Damnit! No good, Comfy. :( I soldered that P2 bad to the main switch pad and nothing. Checked continuity from the two and all is good. Tried the P3 pad instead...nothing! Switched it back to the P2 and tested continuity. Good. P3 no continuity. Good. Checked the board thickness to make sure my soldering on the edge of the ground ring was thick enough to reach the battery tube. Even thicker...check.
Next I popped a 14500 into a cell holder and connected the leads to the LED leads to make sure the polarity is good and the LEDs work. Check. They light up beautifully. Something in the driver is not working properly. Current isn't getting from the contacts on the driver to the LED contact pads. I took a pic in it;s current state hoping you might see something off about it. Any thoughts would be appreciated. Right-click the image and open in new tab to see the full size.
Why doesn't yours have a capacitor? I don't think that would stop it from lighting up at all though. Maybe a bad solder? I'd try holding down the chips while testing.
That Batt+ pad on the LED side is connected to the anode pad on the cell side. The LED+ pads should be as well. I wonder if I need to connect a wire from the Batt+ pad to the L1 pad? I was wondering what that Batt+ pad was there for. There needs to be some decent instructions to go with these things. For complete drivers they are missing a lot of important details that someone who wasn’t a member here wouldn’t begin to know how to figure out.
It’s labeled to have an inductor there. If you don’t use the inductor it needs to be bridged. There are a few threads going up with separate discussions of each board and where they exist I add links to them. Much development and discussion goes on in the Oshpark projects thread where it gets lost in the noise. I have mentioned before the need to have separate discussion threads for each board(or at least each family of boards) and that is being done slowly but these are all new enough that that hasn’t been done for all of them. I’m sure Rich will do what he can to add info on his site as well so thanks for pointing this out.
Now that Werner has found your answer if you would like to retitle this thread it could become the needed discussion thread for the SRK FET board.
Ok…so to be clear, I need to jumper the Batt+ pad to the L1 pad?
[quote=Rufusbduck]
Now that Werner has found your answer if you would like to retitle this thread it could become the needed discussion thread for the SRKFET board.
[/quote]
I'd be happy to. What would be the best title for it?
Oh! Thanks for all the help guys!! I will let you know tonight if it works finally.
Yes, BAT and LED+ must be connected with something, either a jumper or an inductor. Big schottky diode at D2 is not needed if not using an inductor.
with inductor:
without inductor: (ignore wires going off to bottom left, those are for bench testing only, removed before installation)
If Richard is selling these as 'assembled' and complete ready-to-run, they should also really have that jumper installed as well. (For some reason he's against using the inductor/diode parts, I don't understand his reasoning for not using them and he doesn't understand my reasoning for them, I can only tell you I use them because they give a much nicer low end and allow lower sub-1-lumen lows even with reasonable minimum PWM values.)
The big electrolytic cap is no longer needed, that was to patch the issue of whiny noise when using older versions of firmware, no longer needed when using the silent fast-PWM firmwares.
Reduction in max output depends on which inductor you pick, both the DC current rating and the inductance rating. Comparing ~15A-rated 0.58uH inductor to a ~15A-rated 1.2Uh inductor, the 0.56 will have less damping effect on the lowest modes but also the least impact on the straight DC 100% mode. The 1.2uH will give a lower low but with slightly less output on 100% mode. But it's in the range of a few milliohms. With a ~4mOhm Rds(on) FET you can afford to sacrifice a little at the top end and not see a big hit in total output.
I don’t know if you know just how relieved to have this finally working! I was about to completely give up on my modding skills and be so discouraged that I stop again for a while. I can’t WAIT to see some impressive results for once!
On a side note, I tried modding my TR-S700 with a couple of the FETs Richard sells (it's a FET driver to start with) hoping they would allow more current to the emitters but I only gained just under 200 lumens. Still only 3745 OTF with 7 LEDs and 3*26650. Not at all what I was hoping to see. :(
This is the diode needed at D2 if using the inductor - http://www.digikey.com/product-detail/en/MBRS540T3G/MBRS540T3GOSCT-ND/918011 - or, again, something of similar specs. It just needs to be able to shunt the leftovers stored in the inductor when the FET is turned off via PWM in the less-than-100% modes.
Multi-cell buck or boost drivers don't rely on the ultra-low resistance FETs anything like a single cell direct drive driver does. Multi-cell stuff has enough voltage overhead that a high resistance FET is easily compensated for.
Thanks. That's what I needed to know to build one of my SRK's. I had been wondering why the C+ pad was so big. Now it makes sense. It needs to be able to handle different things. I'll place an order on Digikey tonight.
The emitters are wired in parallel. Here is a pic of the emitters on the sink:
Here is the driver. At the time of the pic I'd already upgraded the LED wiring and bridged the sense resistor pads but IIRC they were always empty but don't hold me to that. Since the pic I've replaced the two FETS with two 70N02 FETS. Marginal gains with either mod.
