[Syzygies]

RadioShack 23-349, 18V Charger, is on sale now at stores for $5, listed for $3 but out-of-stock online. At this price, hard to resist buying a bunch to use as trickle charge power supplies, so I bought one for testing. I can't tell if they're discontinued or just on sale.

It has a fairly meaningless NiCD/NiMH switch, and a very helpful 18 hour timer, after which it automatically shuts off. The datasheet claims

18VDC 120mA (NiMH)
18VDC 90mA (NiCD)

In fact, it measures 22V open circuit. The undocumented plug is 5.5x2.1mm, not 5.5x2.5mm.

I first tried it directly connected to my PIMETA 10 AAA battery pack, and at a battery pack voltage of 13.83V I measured 187mA (NiMH) and 170mA (NiCD).

This is too much for a safe AAA trickle charge, even with 1000mAh cells. My guess is that even though RadioShack battery capacities are back in the bronze age, particularly for R/C and phone packs, they have moved forward, and this unit was updated without a documentation change. As they say, "Specifications are subject to change and improvement without notice". After all, this unit is designed for AA cells, and for anyone dumb enough to buy NiMH AA cells at RadioShack, this new rate is about right.

I then tried it plugged into the 9V charging circuit described in my tutorial thread Standard LM317 Charger for 9V NiMh Batteries. It worked like a charm, giving me exactly the same 12mA charge rate as I would get using any other adequate power source. Makes sense if you think how linear regulators in series behave.

Think of this as an adequate, very cheap, small wall wart with a timer for powering your own trickle charge circuits. I don't trust their output anywhere near actual amp circuitry, so I plan to wire "normally closed" power jacks to disconnect the amp whenever a plug is inserted.

Anyone selling rechargeable amps on eBay might consider buying out their area RadioShacks, at this price?

 

[bg4533]

I might have to get a few of these. Any way to make them work with AAAs? Will this thing overcharge or does it actually know when to stop?

 

[Syzygies]

Treat it as an 18V wall wart with an 18 hour timer, and build your own trickle charging circuit. That's two trickle charge circuits in a row; the one designed for a smaller current wins. This works; that's one of the tests I ran before posting.

 

[mono]

I found these instore for $3 ea. Picked up 4 of 'em and may turn one into a linear supply. Unfortunately the casing is cemeted together but a C-Clamp and a bit of patience got it open mostly unscathed, a little scratched up around the rim but easily glued back together.

Inside is a tiny board with 2nd board on a riser. My patience lapsed before I got a chance to trace out the whole circuit but it's based around a MC14541BCP timer IC. The NM/NC switch "appeared" to switch in/out a short tap in the secondary winding. It's a cute little transformer (roughly 35mm sq.)and I might end up using just the transformer out of one to build a regulated supply into a (meant as mostly non-portable) headamp's case. It doesn't have regular posts on it though, the AC plugs are moulded into the bobbin and the secondaries are under the tape but have ~ 30mm insultaed leads coming out. It's servicable but perhaps not esthetically pleasing.

The potential to maybe put in my own circuit board is a real plus though, if it ends up clean enough output...

 

[Syzygies]

Nice! Did you take a picture?

I'd guess that the only role of the timer is to turn off everything after 18 hours, as advertized.

(I own 9 @ $3 now, cleaned out my sources.)

 

[mono]

Heh, I need to work on lighting and macros!

more here

 

[Ozymand]

Has anybody had any luck finding more of these? I tried 4 RadioShacks around me and none of them had it. One of the associates tried using the computer to locate some, and the computer said there were some even if they're well aware that there's none in the location. I'd like to get two to play around with building a LM317 battery charging circuit...

I'll keep trying around here, but if anybody knows a source, I'd be willing to pay for two of them.

 

[Syzygies]

Thanks for the pics, they had me in stitches laughing. Everything I've opened lately was machine-assembled surface mount, I had no idea these components were hand-assembled in Dali-esque poses. Hey, $3 is $3, and they work. Says something slightly scary about our international economy. I'm hoping the workers enjoy some protection from the solder fumes.

The puzzle this invites is to figure out which resistors to swap out, to set a pair of charge rates more appropriate to 750 mAh and 1000 mAh NiMH AAA cells. The NiCD/NiMH switch has a modest effect at present, perhaps a 10% difference.

What I've observed so far is a slanted charge curve that bears only the faintest relationship to the datasheet claims, varies somewhat from unit to unit, and depends on voltage, not a constant current like one would get with a simple LM317 circuit. This makes such a mod problematic for fewer than 12 AAA cells.

 

[helb]

I just bought the last 3 in a RS around the corner. I'm planning to use two so I'll sell one for $3 + S&H. PM me if interested.

 

[mono]

Quote:

Originally Posted by Syzygies Thanks for the pics, they had me in stitches laughing. Everything I've opened lately was machine-assembled surface mount, I had no idea these components were hand-assembled in Dali-esque poses. Hey, $3 is $3, and they work. Says something slightly scary about our international economy. I'm hoping the workers enjoy some protection from the solder fumes. The puzzle this invites is to figure out which resistors to swap out, to set a pair of charge rates more appropriate to 750 mAh and 1000 mAh NiMH AAA cells. The NiCD/NiMH switch has a modest effect at present, perhaps a 10% difference. What I've observed so far is a slanted charge curve that bears only the faintest relationship to the datasheet claims, varies somewhat from unit to unit, and depends on voltage, not a constant current like one would get with a simple LM317 circuit. This makes such a mod problematic for fewer than 12 AAA cells.

I have not taken any significant output readings. ALL of the following might be wrong but I'll throw it out there...

It seems to be a simple VA charger, capable of about 2.2W if the labeled spec is in the ballpark. Due to this, there wouldn't be a resistor swap to decrease current, unless you simply added resistor(s) in series on the output, which doesn't seem all that desirable but given large enough resistors to get suitable mAH cumulative @ 18 hour mark, would work.

It looks like the options for modifying the original circuit itself fall around either adding a linear regulation to the output, whether it be in the casing or out, or varying the frequency of the timer IC, or it's binary count. I've explored these a little bit but the numbers aren't adding up so I'm left scrathing my head on it. At the very least a simple substitution of values should change rates but calculating it out doesn't seem to follow their (Motorola et al's) equation.

MC14541BCP Spec Sheet

A couple things copied from the spec sheet and measurements of parts on the charger PCB:

Where those parts are on the PCB:

Unles I'm misunderstanding the circuit, it seems one could just swap out the 2.2uF cap for a 1uF and end up with an ~ 8 hr. charge interval. Or, they could manipulate the "A" and "B" pins, pulling down to ground for "0" to get 1/8th interval, 2.25 hrs. Or they could swap the frequency resistors, but I don't know the current frequency, what would still be within tolerance. "Maybe" swapping out the 350K resistor for a 110K would yield a ~ 8 hour charge. <yawn> ... gettting late, math now subject to gross errors so I'll end the post.

 

[Syzygies]

Hmmm. I nabbed 5 more of these today at $3, passing a RadioShack on an errand. They're still around. Still haven't taken one apart, busy day.

I did manage to read the MC14541BCP part number off your earlier post, and look up the datasheet. My guess at the time was that the ONLY role of this $0.50 (retail) chip is to shut off the charging circuit after 18 hours. You've found its support circuitry.

In practice, one trickle charges when convenient, just like putting gas in a car. I'm more concerned that I leave for a trip with a full tank, than that I wait for an empty tank before filling. So I'd be using these to "top off" my amps, and simply shortening the interval without dropping the current wouldn't be enough for practical unattended operation.

Moreover, I've observed charging currents from these that depend on the voltage of the battery pack, not the elapsed time. So I believe that the rest of the circuit does discrete AC to DC conversion, and handles the charging. Strange that there's no linear regulator, but it didn't behave like it had one, so I guess I shouldn't be surprised.

Perhaps if one tosses the entire board with the timer on in, the other board provides just a basic wall wart, and this becomes the world's cheapest low-power 18V wall wart?

I still think changing resistor values will drop the output current, and that there will be two switch-selectable choices. If this is true, I'll continue to sweep stores for these, and offer them up to people who can't find them. They'd be a terrible idea for actually powering an amp, but if all one wants to do is charge 12 AAA cells, one of these puppies saves $30 that could be much better spent on components that matter.

For 12 AAA cells, there isn't the overhead to interpose an LM317 circuit. I had thought of putting a resistor in series with the battery pack, to force the charger into partial dropout to lower its final charging current, but modding the wall wart itself would have a certain charm.

If anyone can guess which resistor to change, a nondestructive test would be to use alligator clips to add more resistance in parallel. If this increases current according to one's model, then one's model is probably right, and could be further tested by swapping in a higher value resistor.

 

[mono]

Well I'd thought about replacing the timer circuit board(s) with a linear regulator board... HAH! I'm too much of a glutton for punishment to do that, instead I kept the timer board inside and slapped together an LDO LM1085 regulator, umm, blob.

Target was recharging a Pimeta I have with 6 prismatic cells (~ 8.7V fully charged). Not that it needed an LDO instead of an LM317, but when I started out I'd wanted to try & get peak voltage out of it for multifunction use, then later changed my mind and just wanted a dedicated charger for pseudo-9V NiMH portables, for which a SO LM317 would've done just as well but I already had a couple spare LM1085.

Circuit I used. Note that it is targeted at a narrow voltage range, it was stable from 6V-16V. Outside this you may need different (lower?) feedback resistor than 220 Ohm and maybe different load resistor for unloaded stabilization. Then again it could've just been the old POT I used to size up the pull-down resistors, it wasn't the best at either end of it's travel plus beyond 16V the transformer is reaching near it's loaded limit.

[EDIT:]The surroundings here weren't exactly "quiet" but AFAICT, the charger fitted with below circuit and a couple more tweaks (capacitors added) mentioned below, was clean enough to run a CMoy with TLE railsplitter (CMOY types are, IMO, more susceptible to power noise than more advanced amps) I powered an amp by it alone, no battery, and couldn't hear noise. My initial impression is that it seems viable for powering a headamp but I'd have to be critically listening for quite a while to be 100% sure of it, and I'd not listened to the Cmoy in a while... just don't want to lead anyone astray here, I "think" the output is good for a low-current headamp but you'd have to judge for yourself.. (YMMV).

Added a couple caps not shown in the schematic below, not for the purposes of battery charging but just in case they improved the output for running an amp. I put a 250V AC .22 mfd film cap on the AC input of the transformer (see picture linked at bottom of this post), and a .1uF 50V ceramic INSIDE of the DC power plug itself (I substituted a 5.5mm/2.5mm plug because that's what I needed instead of the original molded 5.5/2.1mm plug). Someone could try putting the .1uF within the wall-wart casing if they think they need it.[/Edit]

EDIT 2: See my post below, this circuit as-is may not be suitable for battery charging.

more pics

 

[psilosome]

Nice! I just picked up four of these buggers. For anyone in the area, the Rat Shack in DC at 1830 K St NW has about 10 left, but I suspect it won't be open on the weekend.

 

[Mitch318i]

Can anyone find proof of this deal in Canada?
my local RS has 2 of these, but they're selling for $15.99...
I doubt our dollar sucks that much
...Mitch

 

[looser101]

Mitch,

Apparently they are getting rid of them in the US. We'll just have to wait our turn.

R