Dismantling the PWM Controller

[martink]

On a whim, I decided to have a look at the innards of the standard PWM controller. This was partly triggered by my own observations of what seemed like a few strange design features and partly by Ozrail's recent issues with the automation features.

Anyway, after determining that a moderate amount of brute force wasn't the answer, some additional fiddling showed that the four feet can be prised out to reveal four screws. These allow the box to be disassembled and the battery clip and baseplate removed. Removing two more screws freed up the PCB.

What I found was a basic unit that I assume comes from the original type of controller, with a really simple PWM circuit based on a 555 timer circuit grafted on with some fairly crude rework. That seems to explain a couple of things that had puzzled me - the fact that PWM output doesn't quite go down to 0% (there is always a faint buzz), and that the autostop/autoreverse feature doesn't include acceleration and deceleration.

There is also no voltage regulator on the external power supply circuit, so if you use this feature then make sure that you use a proper regulated 5V DC plugpack.

[henningz]

so much money for so little content

[martink]

Actually the price isn't unreasonable. There are $30-$40 worth of bits there (as bought from your local electronics store), and design/manufacturing/sales/profit usually puts a x3 or x4 multiplier on top of that, so the numbers do add up. If you compare this with basic low-end controllers for the larger scales, it looks even better - how many of those give you simple automation features?

[Reith01]

Yes, reasonable for commercial.

I made up a basic controller for about £12 – the stripboard, direction switch, pot and enclosure were most of that. It still needs a protection circuit but that’s small money. The official controller has the reverse sensing facility – that would mean extra circuitry in my outfit. I suppose all in all with a dedicated voltage regulator it could be done for £16. Then as you say you have to add on for marketing and inevitable hike for profits as I doubt they produce them on a vast scale.

There’s some advantage with d.i.y but the extras possible in the controller itself are minimal without some elaborate circuitry. Reversing: I haven’t tried to work out how, from your photos but I’m guessing the auto-reverse is an H arrangement looking at the several power transistors. It isn’t something I want on the current ‘learner’ layout but I can see its value in end-to-end.

So if someone wants to be up and running straight away and hasn’t the tools to build and test their own, the commercial one isn’t bad compared with some on the market.