LED Sequencer - NPN or PNP transistors ?

Thread Starter

01-0077

Joined Jan 12, 2021
30
Hi,

Total Noob here, hoping somebody may be able to help me out.

I have been trying to design a circuit which sequences through 5 banks of led's then resets and starts the sequence over. I reviewed many such circuits on the web and came up with this circuit. Originally, I was going to use incandescent lamps (hence the 4A darlingtons) but have now decided to use the 5 banks of led's. I need to power 5 prewired banks of leds with one common ground, this is why I used emitter-follower as I couldn't see how to make this work with the load on the collector side of the transistors. The max current draw per bank will be 100mA.

I built this circuit and it worked with incandescent lamps however the transistors got very hot with a current draw of 1.75A. I did some more reading and believe that an emitter-follower circuit cannot fully saturate the transistors and this leads to increased heat dissipation. I also believe I may not need the base resistors in this circuit?

My questions are:

1. Can I use PNP transistors in place of the NPN's with the load between the collector and ground?
2. My understanding is that the 4017 can sink and source current but I am not sure if sinking occurs when the output goes high or low on the 4017 output pins? Will the sequencing still work with PNP's?
3. If I can use PNP transistors how do I calculate an appropriate base resistor?

Thanks in advance for any assistance,
John.
 

Attachments

AlbertHall

Joined Jun 4, 2014
12,345
You do not need the base resistors in your circuit and deleting them may reduce the heat in the transistors.

If you use PNPs you will have all LEDs except one on at each step.

Bob
To avoid that you could use a circuit like below. Here both transistors should be fully saturated.
1613136232044.png
 

Audioguru again

Joined Oct 21, 2019
6,673
The extra transistor is needed since your LEDs have a common cathode. If the LEDs have a common anode (connected to +12V) then the extra transistor is not needed.
You did not explain what is each "100mA bank of LEDs". You cannot use six 2V red LEDs in series without a proper series current-limiting resistor replacing one or two of the LEDs.
 

AnalogKid

Joined Aug 1, 2013
10,986
What is the current through each LED string? This affects the choice of PNP pull-up transistor (T2), and its base current affects the choice of the NPN driver transistor (T1). For firm saturation (and lowest heat dissipation), the base current should be no less than 1/20th of the collector current. The old rule-of-thumb was 1/10th, but that was back in the 50's when power transistors sucked.

Another design consideration is the driving circuit, and how much current it can supply into the T1 base. Working forward from this, it might be the case that the T1 collector current is not enough to drive T2 into saturation at the required load current. In this case, either T1 or T2 can be changed to a Darlington type, or T2 can be changed to a p-channel power MOSFET. If the load current is over 1 A, a MOSFET will have lower power dissipation than a bipolar type.

ak
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
Thanks everybody for the advice.

The led banks are off the shelf units and will most likely draw closer to 30-40mA each, I was allowing some headroom when I stated 100mA. Each led bank has a current limiting resistor built in.

I had considered an NPN switching a relay but would like to keep the component count as low as possible. Any other suggestions even if it means scrapping what I have already done? Be gentle, my knowledge is rudimentary at best .

Thanks again,
John.
 

Audioguru again

Joined Oct 21, 2019
6,673
What is the current through each LED string? For firm saturation (and lowest heat dissipation), the base current should be no less than 1/20th of the collector current. The old rule-of-thumb was 1/10th, but that was back in the 50's when power transistors sucked.
He said 100mA max per bank of LEDs. Ordinary little transistors will be fine, not power transistors.
 

Audioguru again

Joined Oct 21, 2019
6,673
For an output of only 30mA to 40mA, little NPN emitter-followers without base resistors will be fine if the output of about 10V will allow the LEDs to be bright enough.
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
That's great news AGA, thanks.

My concerns arose when I tested the circuit using incandecent lamps drawing 1.75A this caused the heat issue and got me thinking that the emitter-follower wasn't viable as a switch, even for the leds.

When I tested the circuit with the leds the transistors didn't even get warm and the leds where bright enough for my purposes. I wasn't sure if the longevity of the transistors would be compromised when in emitter-follower configuration (not saturated) and this is what prompted my initial post. Do you see any issues in this regard? Also, how far would you estimate before the current draw would be an issue for the circuit as is (sans base resistors)?
 
Last edited:

Audioguru again

Joined Oct 21, 2019
6,673
Longevity of emitter-followers?
An ordinary 2N3904 or BC547 transistor has a max allowed heating of 625mW at room temperature but maybe 500mW in a desert in summer.
In your circuit it will have a collector current of 40mA max and a voltage across it of about 1.4V so it heats with only 56mW which is nothing.

The CD4017 output voltage drops as its output current increases which causes it and the transistors to heat up. The TI datasheet of the CD4017 has output voltage/current curves for you to calculate the transistor heating.
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
Thanks Again, I was referring to the life of the transistors but it sounds like it won't be an issue. If the collector current was increased at some point would it be advisable to use a darlington NPN with a higher gain to reduce output current drawn from the 4017 thus reducing heat for 4017?
 

AnalogKid

Joined Aug 1, 2013
10,986
For low parts count and complexity consider replacing the five NPN driver transistors (and their base resistors) with a CMOS hex inverter. 14 pins vs 25. Then, with small p-channel MOSFET output transistors, you eliminate another five resistors. Dun git mor simplr'n dat.

ak
 

Audioguru again

Joined Oct 21, 2019
6,673
The CD4017 and almost all other Cmos CD4xxx ICs were designed with tiny low power transistors with a low maximum allowed amount of heating listed in their datasheet. A darlington NPN emitter-follower would reduce heating in the CD4017 a lot and the darlington will be heating instead.
EDIT: I was going to say the CD4049 buffer has a higher output current but here you want the LEDs to have their anodes driven high but the CD4049 has a high current output logic low.
 

AnalogKid

Joined Aug 1, 2013
10,986
Thru hole or SMT transistors?

AND - How locked-in are you to the common-cathode LEDs? Things would be a whole bunch easier if they were common-anode and you were switching the cathodes to GND. Just sayin ...

ak
 
Last edited:

Bernard

Joined Aug 7, 2008
5,784
Using a single NPN, 12N3904 etc., as an emitter follower pull-up resistors could be added to 4017 used outputs, say 4k7, & add a Si diode in series with LED supply to effectively increase base V
giving a harder turn on.
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
Thanks everybody, I really appreciate your assistance and advice.

AK, yes I am stuck with common cathode leds and I am using thru hole at present but hoping to progress to SMT when I get a PCB made.

Per post #8 if the emitter-follower is good for 30-40mA then I am inclined to keep things simple, remove the base resistors and leave well enough alone.

Having said that AK suggested a hex inverter and AGA mentioned a CD4049, would it be possible to use CD4049 to invert the 4017 outputs to switch either PNP or p-channel mosfet thus allowing for my common cathode led setup? If so, anybody care to draw up what this would look like?
 

Audioguru again

Joined Oct 21, 2019
6,673
A CD4069 hex inverter or a CD4049 hex inverting buffer can be used to invert the CD4017 outputs to drive PNP transistors.
Power the hex ICs with the 11.4V and see their inputs and output pins on their datasheets.
 
Last edited:

Thread Starter

01-0077

Joined Jan 12, 2021
30
I just re-read post#12 by AK along with the advice from AGA. I had a look at the datasheet for a CD4069 and to be honest a lot of it was over my head. I am pretty sure the CD4017 outputs are within acceptable input range for the CD4069. Also, I would have to tie the unused input of the CD4069 to either Vcc or ground.

What I couldn't work out was the effect the hex inverter would have on the output voltage & current from the CD4017 and the outcome this would have on the ability to switch the p-channel MOSFET? Also, would the p-channel MOSFETS require gate resistors?
 
Top