached is the circuit for one solenoid.

 

Welcome to AAC!

Just put the other solenoid in parallel with the first.

Note that the diode you're using is rated for 1A average, but it's surge rating will handle 2A briefly.

Some suggestions for drawing schematics. The usual direction for current flow is from left to right and top to bottom, be consistent in the use of connection dots, and try to avoid unnecessary wire jogs.

 

I generally keep 1n4007 on hand to cover all eventualities, the Onsemi spec sheet shows a the peak fwd current for 8ms is 30amps for the 1n700x series!
Max.

 

be careful with power.

a solenoid can take a fair current,

if its on for a while, they dissipate a fair heat

 

be careful with power.

Total solenoid current will be 2A, transistor will be dissipating about 2W and continuous operation wouldn't require a heatsink; but the transistor *will* get hot.

 

So, I tried the circuit. I used the same pin (13) to connect base of both Transistors. It worked, but both Solenoids and both Transistors became very hot in just a few seconds. I don't think I can make them work for 10 minutes at a stretch; I fear mu Arduino too.

Any ideas, how I can avoid the solenoids and transistors getting this hot, other than additional heat sinks. Is there a danger to arduino, I guess not since, I have 1 k resistors on both.

Why does operating one solenoid doesn't heat up both the Solenoid and Transistor?
thanks again.

 

Use mosfets instead like IRF510/540 types, and perhaps a UF5407 for back emf..

 

Try using 2 logic gate Mosfets, one for each solenoid. IRL530.
If the solenoids are fed from the rated voltage then the heating is normal.
Max.

 

Any ideas, how I can avoid the solenoids and transistors getting this hot, other than additional heat sinks. Is there a danger to arduino, I guess not since, I have 1 k resistors on both.

Measure the voltage on the emitters collectors of the transistors. If they're not saturating, power dissipation will go up.

Why does operating one solenoid doesn't heat up both the Solenoid and Transistor?

Could be insufficient base drive when driving two.

 

Yes, you don't have enough base drive to fully turn on the BJT.
It needs about 200mA base current for good saturation.

One logic-level MOSFET with an ON resistance of ≤.1Ω and one 1N4004 diode should be sufficient to carry the 2A from both solenoids.
I see no reason to use two of either.

But note that the MOSFET must be a logic-level type that will fully turn on with the 5V from the Arduino.
The IRF510/540 types require a gate-source voltage of 10V to fully turn on.

 

thanks everyone! great forum indeed.

 

Yes, you don't have enough base drive to fully turn on the BJT.
It needs about 200mA base current for good saturation.

That can't be right, can it? It's a Darlington with Hfe of 1000. So, theoretically, it would only take 2mA at base to provide 2A output, right? I know it's better to provide plenty of headroom, but not a factor of 100.

I'd like to see a schematic of how the two are wired up. Are they sharing 1 resistor? Could the current through the bases of the two Darlingtons be fluctuating wildly with nothing controlling how they share the current from the Arduino output?

I don't disagree with everyone's MOSFET advice, but I also think this project should be very manageable with the existing Darlingtons.

 

That can't be right, can it? It's a Darlington with Hfe of 1000. So, theoretically, it would only take 2mA at base to provide 2A output, right? I know it's better to provide plenty of headroom, but not a factor of 100.

Minimum beta is 1000 under these conditions:




This graph shows typical beta of a few hundred at 2A when used as a switch:

 

That can't be right, can it? It's a Darlington with Hfe of 1000. So, theoretically, it would only take 2mA at base to provide 2A output, right? I know it's better to provide plenty of headroom, but not a factor of 100.

Yup.
It was drawn as a standard transistor and I forgot that the TP120 is a Darlington.

They use a forced Beta of 250 for the saturation voltage tests in the data sheet, so that would mean you should use a base current of a least 8mA for switching 2A.
Since the Vbe(sat) value in the data sheet is about 1.6V@2A, the 1kΩ base resistor shown would give only about 3.4mA base current, thus the base resistor should be reduced to around 400Ω.

I don't disagree with everyone's MOSFET advice, but I also think this project should be very manageable with the existing Darlingtons.

As long as you don't mind it dropping close to volt and getting a little hot (unless it's on a heatsink) when conducting 2A.

 

Yup.
It was drawn as a standard transistor and I forgot that the TP120 is a Darlington.

They use a forced Beta of 250 for the saturation voltage tests in the data sheet, so that would mean you should use a base current of a least 8mA for switching 2A.
Since the Vbe(sat) value in the data sheet is about 1.6V@2A, the 1kΩ base resistor shown would give only about 3.4mA base current, thus the base resistor should be reduced to around 400Ω.
As long as you don't mind it dropping close to volt and getting a little hot (unless it's on a heatsink) when conducting 2A.

Sweet! Thanks for the clarification and the calculations. I still have a lot to learn and I'm always afraid that I'll lead someone astray if I jump in when I shouldn't.

And yeah, the MOSFET would be better for sure - I'm just curious what's going wrong in the OP's current attempt, because two Darlingtons driving 1 amp each shouldn't be struggling at all, unless something isn't wired right.

 

Minimum beta is 1000 under these conditions:
View attachment 130181
View attachment 130179
View attachment 130180

This graph shows typical beta of a few hundred at 2A when used as a switch:
View attachment 130178

Am I misreading the graph? It looks to me like that graph shows 2A collector current with only 2mA of base current while operating at just over 1V collector-emitter drop, very close to its minimum.

 

Am I misreading the graph? It looks to me like that graph shows 2A collector current with only 2mA of base current while operating at just over 1V collector-emitter drop, very close to its minimum.

My design style is more conservative than yours. I'd use at least 8mA.

Consider that the graph is for a typical device. For the design to function without cherry picking, you need to be conservative.

 

My design style is more conservative than yours. I'd use at least 8mA.

Consider that the graph is for a typical device. For the design to function without cherry picking, you need to be conservative.

Oh, I agree completely with playing it safe and providing the extra base current. If I were designing this circuit, I'd definitely do what you and crutschow said, providing 8mA or more for 2 amps (4mA or more for 1,) or just using a suitable MOSFET.

My comment was just because I didn't see where the graph shows what you were describing. The graph seems to support the idea of 1000x current gain (which a cautious engineer would then derate in order to play it safe.)

More to the point, if the OP can operate one coil with one Darlington without issues, there should be no problem operating two coils with two Darlingtons, unless there's some other issue, like sharing one base resistor for both bases.

If the OP can pick up some MOSFETs, great. If not, my advice to the OP would be to either use one Darlington with the base resistor adjusted to provide at least 8mA base current, or use the two Darlingtons and make sure each has its own base resistor, sized to provide over 4mA of base current for each.

 

My comment was just because I didn't see where the graph shows what you were describing.

I picked the base current for the lowest saturation voltage specified by the manufacturer.

The graph seems to support the idea of 1000x current gain (which a cautious engineer would then derate in order to play it safe.)

Not for saturation. I always design for minimum saturation voltage to minimize power dissipation in the transistor to maximize power to the load.

 

I picked the base current for the lowest saturation voltage specified by the manufacturer.

Not for saturation. I always design for minimum saturation voltage to minimize power dissipation in the transistor to maximize power to the load.

Gotcha. Like I said, I have a lot to learn. Thanks for the help!