I have built many of them, but usually all with the same firmware and driver designs. Different drivers behave differently with the tailcap draw, and I’ve never owned/used a guppy driver.
Usually though, if the driver is acting strangely, your balance of bleeder value / tail draw needs to be adjusted. On an OTC-based driver, the higher the tail draw, the lower value bleeder you need.
Since wight is back, maybe he can shed more light on the technical aspects of what is going on with the driver. I really just have my observations.
Thanks again guys. I’m not home now to check anything but I did remove the bleeder resistor. It’s gone, and the tailcap still illuminates. Plus like I said I can put the tailcap on another light that’s never been touched and its driver responds the same way.
So bottom line - its not the bleeder resistor ’cause there isn’t one.
I could put larger resistors in the tailcap and see if that makes a difference. Like I said they’re 3.9kohm right now, but there’s 3 of them because of the 3 discrete channels in the tailcap board.
I’ll also try penciling that cap this evening and see if that makes a difference, but it is the same cap I tried stacking a bleeder on first and caused me to lose mode switching altogether, so I think I don’t want too much current flowing around that cap.
That’s the problem. You usually need the bleeder to make the driver function correctly, not to make the tail light up. Everything you are telling me says you NEED a bleeder.
Higher resistor values in the tail will also help.
Oh, I guess I don’t understand all this as well as I thought :). That’s not surprising.
I thought Matt said his MtnElec Qlite drivers never needed a bleeder resistor. Regardless I can add one back. I did have one at first but removed it after I noted this memory issue, so I can already say that it’ll still do this with a 560 ohm bleeder resistor in there.
I guess it’s just a balance? Gotta play with resistor values until it acts right?
I think what they’re saying is that you need the bleeder resistor, not in the place you had it, but directly to ground, in order to get rid of the ‘next mode’ memory you’re seeing.
I don’t really know it all that well either, just what I and others have experienced. Typically drivers that use an OTC for mode changes need a bleeder resistor between Batt+ and gnd, and other types of drivers may not. The Qlite in stock form does not use an OTC (and may not need a bleeder) , but an OTC is often added to a Qlite to use different firmwares (like guppy, apparently).
It’s definitely a balance. I think your high tail draw was just too much for the 560ohm to “bleed”. So either lower the tail draw (higher value resistors) or use a lower value resistor for the bleeder
Man that looks awesome! I’ll be working with Simon on the possible redesign of the metal switch in the near future but he needs time to get things running smoothly in his new factory first.
If we do redesign it I’m trying to have the internal rubber seal replaced with a translucent material to let the light shine through.
emarkd, can you confirm was the 84ma figure a typo? Was it supposed to be 0.84ma? If so, that’s still high, but it makes more sense.
If it’s actually 84ma, I’m wondering if maybe the Rev5 ring is shorting on the metal switch in that S2+, I know it’s really tight up in there with the little switch button.
Hey David thanks for the input but I did try it directly between positive and ground. My original placement on top of C1 caused me to lose mode switching altogether. Still not sure exactly how that worked but the driver became a single-mode driver (on mode 1, moonlight) with the bleeder resistor stacked on C1. I moved it to a place directly between + and GND and my modes worked again, but then it was next-mode memory. I thought by removing it altogether as Matt suggested I could regain “normal” function but that’s not the case.
Thanks PD, but no that was not a typo. I read 84 mA using my trusty old Fluke 76 meter. Its 20 years old now and hasn’t been calibrated since I bought it, but I think its probably still fairly accurate.
There could be some small short in there somewhere, but I don’t think its against the metal switch. I added waterproofing in the form of a disposable nylon glove just below the metal switch so there shouldn’t be any contact there. I’ll check better this evening though.
ok, 84ma is crazy high, the highest I’ve ever measured was just over 1ma I think. There has to be a small short somewhere I think, I don’t think the resistors you said you used would allow anything close to that current. (and I think the tailcap would be ridiculously bright if it were getting that much current through the LEDs). At 84ma, it’ll totally drain a full 800mah cell in about 10 hours.
You know, now that I’m saying all this I’m starting to doubt myself. My meter has two lead connections for measuring amperage and if I was using the low-amp input, which I should have been, then I actually measured MICROamps. I’m used to dealing with several amps of draw, not microamps, so its very possible that I’m now telling you all a lie. I’ll double-check it tonight but there’s at least a chance that what I read last night was 84 µA without realizing it. :-/
Awesome, so lets assume I’m an idiot (not much of a stretch ) and that my tailcap setup really is pulling 84 µA and is working properly. Then the thought process is that I really need that bleeder resistor on the driver, yes? I’ve already had a 560 in there with no change in behavior. Should I go higher or lower?
Lower, got it. I wish I understood why I need a lower one but honestly I don’t. That’s okay though, I’m learning and I really appreciate you holding my hand through this
I’ll report back later this evening after some experimentation.
I’m no electrical engineer, but the way I think of it is that to make the driver work normally, we have to make it oblivious that the illuminated tailcap exists. So we want all of the power the tailcap needs to bypass the driver and just come from the bleeder. If the driver is still messing up, it must mean that not enough power is coming from the bleeder, and it must be pulling some through the driver, keeping the mcu awake or keeping the OTC charged instead of draining like it should. So we need to lower the resistance to make more go through the bleeder.
I hate to muddy up this thread with a post that doesn’t contribute in any way… but I just gotta say, holy wow have I learned a lot in the last two pages!
Ordering parts Trying to order parts as we speak, I wanna play too! Thank you guys!
Okay, I got home shortly ago and make some quick progress. I dropped in several different sizes of bleeder resistor, just working my way down through the assortment I own. I got to 220 ohms but then my driver started working properly again! Tailcap draw is way up, 944 µA now, and the tailcap brightness is much higher too, so I may crack the tail back open and increase those resistors a bit. If I understand things that shouldn’t affect the driver though, so hopefully I won’t have to touch it again!
That is kind of high. And current going through the driver may be keeping the OTC from draining or could be keeping the SRAM from decaying to its powered-off state. Either of these can break the interface on an offtime-based driver.
I’m not entirely sure how guppydrv (guppy2drv) works, but DrJones’ page says it uses off-time memory. So, it probably needs a bleeder to let power through without keeping the rest of the driver lit up.
It seems you’ve already figured this out though, and are in the process of fine-tuning it. I can at least say that the lower the tailcap power, the less the offtime-disrupting effect. Also, the less-resistive the bleeder, the lower the offtime-disrupting effect. However, less-resistive bleeders also make the normal running modes less efficient, IIRC. So, it should probably be the highest value which still works.
) and that my tailcap setup really is pulling 84 µA and is working properly. Then the thought process is that I really need that bleeder resistor on the driver, yes? I’ve already had a 560 in there with no change in behavior. Should I go higher or lower?