Home made constant current circuit

[kuksul08]

I followed an instructable and tried making my own constant current circuit.

Here it is:

I have a couple questions though:

1. Electronic gurus...does this look like a very efficient design? When I hooked up power to it, the large transistor got VERY hot.

2. According to the ratio/equation, the circuit should put out roughly 890mA. When I hook up 4.8V source, it puts out 740mA. When I hook up a 7.2V source, it puts out 860mA, and when I hook up 9.6V source, it puts out 880mA. Does this sound relatively accurate for this type of circuit, or should it be more consistent. I expected a much flatter current.

Thanks :)

 

[kuksul08]

Here's the instructable BTW

http://www.instructables.com/id/Power-LED_s---simplest-light-with-constant-current/

 

[Norm]

I have a couple questions though:

1. Electronic gurus...does this look like a very efficient design? When I hooked up power to it, the large transistor got VERY hot.

2. According to the ratio/equation, the circuit should put out roughly 890mA. When I hook up 4.8V source, it puts out 740mA. When I hook up a 7.2V source, it puts out 860mA, and when I hook up 9.6V source, it puts out 880mA. Does this sound relatively accurate for this type of circuit, or should it be more consistent. I expected a much flatter current.

Thanks :)

Part answer to Q.1

Maximum limits:

the only real limit to the current source is Q2, and the power source used. Q2 acts as a variable resistor, stepping down the voltage from the power supply to match the need of the LED's. so Q2 will need a heatsink if there is a high LED current or if the power source voltage is a lot higher than the LED string voltage. with a large heatsink, this circuit can handle a LOT of power.

The Q2 transistor specified will work up to about 18V power supply. If you want more, look at my Instructable on LED circuits to see how the circuit needs to change.

With no heat sinks at all, Q2 can only dissipate about 1/2 watt before getting really hot - that's enough for a 200mA current with up to 3-volt difference between power supply and LED.

 

[kuksul08]

Part answer to Q.1

:D Yeah, I read that part. Maybe I should have asked: is that normal for transistors to get hot? and... does it damage them?


edit: also any suggestions/comments for my soldering would be helpful. The first time I tried this it was a disaster, and I had to redo it. Actually the first 3 times were disasters :(

 

[Mr Happy]

is that normal for transistors to get hot?

No :shakehead

and... does it damage them?

Yes :poke:

 

[rushnrockt]

:D Yeah, I read that part. Maybe I should have asked: is that normal for transistors to get hot? and... does it damage them?

You are good for up to ~175C, so yeah, it can get really hot. With the TO-220 package, you can easily attach a VERY large heatsink as necessary.

 

[Norm]

:D Yeah, I read that part. Maybe I should have asked: is that normal for transistors to get hot? and... does it damage them?


edit: also any suggestions/comments for my soldering would be helpful. The first time I tried this it was a disaster, and I had to redo it. Actually the first 3 times were disasters :(

I would have said that it is normal for a transistor that requires a heat sink and doesn't have a heat sink to get hot and as already stated heat will damage a transistor, keep in mind that the maximum operating temperature for some power transistors is over 100 C :devil: hot enough to take the skin off you fingers.
Norm
Beaten to the punch, I need to type faster :) or waffle less :ohgeez:

 

[Mr Happy]

I followed an instructable and tried making my own constant current circuit.

Congratulations. Excellent way to learn :thumbsup:

I have a couple questions though:

1. Electronic gurus...does this look like a very efficient design? When I hooked up power to it, the large transistor got VERY hot.

Well no, it is not super efficient. Power is lost in the current set resistor R3 and in the transistor Q2. This doesn't matter much for a mains powered design as shown, but it is not ideal for a battery powered design.

2. According to the ratio/equation, the circuit should put out roughly 890mA. When I hook up 4.8V source, it puts out 740mA. When I hook up a 7.2V source, it puts out 860mA, and when I hook up 9.6V source, it puts out 880mA. Does this sound relatively accurate for this type of circuit, or should it be more consistent. I expected a much flatter current.

It depends what you used as a current drain. Did you connect an LED, or did you just short the output? In any case, the word "constant" is relative. It might be better to think of such simple circuits as "constantish" current circuits. But variation of a 50 mA or so is not going to matter for an LED driver though.

 

[kuksul08]

Interesting.

So my friend just informed me that the little piece of clear plastic that came with the transistor was actually a 'mica insulator'. I threw it away... what now?

I made this little heatsink, but with no airflow it's only going to do so much.

 

[MorePower]

Why not just use an LM317 or an LM338 linear voltage regulator as a constant current source? Take a look at http://www.reuk.co.uk/LM317-Current-Calculator.htm for more info.

 

[kuksul08]

Congratulations. Excellent way to learn :thumbsup:

Well no, it is not super efficient. Power is lost in the current set resistor R3 and in the transistor Q2. This doesn't matter much for a mains powered design as shown, but it is not ideal for a battery powered design.

It depends what you used as a current drain. Did you connect an LED, or did you just short the output? In any case, the word "constant" is relative. It might be better to think of such simple circuits as "constantish" current circuits. But variation of a 50 mA or so is not going to matter for an LED driver though.

Yes although I may not end up using it, and wasted about $9, it was fun soldering and learning about it :)

I did not use a load, just shorted it. The last time I put an LED on there I didn't test it first, and it blew it up.... my only led :mecry:

Thanks

 

[Mr Happy]

also any suggestions/comments for my soldering would be helpful. The first time I tried this it was a disaster, and I had to redo it. Actually the first 3 times were disasters :(

Yes, I noticed you seemed to have a bit of trouble with that. You need to read all the instructions on how to make good solder joints, and practice.

If you are still using that lead-free solder junk, put it aside and get some nice 60/40 tin/lead solder. Make sure it is rosin-cored electrical solder in a small wire, 0.5 mm or 1 mm. If you have any trouble melting the solder so that it flows easily your iron is not hot enough.

 

[kuksul08]

Why not just use an LM317 or an LM338 linear voltage regulator as a constant current source? Take a look at http://www.reuk.co.uk/LM317-Current-Calculator.htm for more info.

That looks better :whistle: I saw an LM317 at radio shack for cheap too. Oh well, there must be some advantage to using this one?

 

[kuksul08]

Yes, I noticed you seemed to have a bit of trouble with that. You need to read all the instructions on how to make good solder joints, and practice.

If you are still using that lead-free solder junk, put it aside and get some nice 60/40 tin/lead solder. Make sure it is rosin-cored electrical solder in a small wire, 0.5 mm or 1 mm. If you have any trouble melting the solder so that it flows easily your iron is not hot enough.

Hah, you should see my last attempt. It's really sad, but for my first or second time soldering ever... yeah it still sucked haha. http://i106.photobucket.com/albums/m267/kuksul08/IMG_1699Large.jpg

Is the lead stuff really that much better? Does it just stick and flow easier at the cost of being toxic?

 

[Mr Happy]

So my friend just informed me that the little piece of clear plastic that came with the transistor was actually a 'mica insulator'. I threw it away... what now?

I made this little heatsink, but with no airflow it's only going to do so much.

You have much to learn, grasshopper :)

Actually, you want a heat sink to be large and open, rather than an enclosure round the circuit. The method is to apply some special thermal conductive grease to the back of the transistor and then bolt it tightly to the heat sink using the hole provided. The mica washer is for those cases where the transistor has to be insulated from the heat sink. This is not always necessary and can be omitted if the heat sink is free floating and will not touch any wires or live parts.

 

[spencer]

The lead free solder really should only be used in plumbing. If you want to solder electronics decently then the lead stuff is the only way to go. Also, if you are using a big soldering gun (like more than 50 watts) then go to Radio Shack and pick up a cheap pencil iron for like 15 bucks. It will solder many times better then a big gun with a big tip.

 

[kuksul08]

You have much to learn, grasshopper :)

Actually, you want a heat sink to be large and open, rather than an enclosure round the circuit. The method is to apply some special thermal conductive grease to the back of the transistor and then bolt it tightly to the heat sink using the hole provided. The mica washer is for those cases where the transistor has to be insulated from the heat sink. This is not always necessary and can be omitted if the heat sink is free floating and will not touch any wires or live parts.

Indeed:crackup: Ya know, at first I did exactly like I saw in an instruction video. I touched the soldering iron to the pad/resistor lead, then on the opposing side I touched solder and it flowed around completely, making a nice conical pyramid and coating the copper. But when I tried connecting multiple leads, it got ugly and globbed up.

I considered the larger heatsink, and surrounding the circuit will simply bring the heat back around heating up the air quickly, but there are two things that came to mind: size restrictions, and if there is zero air flow, will it really make a difference? I would think the mass of the heat sink becomes more important when it has no airflow, so location doesn't matter. I dont know, just thinking :thinking:

Thanks for the advice :)You have always been helpful. I will eventually use some silicone based heatsink compound to connect it.

 

[kuksul08]

The lead free solder really should only be used in plumbing. If you want to solder electronics decently then the lead stuff is the only way to go. Also, if you are using a big soldering gun (like more than 50 watts) then go to Radio Shack and pick up a cheap pencil iron for like 15 bucks. It will solder many times better then a big gun with a big tip.

Good thing to know. I may have some back at home, but all I currently have is this lead free stuff.

I have a weller wlc-100 adjustable power (from about 10-40W i believe). The tip is more like a straight slot screwdriver though.

 

[Mr Happy]

Is the lead stuff really that much better? Does it just stick and flow easier at the cost of being toxic?

Lead-free solder:

Tin-lead solder: :cool:

It's not actually toxic unless you swallow it. One of those rechargeable spotlights with an SLA battery has thousands of times more lead in it than a little bit of solder. The whole lead-free solder thing is a complete crock, dreamt up by politicians with nothing better to do than waste people's time and money. To get an idea how good lead-free solder is, you only have to look at the list of organizations with exemptions from needing to use it: police, emergency services, medical equipment manufacturers, military, security services, government agencies, aerospace industries ... in short, anyone who needs electronic equipment to work reliably without failing.

 

[likeguymontag]

Lead-free solder:

Tin-lead solder: :cool:

The charging port on my last cellphone came unsoldered from the board inside. It was supposedly a problem with the lead-free solder used, as my phone was one of the first generation of electronics subject to European RoHS regulations. I've never soldered with lead free, and I don't plan to unless I'm doing plumbing. Lead is only toxic if you eat it, and I support efforts to keep it out of our water supply, but RoHS... :mad:

Oh, and so that this post has some useful content: To the OP, you should use rosin-based flux when you solder. It helps the solder stick and flow. I always dip my wires in flux before I tin them.