Stay away from circuits from the likes of Mr. Ahmed. He couldn't even draw a decent cartoon. Too many unnecessary wire jogs and several unnecessary wire crossings.

He could have drawn it more like this:

 

multi-output supply

No, don't feel bad. They're small cheap micro controllers. Their fate was sealed on creation. Most will die as sacrificial chips on the alter of learning. That gentile blue smoke as it passes on lights understanding in the next generation. Some lucky few will live on as successfull projects, standing on the bones of its predecessor.

wipes a tear from his eye!

 

Here is an update for my original post, this circuit will work as long as you don't have all of the rows lit at the same time for long periods of time.

This is just a simplified block diagram...details will be provided on request.

You could use a multi-output supply or just get a 24-28 volt supply and a DC/DC module for the Nano.

Measure those ring terminals and get an appropriate sized barrier strip for easy construction.

View attachment 248470

This one makes sense to me, i think, but I;ve never used a ULN2003, I've looked online but not sure I'm looking at the right thing

 

ULN200x, ULQ200x High-Voltage, High-Current Darlington Transistor Arrays datasheet (Rev. P) (ti.com)

 

ULN200x, ULQ200x High-Voltage, High-Current Darlington Transistor Arrays datasheet (Rev. P) (ti.com)

THank you!

 

There is no reason if you desire to use the original lamps. It will give you the original aircraft look. Each one uses 28V at 40mA from what you indicated that would be about 0.8 Amps total for all on at the same time. To turn any one of them on all you need is something as simple as this:


You can do this for each one you want to control (all 20). If you want to power 4 at a time just put 4 of them in parallel as shown here with change in base drive resistor:

Then all you need is a 28V power supply at 0.04A * 20 = 0.8 Amps

The following power supply is only 18 bucks and is capable of 28Vdc at 25W or 0.892 Amps. Or you may want to go with another with a little more power:

https://www.digikey.com/en/products/detail/mean-well-usa-inc/GST25B28-P1J/7703661

 

There is no reason if you desire to use the original lamps. It will give you the original aircraft look. Each one uses 28V at 40mA from what you indicated that would be about 0.8 Amps total for all on at the same time. To turn any one of them on all you need is something as simple as this:

View attachment 248545
You can do this for each one you want to control (all 20). If you want to power 4 at a time just put 4 of them in parallel as shown here with change in base drive resistor:
View attachment 248546
Then all you need is a 28V power supply at 0.04A * 20 = 0.8 Amps

The following power supply is only 18 bucks and is capable of 28Vdc at 25W or 0.892 Amps. Or you may want to go with another with a little more power:

https://www.digikey.com/en/products/detail/mean-well-usa-inc/GST25B28-P1J/7703661

The solution that I offered has a major flaw. Now that I have though about it. The inrush current for an incandescent bulb is like 10 times the rated current for about 100ms! This could destroy the transistor and decrease the life of the bulb. There is an easy fix for this. Here is a models simulation with a capacitor and resistor added as the load to 'model' the behavior of the bulb. Take a look at the inrush current.


Here is the inrush current. A 5V 1/2 second pulse is applied to Q1:



Here is the fix:



Here is the current with the fix:


The fix simply limits the current in Q1 via Q2. If the current goes above 0.6/R amps (in this case around 55mA) Q2 holds back Q1 for a little while.

I am sorry I did not catch this originally. But it has been a long time since I had to work with incandescent lighting. Attached in the LTSpice of it.

 

The solution that I offered has a major flaw. Now that I have though about it. The inrush current for an incandescent bulb is like 10 times the rated current for about 100ms! This could destroy the transistor and decrease the life of the bulb. There is an easy fix for this. Here is a models simulation with a capacitor and resistor added as the load to 'model' the behavior of the bulb. Take a look at the inrush current.

View attachment 248785
Here is the inrush current. A 5V 1/2 second pulse is applied to Q1:

View attachment 248786

Here is the fix:

View attachment 248787

Here is the current with the fix:
View attachment 248788

The fix simply limits the current in Q1 via Q2. If the current goes above 0.6/R amps (in this case around 55mA) Q2 holds back Q1 for a little while.

I am sorry I did not catch this originally. But it has been a long time since I had to work with incandescent lighting. Attached in the LTSpice of it.

The solution that I offered has a major flaw. Now that I have though about it. The inrush current for an incandescent bulb is like 10 times the rated current for about 100ms! This could destroy the transistor and decrease the life of the bulb. There is an easy fix for this. Here is a models simulation with a capacitor and resistor added as the load to 'model' the behavior of the bulb. Take a look at the inrush current.

Here is the inrush current. A 5V 1/2 second pulse is applied to Q1:

Here is the fix:

Here is the current with the fix:

The fix simply limits the current in Q1 via Q2. If the current goes above 0.6/R amps (in this case around 55mA) Q2 holds back Q1 for a little while.

I am sorry I did not catch this originally. But it has been a long time since I had to work with incandescent lighting. Attached in the LTSpice of it.

Attachments

• IncandesentLightBulbDriver.asc

 

My $.02

#1. There are plenty of LED dimming integrated circuits around. It might be wise using one of them.

#2, Different color LEDs have different forward voltages.

#3, The forward voltage by color is a measure of uniformity.

#4. Don't wire LED;'s in parallel. e.g. Rseries = (Vsupply-(Vled1+Vled2+.....Vledn)/20 mA (typical)
there are always rules that get broken.

#5. Leds are best dimmed using PWM. Pulse width modulation.

#6. Is this a battery powered thing.

#7. Use a driver.

#8. You can get a little fancy and use the ambient light to modify the brightness.

#9. The power supply voltage and whether battery powered means a lot as well as how long it needs to operate is a lot.

 

so... bit of an update
I have this working but wondered if anyone could cast their eye over to see if there is anything I should fix or correct, plus I'm having an issue with the nano code.

I hope what I’m describing is correct, sorry im very much learning on the fly!!!

I have changed the lamps to 6.3 V O.2mA. The nano is going through a ULN2003A transistor, and the lamps are being driven by a 9V battery at the moment, I might double it up to 18V.

This all seems to work correctly, but it's the lamp row Grounds that are connected to the ULN2003A and they all share the same positive, from the battery is this correct? as that’s just the way the old indicator is wired up.

As I said this seems to be working fine, at the moment but I just want to be sure that I don’t burn out the nano and the transitor is doing its job, or i;ve just wired it up wrong but it seems to be working. I'm going to try and draw the circuit so you can maybe see it better.

 

Your wiring looks correct, but I would be concerned about running 6 volt bulbs @ 18 but they don't look overly bright @ 9 so you might be ok.

As long as you have the Nano outputs connected to the driver chip inputs your Nano will be fine.

I see you have left pins 6, 7, 9, 10 and 11 open...this is fine.

That one row seems to be partially lit, is that the issue you mentioned with the Nano code...you need to fix that. Could it be a faulty output?

 

Your wiring looks correct, but I would be concerned about running 6 volt bulbs @ 18 but they don't look overly bright @ 9 so you might be ok.

As long as you have the Nano outputs connected to the driver chip inputs your Nano will be fine.

I see you have left pins 6, 7, 9, 10 and 11 open...this is fine.

That one row seems to be partially lit, is that the issue you mentioned with the Nano code...you need to fix that. Could it be a faulty output?

i;ve made a poor atempt at the circuit as i think it is, yes THe bulbs arent overly bright at all, I just assumed ast there were 4 in a row being lit at once it would need 24v so I was playing it safe at 18, is this not the case?

So weirdly this circuit works, if i use the delay function, but if i try using the exmaple sketch blink without delay, and chaneg the pin to 12 it dosen't work

 

You are showing the bulbs in series, I'm pretty sure they are in parallel going off of the image in post #6 and the fact that aircraft specs would never allow that kind of failure mode...one goes out they all go out.

 

yes you are correct - basically if one bulb is pulled out the others wrk so thats parallel? im guessing, sorry i was having issues with how that circuit should look, ive thrown myself in at the deepend lol I ave been learning to use fritzing too lol - head is blown!

Yes the circuitry and build quality is a thing to behold, so well put together which is why im reluctant to rip it all apart, i'll attach a pic of the back, is it possible to show me how that circuit should look?

 

basically if one bulb is pulled out the others wrk so thats parallel? im guessing, sorry i was having issues with how that circuit should look

They'll be in parallel so there wouldn't be a single point of failure.

ive thrown myself in at the deepend lol I ave been learning to use fritzing too

I'm not a fan of fritzing. I tried using it to show some members how to draw less confusing wiring diagrams and the program kept aborting. You'd be better off learning how to use a better schematic/PCB editor. The pros don't use color coded schematics. Monochrome has been the standard since the beginning.

Yes the circuitry and build quality is a thing to behold

You'd hope that people designing and building things for aircraft would know what they were doing.

 

Yes parallel.

You should be able to make your schematic changes from the example in post 19. Note how the bulbs are wired.

 

Yes parallel.

You should be able to make your schematic changes from the example in post 19. Note how the bulbs are wired.

I've updated the circuit from your exmaple, is this correct?

 

If I were doing this, I would simply replace them with high brightness LEDs and resistors. These would be bright at 5mA, so no need for a transistor.

Bob

 

If I were doing this, I would simply replace them with high brightness LEDs and resistors. These would be bright at 5mA, so no need for a transistor.

Bob

Yes that was an option but i wanted to keep it as original as possible

 

I've updated the circuit from your exmaple, is this correct?

Looks like it. I like how you eliminated the unnecessary wire crossings.

It would be nice if you moved the component designators for Nano and IC1 closer to their symbols.

Does fritzing have a print option? Colored schematics seem a bit amateurish to me.