06-05-2015, 07:13 PM
Hi David,
Afraid to say that I've spotted a mistake with the PCB. The output from the bridge rectifier should feed the relay coil and the 330R resistor, plus the 220uF smoothing capacitor. Currently, it doesn't...
Your schematic looks correct at a glance, so it's a question of making the PCB match the schematic.
Looking at the PCB as it stands, it might be a little bit tricky to route the +UNREG (as I hereby christen it) to the relay coil without a jumper. So - a top tip here - do just that; use a jumper. Why? Because, as I mentioned earlier, depending on the exact details of the transformer, and hence the exact value of +UNREG, you might find that a resistor in series with the relay coil is beneficial. And it's a lot easier to replace a jumper link with a resistor than it is cutting a track. Does that make sense? That sort of "forward thinking" is a valuable skill when doing this sort of thing.
Irrespective of the transformer, I would still retain the Zener diode and series resistor. Why? Because without building and debugging a prototype, I'm not 100% sure about how the 555 responds to a ripply supply in this application (I alluded to this in post #29). And even if there is no problem, the Zener provides protection to the 555. Mains surges happen
Additionally, I'd be a little wary of the trigger input (as mentioned by Peter in post #8). I think I'd provide the pads and footprint for a small capacitor between the switch contacts - perhaps 100n or similar? The wiring to the switch will act like an antenna, and might pick up all sorts of nonsense. A useful "dirty" test for this sort of thing is to hold a mobile phone next to the wiring while it is in a call.
I'd also be tempted to replace the link (into pin 2) with a resistor, as that will limit the current that can flow in the event of an electro-static discharge. Especially if you are using the CMOS version of the timer. Of course, the value of that resistor (at least 1k) might have an impact on the pull-up...
Isn't it amazing just how complex things become when you want them to work in the real world? That's why my 2-transistor circuit appeals to me - despite the apparent simplicity, it's actually pretty well "sorted". Mostly because you know how it works. Who knows what's actually inside the 555 - that is, all versions, from all vendors? Been there, got the t-shirt
All the best,
Mark
Afraid to say that I've spotted a mistake with the PCB. The output from the bridge rectifier should feed the relay coil and the 330R resistor, plus the 220uF smoothing capacitor. Currently, it doesn't...
Your schematic looks correct at a glance, so it's a question of making the PCB match the schematic.
Looking at the PCB as it stands, it might be a little bit tricky to route the +UNREG (as I hereby christen it) to the relay coil without a jumper. So - a top tip here - do just that; use a jumper. Why? Because, as I mentioned earlier, depending on the exact details of the transformer, and hence the exact value of +UNREG, you might find that a resistor in series with the relay coil is beneficial. And it's a lot easier to replace a jumper link with a resistor than it is cutting a track. Does that make sense? That sort of "forward thinking" is a valuable skill when doing this sort of thing.
Irrespective of the transformer, I would still retain the Zener diode and series resistor. Why? Because without building and debugging a prototype, I'm not 100% sure about how the 555 responds to a ripply supply in this application (I alluded to this in post #29). And even if there is no problem, the Zener provides protection to the 555. Mains surges happen
Additionally, I'd be a little wary of the trigger input (as mentioned by Peter in post #8). I think I'd provide the pads and footprint for a small capacitor between the switch contacts - perhaps 100n or similar? The wiring to the switch will act like an antenna, and might pick up all sorts of nonsense. A useful "dirty" test for this sort of thing is to hold a mobile phone next to the wiring while it is in a call.
I'd also be tempted to replace the link (into pin 2) with a resistor, as that will limit the current that can flow in the event of an electro-static discharge. Especially if you are using the CMOS version of the timer. Of course, the value of that resistor (at least 1k) might have an impact on the pull-up...
Isn't it amazing just how complex things become when you want them to work in the real world? That's why my 2-transistor circuit appeals to me - despite the apparent simplicity, it's actually pretty well "sorted". Mostly because you know how it works. Who knows what's actually inside the 555 - that is, all versions, from all vendors? Been there, got the t-shirt
All the best,
Mark
