555 and 4017 help required - recovery beacon

[tinytim]

Hi guys,
i had a friend of mine pop over a recovery beacon that sits on top of a vehicle, its 12volt dc powered.
The original circuit is powered by a microcontroller. My knowledge of MCUs is almost nothing, although i am sort of ok on the theory behind them.

I figured it would be easier for me to create a clock signal derived from a 555 and pump the output pwm into the clock input of the 4017, the idea is that each pin out of the 4017 can trigger a connected led board. The system works fine as a standard chaser but i am having some issues trying to power something a little heftier. My idea was to use a darlington (BD121) but i just cant get the damn thing to strobe.

i'll post some pictures up very shortly after i tidy the bench up

The leds i want to power are the original ones in the beacon box. I understand the idea of a darlington is to connect the collectors together and use the emitter of the first to drive the base of the second thus allowing full saturation.

top photo - original board with eprom and mcu
2nd pic - 12 of these boards installed in the box
3rd pic - 555 pwm controller for clock into 4017
4th pic - 4017 works and walks through 0 to 9 as expected

if i take a pinout of the 4017 i have to use an NPN to increase the current available to the emitter if wanting to power something beefier.
Looking at the led boards, maybe i have overlooked the 44H11 and the role it plays. maybe its negatively switched with the original mcu sinking rather than sourcing...  So, what next,  some ideas guys please.  regards, Tim

[DJohn]

You've only posted pictures of the bits that work.  How do you have the BD121 connected?  Could you draw a schematic of the whole thing?  How much current do the LEDs need?

I assume you're doing this because there's something wrong with the original.  Are you sure the fault isn't in the LEDs?

4017 through a resistor to the base of an NPN darlington, emitter to ground, collector through a resistor and LED to the positive supply should work.

[tinytim]

ok, will tag some more pictures in a bit. Not sure on how much current each led board requires but there are 14 (not 12) of the darn things.  The original board has an ATMEL 905 24c02b with a CF745 (a pic) powering 8 x trans that i cannot make out the id on. looks like D16N01 HO44 C3KU but it is really feint laser markings.  Will try your suggestion in a moment with some measurements and will swap the led boards around to see whats common between them.

Was not keen on trying to get the mcu board up and running but i keep bumping into projects i want to experiment with and mcu's are everywhere so it makes sense for me to expand my skillset somewhat.

back in a while. Cheers

[tinytim]

right then,  didnt have any darlingtons left, if i do i cant find them, i tried using 2 npn bipolars as a darlington pair and i did have some success, i connected the scope up and you can see the drop to gnd when the base is pulsed high.  I cant seem to get enough current through to make the damn thing trigger (a very weak illumination is present though)  and i have tried a few other led units too.  The arrangement works fine with standard leds connected with a load resistor so i think i'll go for a drive and pick up an assortment of darlingtons / npns etc

tim

[6502nop]

A long time ago...

I had a similar problem with a Larson scanner I designed (my "KITT kit"), using the 555 and an LM3914 (I think the schems are posted in another thread). Anyway, everything was working fine, until I hooked up the actual lights. I planned on using T194 marker lamps driven by some TIP120 Darlingtons. To make things easier on the wiring, I figured I'd just switch in the hot to each bulb, and tie all the grounds to a chassis point nearby. Big mistake. There's a reason all the hot switching in a car is done with relays. Anyhoo, I went through quite a few tests, nothing worked. It wasn't until the ballast of my so-so worklight overhead went flaky, and went out, leaving me in the dark. That's when I noticed that my lights were lighting up - just not bright enough to notice unless you're in a Pennsylvania coal mine at midnight.

That's when I remembered the difference between "source" and "sink". Sure enough, swapping the load from switched hot to switched ground did the trick. Look at the attachment to see what this means. Print it out. Make a poster of it. Throw it on a t-shirt and wow all the ladies with it. Most importantly, change your circuit accordingly, and see what happens. Now, then...

* Make sure the 555 is also a CMOS type (TLC555 or 7555) to match the full 3-18V supply range of the 4017. If the car's regulator should fail, the Vcc can see up to 16-17V, and the standard 555 is only rated for 15. It also uses less power, and has a higher frequency capability.

* When calculating the bias resistors (4017-Darlington base), use the 15V level for the calculations. That'll give you the necessary "fudge factor".

* To calculate, find the load current (measure current draw), and divide by the Hfe (aka "Beta") to get the base current required for the load current. Make sure the transistor used can handle the current at that voltage. For example, we'll assume that each module pulls an amp at 12V, and we're using a TIP120:

Vce=60V, Vbe=5V, Ic=5A, Ib=120mA, Hfe=2500

Okay, the voltages and current are more than we're using, so this'll work. Now, to figure out the Rbias, we'll take 1A (1000mA), and divide by the Hfe (2500), to wind up with 0.4mA (400uA). For the 4017, that ain't squat. It also means we can use a fairly high value resistor to the base of the TIP120. Using 15V as our guide, let's see what a 1K would do:

15V/1000 = 0.015A (15mA).

That's about 37 times more base current than we really need, but still far less than the max current we can send it (120mA). We also should consider the limits of the 4017 output, and limit ourselves to about 4mA load. So, we can redo that a bit:

15V/.004A=3750R  (and to double-check:)
.004A*2500Hfe=10A  (plenty of base current!)

I'd just use a 3K9 or 3K3 and call it done. Either will have enough current to turn on the TIP120 without over taxing the 4017.

* Of course, DJohn does raise a few valid points: Is there a trigger pin on those modules that can be driven by the 4017 directly? That is, if each module is ready to go, just drive the /Enable pin low (or high), and skip the Darlingtons.

nop

[tinytim]

6502nop

some good points there. I have it all on the bench today. I have managed to source various npn darlingtons / power switching transistors so hopefully i can get this damn thing sorted out once and for all today. Thanks for your help. I will post back my findings later.

[tinytim]

update:  Stuffed a BD33B darlington in there to switch the power though the load then to ground. Works a treat.
Now all i have to do is repeat a few times  and also order some of those damn 44H11s - some of the led boards have blown over time so again, i will post up some pics. Actually, thinking about it, i'll do a video on it and share what i have learned.

On a different note, i picked up a velleman K8048 pic programmer and experiment board to cut my teeth on, i will do that as a different topic at a later date.

Big thanks to everyone who helped, sometimes a problem needs fresh eyes to glance over it.
Regards,
Tim

[tinytim]

update on things.

[Kdog44]

Depending on how much current you need; you could use a LM293D (H bridge) . I believe it can source and sink 600mA plus has fly back diodes. It works on 5V logic and can supply up to 36V to your output.
 
http://www.ti.com/lit/ds/symlink/l293d.pdf

[tinytim]

thanks kdog, those look pretty good. Will bear it in mind when working the new board.