Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Want to build large LED chasing, how To drive LED?

Status
Not open for further replies.

wilykat

Member
I plan to use a simple 2 flip flop design to make chasing LED:

8815-flipflop.jpg


Which should result in:

8816-LED_chase.gif


And there will be around 60 at the time. The V supply is 5v and I'm trying to figure out the way to drive about 15 LED per output. Since the flip flop circuit isn't designed to handle that many LED, and I don't want to build 8 sets of flipflops too much space wasted.

LEDs are max 2.5v 30mA and I plan to limit to 2v 20mA to keep them from blinding people and being distracting.

There are 3 possible ways. One is using H bridge like this:

8817-hbridge_LED.jpg

(forgot to add resistors to transistor base)

4 transistors control half of all the LEDs, one set of 15 lights up and other is off and since both are connected to just 1 flip flop it will always be alternating. A second set of H bridge controls the remaining LEDs. On paper it looks good.

But I'd need transistors that are capable of around 3A each minimum. Advantage: since transistor has typical 1.4v drop each or 2.8 total so I'd only need to drop .2v or so but finding resistor that is high watt enough is going to be tricky. Would zenier diode in reverse bias at Vcc work to lose that 0.2v? Or is there something else? Or I could just leave LED at 2.2v and pray it won't be too bright anyway.

There's also using MOSFET instead of transistor but I have no experience with them. So I don't know what are the typical V drop from source to drain or what are required to use them. If you think MOSFET is better, please advise.

3rd option is to use 4 transistor or MOSFETs, each one controls set of LEDs. Fewer transistors total:
8818-one_transistor.jpg



I'd have to drop quite a bit, about 1.6v so 2 regular diodes in series from Vcc? Or I could use 16 LEDs in 2 parallel x 8 series configuration? About 1.7v each LED this way so defiantly won't be bright but still usable.

Any advices would be appreciated. :D
 
Last edited:

BrownOut

Banned
You'r first two options have no merit. The last one is correct, but you need series resistors, one for each LED. You need to limit the current, not the voltage. Find out what current should be for the LED's then use one resistor per LED of the appropirate resistance.
 

wilykat

Member
Then I would need to either wire individual resistors to each LED directly or have 60 wires from the circuit board to the LED. Those would take up quite a bit of space along the LEDs. Assume 2v 20ma per LED on a 5v circuit and taking in account the transistor takes 1.4v I would need about 80 ohms per resistors, or 82 nearest standard value. Also what transistor should I get that can handle 15 LEDs? I missplaced my handy transistor book and when I tried googling I kept getting lots of web site that wants to sell bunch of stuff without listing any specs.
 

BrownOut

Banned
First of all, make no assumptions. The voltage drop through a transistor is about .7V typically. I don't know where you're getting 1.4V from. Find out for sure what the current and voltage are for your LED's. Different colors have different ratings. Then you can caluclate the value of the series resistor. If you devices are really 20mA, then that's about 300mA total. Most general purpose transistors can handle that current. But you can always use more than one, if the current gets too close to the edge, just connect 7 LED's to one transistor and 8 to the other. Otherwise, keep looking over datasheets. Try to find the websites of major transistor manufactures, like National Semi, etc.
 

audioguru

Well-Known Member
Most Helpful Member
First of all, make no assumptions. The voltage drop through a transistor is about .7V typically. I don't know where you're getting 1.4V from.
I don't know where your 0.7V comes from. A little 2N4401 or 2N4403 has a typical saturation voltage loss of only 0.2V when the collector current is 300mA and the base current is 30mA.

Every LED might have a different forward voltage. The forward voltage is a ranges of voltages. If there is enough voltage across each current-limiting resistor then the differences in the LEDs will be swamped.
 

wilykat

Member
ok now I am confused. It had been a while since I tried inventing something using transistors on my own. I usually stick with TTL and CMOS.

Assume I get 2N2222 for 600ma max (seen some brand that can go 800mA to 1A but 600 is lowest max), and trying to understand the spec (Fairchild PDF), it seems at higher current draw it would be 1v collector to emitter?
 

BrownOut

Banned
What exactly are you looking at on the spec? The VCE curve shows about 300mV at IC=500mA. The curve is steep at that point, but I don't think you can extrapolate more than about 500mV at 600mA?
 

audioguru

Well-Known Member
Most Helpful Member
The Fairchild datasheet is wrong (made by little school kids).
The Collector-Emitter Saturation Voltage is supposed to say that its base current is 1/10th its collector current but it wrongly says that its VCE is 10V.

Its maximum collector-emitter saturation voltage is 1.0V at a collector current of 500mA and a base current of 50mA. The curves show that its typical voltage is 0.3V.
 

BrownOut

Banned
HAHA! I would go by the graph then. It seems "right" from what I can remeber about the part. At least you can coorelate it with other datasheets.
 
Last edited:

audioguru

Well-Known Member
Most Helpful Member
The graph only shows "typical" transistors. You might get ones that are weaker than typical and they might have a voltage drop that is too high for your circuit or they might get too hot.

If you design a circuit with minimum spec's then every passing transistor will work.
 

BrownOut

Banned
Always leave yourself a good margin with respect to power dissapation. By guardbanding for power, you don't have to eat up your very small voltage margin by trying to consider every part you might install. One way to do that is, as I suggested, by distributing the load among more than one part. If you want to run to the edge, then you might want to screen parts, or just make sure you're buying "A" grade parts. You may also want to understand how the transistor's tolerance affects the power disspated. Power may or may not be all that sensitive to the transistor's tolerances. But if you don't want to think about all of that, just leave yourself generous power margin.
 
Last edited:

ronsimpson

Well-Known Member
Most Helpful Member
In the h-bridge circuit only the bottom two transistors will have a small drop across them. The two top transistors will have almost 1 volt of loss because they are in an emitter flower circuit.

I am using 300mA and 1 amp LEDs. It takes a lot less wiring.

It sounds like you are going to parallel 15 LEDs at a time. Don’t parallel LEDs. If all your LEDs were made at the same time on the assembly line it might work. If they came from different batches some will not light and others will be too bright. I know of a project where three LEDs are in parallel and we have to sort the LEDs for on voltage. We sort into 8 different bens, and then use LEDs with the same on-voltage.
 

BrownOut

Banned
In light of our discussion on device tolerances, I tend to agree w/ronsimpson. A good alternative would be to use one drive transistor per LED. Might seem like overkill, but the transistors are small and cheap. Alternatively, you can use an integrated LED driver.
 
Last edited:
Status
Not open for further replies.

EE World Online Articles

Loading
Top