Due to my openly-admitted laziness, a while back i got myself a trimmer - it's way quicker and easier than a full shave, and i rarely need (or want, for that matter) to be totally clean shaven. But that's enough background babbling, on to the nitty-gritty. This one's mainly "what for increasing the chooch factor", as it were - thank you AvE, you crazy canuck.
This thing must've cost me something around 30 bucks or so (if not less), a couple years ago. No frills, no bling, pretty straight-forward. I'm sure i used it a few times first, but i just couldn't help myself and just "had to" crack it open at some point.
The internals revealed a great deal of simplicity: two NiMH cells in series, a DC motor, and merely a handful of components on the charging side. Now, isn't that a(nother) prime candidate to replace those wonky rechargeables with a (slightly more beefy) lithium cell?
The charger is, on one hand, a lot simpler than i would've expected. That being said, it IS "only" rated for 3.2v @ 1.5A.
On the other hand, it's reasonably well designed and built, i must say. It's your run-of-the-mill flyback converter (which is the most frequent topology you'll see in power supplies up to 100W or so - phone chargers, laptop power bricks etc). But let's not jump the gun just yet, one step at a time.
The charging circuitry inside the trimmer's quite interesting - simple, but good enough, i suppose. A few resistors, two diodes, an LED and a switch.
Weeeeeell, wouldn't you know it? Can't cram an 18650 in there because it's too long. Lucky i'm the gathering type, though. A year or two ago, i had to replace the battery pack in someone's 13" MacBook Pro; the factory one had crapped out (the fancy fuse on the control board had blown), but since the cells were all fine, i decided to keep it. The smallest of the six cells fits the bill (or in this case, the physical space available) juuuuuust right. Score!..
Markings say 7.07Wh, so something in the neighbourhood of 1800mAh or so - well over the (alleged) 1100mAh of the stock NiMH cells. That being said, i'm starting to slightly wonder whether i'll need to knock down the voltage some.
As it stands, i'm looking at roughly 60% more-than-stock, which is a smidge more than a more reasonable 10-20% tolerance band. The motor itself would appear to be rated for 2.4v, from what i can decode from the markings. Online i could only find data on the 7.2v and 24v versions.
Then again, the bench supply says the motor draws about 360mA @ 2.4v, and around 400mA @ 4.2v. Sounds a bit more "angry" too, but no noticeable heating (albeit after only a handful of seconds seconds), or strange smells. And considering this will be under sporadic load, i'm thinking "she'll be 'roight" (read with an Aussie accent for maximum effect). So that's that then.
Next, figuring out the switching & wiring. And then altering the power supply circuitry, in order to obtain 4.5-5v for charging. There should, at least in theory, be plenty of juice available in the thing - 3.2v @ 1.5A is well north of 4W, which would be in the region of the max the TP4056 charging chip can handle anyway. And, well, that turned out to actually be pretty straightforward, as can be seen below.
For mechanical reasons (read: to get the charging board to fit, AND to get the two LEDs on the charger roughly kinda-sorta in the region of the little window where the stock LED used to shine through), i decided to chop off the connector end of the charging board (which is unused). The wires from the power input connector got wired straight to the chip's pins as well. I stuck the board to the factory PCB with a nice big patch of 3M 9087 double-sided tape - strong stuff.
The charger board "resizing", in turn, prompted the need to reorient the charging current setting resistor - no biggie, i just tacked it straight onto the TP4056's pins. I decided to go with a 3.3k resistor, to drop the charging current down to 400mA or so. That's partially because, after i get my hands on some, i'm thinking of sticking a micro-USB connector instead of the stock barrel one. That, plus i doubt i'll ever be in a great rush to charge this thing, and taking it easy on the cells and the charger. and i can always tack on an extra resistor across that, for an increase in current.
All wired up, all buttoned up. Had to snip off a few corners from the top half of the casing (which would've held down the stock NiMH cells), but no other hindrances. Flicked the switch, it turned on just fine. Flicked it a few more times, "to make sure", and nothing. It would appear that the overcurrent sensing of the DW01 chip is triggering. Kinda-sorta makes sense, DC motors do indeed have a considerable inrush current, at start-up. Weird part is, though, that the overcurrent cut-off doesn't seem to want to go away on its own (as it should, from what i can gather from the datasheet).
Not entirely sure it's related, but it turns out that, for whatever reason, they swapped (or mixed up?) the two resistors associated with the DW01, when populating these charger boards. At least in the datasheet, there's a 100 ohm in series with the Vcc input, and a 1k on the current sense input. On these charging boards, those two seem to be reversed. I'll swap'em around, to make the real circuit as in the datasheet, and then i'll tack on a resistor from the current-sense input to ground, to "derate" the apparent current draw. Also, just to be on the safe side, i think i'll tack on a reverse / "freewheeling" diode across the motor, to mitigate whatever inductive kickback might occur when stopping.
Ok, 100 ohm's on the Vcc line, and i stuck in a 680 ohm from CS to GND, in preparation for the current-sensing attenuation. As it stands, that particular feature is bypassed, and sure enough, it powers on and off just as well and normally as you could expect. I added in the 1k from CS to where it's supposed to go (ie. OUT-), and still all is well, as it should be. At long last...
On a side-note, this right here is where i envision the main failure point to be. The lack of "recoverability" is what bothers me most.
But the trimmer's all buttoned back up again, and workin' like a charm. Hell yeah! Now that that's out of the way, moving (back?) on to the power supply / charger.
There's no fuse as such, save for a heatshrink-wrapped 10ohm resistor (says so on the silkscreen) on the input. Also there's no kind of filtering (to prevent switching has finding its way back into the mains), or non-populated spots for that. Four discrete diodes make up the rectifier bridge - no surprise there, this thing's really low power to begin with. What they did do, though, was to use an actual inductor (as opposed to a wire jumper) between the two high-voltage reservoir caps, which i thought was a nice touch. With a 4.7k resistor in parallel with it, no less.
One (other) surprise was that the two Y-class EMI-suppression capacitors (between the primary and secondary) ARE in fact properly safety-rated, with all the requisite markings. You don't see that everyday, especially in bargain-barrel gear.
Despite not having optocoupled feedback from the secondary, it does actually employ a dedicated PWM controller chip, and a high-voltage NPN transistor (BU3150F) that does the actual switching.
You'll have to excuse the hand-drawn schematic, but it was way quicker than messing around creating the proper part in the software and all that malarkey. Not like i managed to find any sort of data on the PWM chip anyway - if you can figure out what the "FBBC" markings on a SOT23-6 (aka. SOT26) package mean, your Google-Fu is way stronger than mine. <bows>
Quite a funky-looking topology though - not 100% your bog-standard current-mode controller flyback. You'd expect a current-sense input, and a dedicated ground pin, but i'm not sure i can see one. The PWM chip is driving the high-voltage NPN directly, and there's some sort of signal picked up from its collector.
The chip would appear to be supplied via what looks like a discrete voltage regulator, made with the MMBT3904 SOT23-packaged NPN transistor. I measured about 7v after the diode (and at the collector), about 3v in the base (coming from pin 5), and 3.3v on the output (and pin 6 of the IC).
I'm starting to be slightly unsure whether i can be bothered fiddling with this thing any further, actually - it might be less of a headache to just transplant the cable onto / into a phone charger or something. Or actually, just to go ahead with my initial plan of sticking in a micro-USB connector.
Either way, the trimmer itself works just great. The blade assembly gets warm a bit sooner than before, but that's just normal - the motor's working faster than before. I considered loosening the screws that hold the static blade on, but that wouldn't change anything, since the cutting blade is held pressed with some pretty beefy springs.
All in all, i'm happy with how this turned out, and at least i found some use for some of that <ahem> "dead" MacBook battery.
No comments:
Post a Comment