I pulled the ATiny13 and put a new ATiny13A on the board. Also put a new voltage regulator at U3. Ended up going with my 4th emitter and finally got light. With the clamp meter around the negative lead I got 2 levels out of the 5 that are supposed to be there, .02A and 3.12A…emitter amps. And from a Panasonic PF. (The Sanyo does the same thing by the way) The STAR firmware is set at 1, 4, 25, 120, 255. Similar to the BLF 17 DD board with no gate resistors for the FET, there are no middle modes. The 1 for moon is not activated (no stars on this driver) so I’m getting the 4 and the 255, nothing on the 25 and 120.
Edit: The clamp meter was around the negative lead from the driver to the emitter.
And when I removed the ATiny13, it appeared that the ground (pin 4) was not making contact…solder had blobbed onto the outer portion of the leg over the mask and had the leg or pin hovering over the actual contact pad which was clean of solder, somehow. That particular pad is quite small, while all the others are longer rectangles this one is a small square. So it would appear that when I heated the pins one at a time with the .5mm soldering iron tip, the solder on that one “jumped” the pad. (beats the hell outta me) You can see the clean pad in the pic, but it appears to be soldered at the end of the pin. This must have been a cold joint and wasn’t making contact as it came up with the mcu and left nothing on the pad, not a mark.
Edit: Ooops! That’s my solder wire, I’m using Kester EP256 solder paste in a syringe. It doesn’t specify any more than that on the syringe, can’t remember which variant it is. (63/37 I believe, with no clean flux)
Thanks for doing all this Dale. A continuity check is a royal tedious pita but will reveal cold joints like that. You just have to make sure the probe is on the pin and not the pad and sometimes probe pressure can create a connection that doesn’t exist with pressure removed. I’ve had emitters that looked fine but had the same problem so a visual inspection won’t reveal everything.
Good work! Thanks for testing with the PF cells. Did you happen to take any readings that showed the increasing output thing again?
Can you flash a STAR firmware build with a lower freq PWM? HarleyQuin posted a little chart with a writeup on what does what. I’m pretty sure that all you’d need to do to drop from 19kHz to 2.4kHz is set your low fuse to 0×65. CPU freq would not change, only the divider.
Actually, upon re-reading Mattaus’s post on the issue and glancing at the datasheet, I doubt that flashing with a lower freq PWM is going to do much. Shouldn’t hurt anything of course. High frequency PWM should just limit how low the PWM’d output actually gets. My bad!
I would like to develop your level of patience. Your attention to detail is helpful and I am learning second hand from your pictures and explanations. Thank you for that.
Dale, I feel that a 16lb sledge while effective, it lacks character. If it must be put down, I propose that as it has received so much attention and care during its life that it’s demise should be equal. As someone so skilled with a camera perhaps a series of high speed photos to document it being rendered asunder by a precise and carefully applied pyrotechnic charge. J)
Hmm, 16lb sledge. While I sympathize with the plight of the crash test dummy, it is their preordained lot in life to suffer for the benefit of others. Not alone in this, others throughout history have paid the ultimate price, sometimes merely for entertainment. Some of you may remember Wile E. Coyote, Super Dave Osborne, and who could forget “Oh No, Mr Bill!” Whomp! 16lb sledge meets playdoh.
However, the cost of the parts that went into this Wile D. Bill driver is not insignificant so I ask you Mr Hands, is it worth it?
Dale, you really ought to get some extra flux, it makes working with these smaller components a lot easier. I use the MG Chemicals 835 Rosin Flux (non-corrosive, non conductive) liberally on most projects working with smaller parts. It makes a sticky mess but it works well and comes off easily with non-chlorinated brake cleaner (wear gloves, spray outside!) A $8 bottle will last for years and years.
Of course you don't need extra flux for everything, that's definitely not true, but when you're working with smaller packages it allows you to drag solder and use other techniques that are a bit faster and give a cleaner result in the end. Lots of flux is key to easily hand soldering surface mount components.
It is interesting that the middle modes don't work. I guess we'll have to figure out what PWM frequency this chip likes best, if you're running the phase correct right now and it doesn't like it I guess you could try the fast PWM (switch 0x21 to 0x23) and try that. If that doesn't work, we could try halving the frequency and going back to phase correct. I guess what we have to figure out is how the buck IC responds to different signals. Just like the FETs (and each of those is different) is different than the 7135s, the buck IC also probably has its own set of preferences.
Thanks for linking that post. I guess what also remains to be seen is how the output from the buck IC correlates with the PWM frequency coming from the attiny13a. I.e., how close would 200hz input match with the output frequency?
