As you may know, I design and build my own controllers; a number have been described here:
7555-based PWM controller: http://www.009society.com/forum/index.php?topic=1119.0 (http://www.009society.com/forum/index.php?topic=1119.0)
digital PWM controller: http://www.009society.com/forum/index.php?topic=1156.0 (http://www.009society.com/forum/index.php?topic=1156.0)
digital PWM with H-bridge output: http://www.009society.com/forum/index.php?topic=2325.0 (http://www.009society.com/forum/index.php?topic=2325.0)
(I seem to have omitted my LM317-based d.c. controller)
One thing that my controllers did not have was some method of detecting an over-current situation (such as a short caused by a wheel set parked across an isolating gap). I am not really concerned abut this with Angst-Lesspork as it rarely happens and when it does it is easily rectified before any damage is done. However, that's not an excuse for not looking for a solution!
I left it on the back-burner as it were, until I came across a circuit that fitted the bill and would allow me to retro-fit it to my existing units (always provided there was space in the box). For completeness, the circuit can be found on the MERG web site at https://www.merg.org.uk/merg_resources/other.php (https://www.merg.org.uk/merg_resources/other.php) and it is called the Basin Street Controller. The author is happy for it to be freely used.
I'll go into how I applied the circuit to my controllers later, but I have been dabbling again and come up with a further modification.
Whilst using the controllers I had modified to indicate overcurrent conditions, I noticed that the little warning LED was not the first indication that a short had occurred. Now, all my controllers are hand-held devices; could I make them vibrate (like a mobile phone does) when the overcurrent circuit switches on?
The answer is a very simple "yes". The devices that make mobiles vibrate are tiny little discs 10mm diameter and 3mm thick. They contain an electric motor with an eccentric weight on the shaft; when power is supplied, the motor spins the weight which causes it to shake. The device comes with a pull-off tab on one side, which reveals an adhesive pad so that it may be stuck to a flat surface such as the inside of the controller. They work from 2 to 5 volts. And they cost around
KPC used a autoreset cutout fuse. This appeared to work very well in detecting short circuits.
These are available in various ratings from 0.4A to 12A in the same basic size.
Brian Meldon
That's an interesting circuit Hugh, thanks for posting it.
I built battery controller for a young grandson a year or two back - the idea was to have it independent of the mains for safety. So I was worried about the effect of a short circuit on the battery life.
I built in a similar protection circuit but it used a dual op amp instead of the two transistors you have in what is I think in effect a flip flop. I used one op amp as a comparator to detect the overcurrent and switch off the drive with a similar arrangement. The second one I connected to the first in what was in effect a monostable, with a period of about a second. The result was that it cut off the drive for about a second then reconnected it, and of course immediately cut it off again if the short was still there. So there was no need for the reset button, it reset itself once the fault had been cleared.
I'm sure I have the circuit somewhere ...
I have been looking at the circuit you used; it must be possible to do the same with that but I can't see how just at present. It probably needs just a capacitor adding, maybe a diode as well. I will keep thinking about it I expect.
Frank
I'm working on your Simple Controller and managed to short out my transformer last night, it wasn't damaged but the plastic case was getting hot (so of course I unplugged it!) I'm having trouble with the insulator kit. So some kind of short circuit indication would be desirable.
Mal.
Perhaps this simple devise will help? This is the same component used in the KPC controllers as I mentioned in my previous post. It is quick acting and auto re-sets it self:
(https://forum.009society.com/proxy.php?request=http%3A%2F%2Fwww.gaugemaster.com%2F_upload%2Fimgs%2Flrg%2F34505%2FGM41.jpg&hash=4ebf1d84df3d5224d45b135ee8238c347decc378)
http://www.gaugemaster.com/item_details.asp?code=GM41
I have purchased these for less money from Maplins in the past, but obviously Maplins are not an option any more, but other electronic component suppliers may well have them for less cost. This gaugemaster item is 1A but they also make them in other values.
Brian Meldon
Hmmm. I think perhaps you did not understand my post.
I was looking at a method of getting the controller to alert me when a short (overcurrent) was detected, as the presence of a LED glowing was not immediately obvious. Using haptic feedback looked to be a method worth investigating. I presented this as a (novel) solution.
I enjoy playing with electronics and this was just an exercise in design and implementation.
Quote from: WALLSHIRTS on April 12, 2018, 00:59
I'm working on your Simple Controller and managed to short out my transformer last night, it wasn't damaged but the plastic case was getting hot (so of course I unplugged it!) I'm having trouble with the insulator kit. So some kind of short circuit indication would be desirable.
Mal.
Do you mean the LM317 one? You can get an internally isolated version of the IC that does not require the isolation kit The package is called TO-220FP (rather than TO-220) The actual part number is LM317P (note the "P") You may have to hunt around fro them (but I see you can get them through amazon)
You will notice also that in the photo there is a sheet of thick paper under the circuit board, keeping it off the metal of the case!
{I have now updated the page on my web site to mention these}
Quote from: Frank on April 11, 2018, 21:37
That's an interesting circuit Hugh, thanks for posting it.
I built battery controller for a young grandson a year or two back - the idea was to have it independent of the mains for safety. So I was worried about the effect of a short circuit on the battery life.
I built in a similar protection circuit but it used a dual op amp instead of the two transistors you have in what is I think in effect a flip flop. I used one op amp as a comparator to detect the overcurrent and switch off the drive with a similar arrangement. The second one I connected to the first in what was in effect a monostable, with a period of about a second. The result was that it cut off the drive for about a second then reconnected it, and of course immediately cut it off again if the short was still there. So there was no need for the reset button, it reset itself once the fault had been cleared.
I'm sure I have the circuit somewhere ...
I have been looking at the circuit you used; it must be possible to do the same with that but I can't see how just at present. It probably needs just a capacitor adding, maybe a diode as well. I will keep thinking about it I expect.
Frank
Frank, if you leave the first header open (so it does not latch), but leave the second header closed, in the event of a short the LM317 is shut down by the 2N3904 transistor until the short is removed.