AngusMcDoon
Well-Known Member
For those following this project, and maybe wondering what's happening, here's an update. Here's the original thread...
http://www.ybw.com/forums/showthread.php?381831
Last month I requested for 2 testers to have a go with the first 2 complete prototype systems to be made, and 2 complete systems were sent out. Here's the feedback from the testers...
1) There should be a confirm step when cancelling an alarm as it's too easy to choose cancel when meaning to choose silence. Choosing cancel by mistake means that the MOB position is lost.
2) After an alarm has been cancelled no further alarms will be raised for 90 seconds, a design feature. It is not indicated on the display when this 90 second period is in effect so it looks like the system is not working.
3) The radio range between the victim unit and the central unit is sometimes insufficient leading to false alarms, for example if someone goes behind a closed door on a boat.
4) There should be a power switch, or at least an options for one.
5) The board and battery holder in the victim unit come loose and start rattling around in their case.
These are all valid points and I will attend to them all. Here's what's to be done...
1) Easy software change to make. Now there's a confirm cancel alarm on the display with a YES or NO button press. Pressing cancel then NO silences the alarm (as that was probably what you intended) but doesn't cancel it.
2) Another easy software change. When in this 90 second period it is shown on the display along with a count down to re-activation.
3) This is the biggest problem, and requires a new PCB. I've described it with pictures at the end of this list.
4) A power switch is easy to add but there was no provision on the PCB to wire it in. However, because of 3) requiring a new PCB I'll add a couple of pads just after the power connector. If a switch is required it can be taken off from these pads. If not required, then a solder link across the gap between the pads will be used instead.
5) The items in the victim unit were attached using sticky foam pads. I don't want to have to drill any holes in the victim unit because my drilling is not very good. The sticky pads seem to come unstuck after a few weeks. I have bought a glue gun and will use that instead. First tests seem fine.
The radio distance problem...
For the radio circuitry I use pre-assembled RF modules which have all the radio circuitry and an aerial on the board, like this...
This is fine in the victim unit which transmits, but the aerial is a bit puny to use in the central unit that receives. The alternative is in the central unit to use a receiver with an external aerial and a signal amplifier like this...
There are 2 problems with this - firstly it costs about £4 more which will blow my minimum-configuration budget of £20 for the central unit (so maybe I'll offer a small boat version with the original module and a big boat version with the alternative module), but a bigger problem is that the alternative RF module does not fit on the current PCB. Here's the original module in place...
and here's the alternative, which is bigger. It fouls one of the connectors...
So it's yet another PCB design needed - the third, all the previous PCB's end up as inadequate beer coasters and table leg shims...
The RF module in the bottom right is rotated by 90 degrees to fit within the board boundary and not clunk into any connectors. I've added the power switch option while I'm at it, and rotated a transistor outline the right way up as well, although that's not a functional change. I'll send it off to the factory today, but that will be another 3 week delay.
http://www.ybw.com/forums/showthread.php?381831
Last month I requested for 2 testers to have a go with the first 2 complete prototype systems to be made, and 2 complete systems were sent out. Here's the feedback from the testers...
1) There should be a confirm step when cancelling an alarm as it's too easy to choose cancel when meaning to choose silence. Choosing cancel by mistake means that the MOB position is lost.
2) After an alarm has been cancelled no further alarms will be raised for 90 seconds, a design feature. It is not indicated on the display when this 90 second period is in effect so it looks like the system is not working.
3) The radio range between the victim unit and the central unit is sometimes insufficient leading to false alarms, for example if someone goes behind a closed door on a boat.
4) There should be a power switch, or at least an options for one.
5) The board and battery holder in the victim unit come loose and start rattling around in their case.
These are all valid points and I will attend to them all. Here's what's to be done...
1) Easy software change to make. Now there's a confirm cancel alarm on the display with a YES or NO button press. Pressing cancel then NO silences the alarm (as that was probably what you intended) but doesn't cancel it.
2) Another easy software change. When in this 90 second period it is shown on the display along with a count down to re-activation.
3) This is the biggest problem, and requires a new PCB. I've described it with pictures at the end of this list.
4) A power switch is easy to add but there was no provision on the PCB to wire it in. However, because of 3) requiring a new PCB I'll add a couple of pads just after the power connector. If a switch is required it can be taken off from these pads. If not required, then a solder link across the gap between the pads will be used instead.
5) The items in the victim unit were attached using sticky foam pads. I don't want to have to drill any holes in the victim unit because my drilling is not very good. The sticky pads seem to come unstuck after a few weeks. I have bought a glue gun and will use that instead. First tests seem fine.
The radio distance problem...
For the radio circuitry I use pre-assembled RF modules which have all the radio circuitry and an aerial on the board, like this...
This is fine in the victim unit which transmits, but the aerial is a bit puny to use in the central unit that receives. The alternative is in the central unit to use a receiver with an external aerial and a signal amplifier like this...
There are 2 problems with this - firstly it costs about £4 more which will blow my minimum-configuration budget of £20 for the central unit (so maybe I'll offer a small boat version with the original module and a big boat version with the alternative module), but a bigger problem is that the alternative RF module does not fit on the current PCB. Here's the original module in place...
and here's the alternative, which is bigger. It fouls one of the connectors...
So it's yet another PCB design needed - the third, all the previous PCB's end up as inadequate beer coasters and table leg shims...
The RF module in the bottom right is rotated by 90 degrees to fit within the board boundary and not clunk into any connectors. I've added the power switch option while I'm at it, and rotated a transistor outline the right way up as well, although that's not a functional change. I'll send it off to the factory today, but that will be another 3 week delay.
