I just fitted a new 105c 17mm 8 * 7135 3 amp board from James3 on the forum here. It worked fine and cycled through the default modes when out of the host.
Soldered star 4 to the earth track and assembled in to the host
It did not work.
Took it out and it still did not work, so for now have put the old OEM driver board back in . LED is still good.
Noticed a small solder bridge between Star 4 and the inner contact pad of the cap C1.
Remove solder bridge, but it still is dead.
Would shorting that cap to the star kill the driver / MCU chip?
How could I test the board apart from finding another LED to connect it to? no spare LED’s at the moment
.
.
.
.
.
Second question:
Are they constant current out? so if the output is “short circuited” via a multi meter in current mode, will they output the set current and the meter read the correct amperage?
I bought some more advanced constant current driver boards on e-bay a while ago, and that was the CC setting test method. First set the CV output, then ’short the output with an ammeter to set the constant current.
Pics really help with diagnostics. Sounds like you had a dead short to B+. Did the springs get hot and lose their temper? Check the resistance across the input diode, D1. It might have fried. Alternately, if it works when you bypass the diode then that’s probably it.
I suggested a bad input diode because a short from C1 to gnd would dump the short circuit current through the diode bypassing the mcu. The 7135’s don’t turn on because there’s no voltage to the 7135 Vdd pins to turn them on. Fried diode = open circuit and no power to the mcu. Check voltage from pin 8 to gnd, it should equal Vb minus a small drop across that diode. If you still have voltage at pin 8 then mcu may be toast.
Found a diode that works and all good….but before I put it back i the pill…it looks like the pill, although the correct diameter, the shoulder interferes with the tops of the 7135’s and touches the outer tab . The 7135’s are too close to the edge of the board really.
Pill is just a hollow tube…the LED star is not connected to a heat sink, it is just sitting on the edge of the star.
What is the diode there for ? Is it merely reverse polarity protection.
That diode produces a voltage drop of approx 0.7 volts right? So does this not impact on run time?
The MCU will be detecting a cell voltage less that diode drop. ?
What am I missing? Could I run without that diode? Need to know the power requirement of the chip I suppose.
Yes you could run it without the diode, no reverse protection and the low voltage protection flashes or stepdown or whatever won’t be right, but yes…it is possible
Oh yes of course, the MCU is programable, so the LVC setting will have been set to take account the drop. I went and grabbed the Atmel ATtiny13 pinout and went to Dr.Jomes page about his firmware and slave board wiring. I got my head around things a bit better now.
Oh yes of course, the MCU is programable, so the LVC setting will have been set to take account the drop.
I went and grabbed the Atmel ATtiny13 pinout and went to Dr.Jomes page about his firmware and slave board wiring. I got my head around things a bit better now.
[/quote]
Look in on JohnnyC’s STAR firmware, you can tweak the level it goes into it’s low voltage mode…you would have to test it out w/o the diode set it from the stock 130 to say 120 or lower or something then compile but it is adjustable
I’ve been looking into firmware flashing and think I’ll pass.
On the e-bike controllers the job is very easy, TX/Rx pins on the board, nice GUI interface, adjust settings and hit the flash button, but that is only changing certain settings. Looks like changing making firmware on these is a little more complex. More than I need to get in to.
I finally repaired a farm tractor exhaust manifold the other day, bought the parts 18 years ago. Got a Harley in bits 4 years, and two e-bike builds on the go, plus Fordson Dexta restoration. I took 3 weeks holiday, now over, and all I have done is start a new project… flashlights… don’t need to get deeper into this !
Grrrr. What is it with this damn forum software that it had to hijack the first line of each post, use it as a subject line , and in doing so, screw up and quotes.