Hallo Mahmood,

driving 12V/180mA lamps is not easy! If a lamp is switched-on, at first moment impedance of lamp can be about ten times smaller than rated, means up to 2A can flow at first moment. This heavy inrush current must also be taken into consideration when planning to introduce a suited lamp test routine...

Due to heavy inrush current this standard Darlington driver ULN2003 cannot be used. Even not, if two outputs are connected together.
Another disadvantage of Darlington driver is rather high voltage drop, which creates rather high heat dissipation with 180mA. It's better to use a single transistor solution with either bipolar transistors or power FETs.

Another problem is shortcircuit protection: Fuses are normaly too slow and need transistors with lot of headroom in terms of current drive level. So, too expensive or large (or both!) transistors must be choosen. Of course, some current limiting resistor can be used, but to keep an inrush current of ten times the rated current to a save level means to use an idiotic high resistance, which makes the lamp not shining bright any more.

Inrush current is also the major cause of failing of a lamp. Lamp is stressed enormeously by this inrush current. If the working live of a lamp shall be maximized, inrush currents should be suppressed and applied voltage should be choosen a bit smaller than rated voltage.

If you want a circuit, that only uses cheap standard parts, you could experiment with this one:



Each lamp is driven by a constant current source arround BD135: If the current through 3R0 resistor reachs about I = 0.6V / 3Ohm = 0.2A, then the current cannot rise furtherly. When the lamp is switched-on, current through it is limited to about 0.2A and up to about 2.0W will drop across the transistor, but only for a very short time. It's wise to mount the BD135 on a piece of metal, to keep the junction adequately cool.

In order to reduce the voltage drop at transistor with rated current (180mA), the threshold level of current limiting should be choosen in such a way, that the current limiting is NOT invoked with rated current, but only with higher levels. So, threshold level should be designed to be about 10...20% above rated current of lamp.

BC807 is needed to present enough base current for the BD135, about 1/20 of collector current. BC807 is a SMD transistor. Also both 4k7 resistors, the 330R resistor and the both LL4148 diodes are SMD versions There are 'digital transistors' available, containing the pnp transistor and both 4k7 resistors, for instance BCR166 or DTA143EKA. They are really cheap, about 0.05Euro. BD135 is also cheap, when having a good supplier...

I would use 8 x 74HC595 for each floor for the series to parallel conversion. I would also use a +12V bus for all lamps in a floor, so that only one wire for each lamp must be routed from driver board to the building model.

On the right side of schematic there's a row of eight 74HCT597. This row is used for a self diagnosis and short circuit protection: As 2.3W can drop across a BD135, if collector is short circuited to +12V, transistor will heat up enormeously. Well, BD135 can withstand 12.5W with infinite heatsink and 1.25W without heatsink, so 2.3W will cause relevant heat, even when having BD135 mounted on a piece of metal.
74HC597 helps now to detect such a short circuit condition, by 'measuring' voltage between collector and ground. If a HCT type is used, gate will interpret a voltage >= 2V as logical high, and a voltage <= 0.8V as logical low. If everything is fine, about 0.1V will drop across collector emitter path and about 0.18A x 3Ohm = 0.54V will drop across 3R0 resistor, makes 0.64V. But when something is wrong, a much higher voltage will appear at collector. At limited current of 0.2A, which causes a voltage drop of 0.6V across 3R0 resistor, collector emitter voltage of 2.0V - 0.6V = 1.4V corresponds to a heat dissipation of only 0.2A x 1.4V = 0.28W of BD135, which is well within the capabilitiy of this transistor.

Or by other words: Whenever voltage at collector rises above 2.0V, then this corresponds to a failure condition, which results in a possible overrating of BD135, a situation that should cause a reaction by your 8052...
Of course, you can also use HC type of '597. Then, logical high is detected at some higher voltage, namely 3.5V. HC type would also allow a bit higher low voltage, namely 1.0V.

You will find a 10k resistor at each input of 74HCT597. This one is needed to limit input current, as 74HCT597 is supplied by +5V! Internal clamp diodes prevent damage.
For the 10k resistors you can take resistor arrays, of course.

How is the 74HCT597 controlled?
First, you output the data to 74HC595 (inverse logic!). After all 60 bits are shifted, outputs are clocked and actualized. Then, lamps should be given some milliseconds (or more, I don't know...) for turning-on. Then, inputs at 74HCT597 should simultaneously be latched and shifted to your 8052. Now, your micro should compare sent and recieved data: Recieved logical high level can mean both, turned-off lamp and failure with turned-on lamp. So, only by comparing of sent and recieved data the short circuit condition can be detected!

If short circuit condition was detected, then you can put the outputs of 74HC595 into tristate mode, resulting in a fast switch-off of all BD135s of this row, to prevent any damage.

All the BD135 should be mounted on a cooling plate. Althoug only about 0.1V drops between collector and emitter with 180mA, 0.1V x 180mA = 18mW must be multiplied by 3000, which gives about 54W!

A final word: Parts arround BD135 are only calculated, not tested! I could imagine, that 3R0 might be decreased a bit. Please test it yourself. Also, if you test longer lasting short circuits in combination with BD135, don't forget to apply adequate cooling!

Good luck,
Kai