(11-08-2019, 01:11 AM)Csory_imp Wrote: [ -> ] (11-07-2019, 11:35 PM)Lord Popgun_imp Wrote: [ -> ]Basically a spark gap transmitter.
...Marconi: "Hold my beer..." 
...Marconi: "Hold my beer..."
Tesla: "Marconi, bring me beer!"
Csory is there a better point to sense the voltdrop ( like close to the "relay" or in the headlight unit ? ) and can you not try an external in-line filter close to that supply point to mittigate noise?
Or grab the supply from both front indicators ( orange and blue ) with two diodes, that should give a good rise and sink of several volts and still enough for the 5 or 3.3v on board regulator supply since they are on low all the time.
(11-08-2019, 07:22 AM)max_imp Wrote: [ -> ]Csory is there a better point to sense the voltdrop ( like close to the "relay" or in the headlight unit ? ) and can you not try an external in-line filter close to that supply point to mittigate noise?
Or grab the supply from both front indicators ( orange and blue ) with two diodes, that should give a good rise and sink of several volts and still enough for the 5 or 3.3v on board regulator supply since they are on low all the time.
Hall-effect sensors behaviour will likely still be erratic.
Some carefully provisioned shielding may improve hall-effect behaviour, but now already talking about a new project.
Max: Yes, close to the relay should be better for sensing voltage drop. However, I'm pretty sure that it would be practically undetectable both because of the electromagnetic noise generated by the ignition system, and because of the generator "overpowering" it.
Your second tip would be the right way to go - patch into both the left and right indicator cables, unify the signal with two diodes and use that as a separate input for the BikeGadget. However, that requires a redesign of the circuit board and also a new 5-wire harness. Alternatively, (and if I understand correctly, this is what you meant) I could come up with some sort of a "piggyback" module to be hooked into the indicator circuits and also the +12V wire of the BikeGadget, and somehow "encode" the indicator status into that +12V input.
Eventually, I'll provide one of these solutions - for now, just working on sending out the first units as per the original specs to those who were interested
You know Csory, if you are going to work at the "signal detectors" in a subsequent project, you might as well expand it to intervene and also self-cancel the signals too.
I believe a product already exists and posted here at the Forum this year (a Czech product??). Anyway, just thinking aloud.
If I weren't so stingy, I'd think about installing one. Seems like a great idea to me.
Cormanus: thanks for the link! Not being able to download an installation manual I can only guess how the device copes with the mechanically locked nature of the original turn signal switch. (My guess: it doesn't...) I.e. one will still have to manually cancel the signal by pressing the button, and only after that can be used again for the same direction. Manufacturers should really replace these locking switches with pushbuttons and go with fully electronic "blinker relays"...
(11-11-2019, 10:55 PM)Csory_imp Wrote: [ -> ]Cormanus: thanks for the link! Not being able to download an installation manual I can only guess how the device copes with the mechanically locked nature of the original turn signal switch. (My guess: it doesn't...) I.e. one will still have to manually cancel the signal by pressing the button, and only after that can be used again for the same direction. Manufacturers should really replace these locking switches with pushbuttons and go with fully electronic "blinker relays"...
Looking at a review on YouTube, it looks as though you have to insert two tiny foam buffers within the handlebar switch that prevent the switch latching to either left or right. This allows the unit's electronics to control the "off" point.