battery chargers, again
in [Design]
|
From: Paul Hovnanian P.E. on 28 Sep 2009 14:00 John Larkin wrote: > Last time a car went dead in the garage, my wife's Fit, I hacked up a > charger from an old DSL wall-wart and a sabre saw as a series current > limiter. The garage geometry makes it essentially impossible for us to > push a car uphill to the street to jump it. Now The Brat left her Echo > in the garege for a month or so and it went dead, too. So I figure > it's time to buy a real charger. Went to Kragen Auto Parts and bought > two (one for here, one for Truckee) chargers. They are all "smart > chargers", namely switchers with electronics, these days. > > The battery is really dead, 1.8 volts. The first charger hums and > outputs nothing. Tried the next one: it hummed for maybe 3 seconds > then sparked and smoked inside. One month? Dead battery (1.8V)? Forget it. Buy a new battery. In the meantime, hook up a fresh battery through a test lamp and make sure there isn't a short somewhere in the car's electrical system. It may just be a dome light left on, but I've parked cars with clocks, radio memory and assorted microprocessors always on. They start just fine after a month. -- Paul Hovnanian paul(a)hovnanian.com ---------------------------------------------------------------------- Have gnu, will travel.
From: John Larkin on 28 Sep 2009 14:44 On Mon, 28 Sep 2009 10:08:54 -0400, Phil Hobbs <pcdhSpamMeSenseless(a)electrooptical.net> wrote: >Spehro Pefhany wrote: >> On Mon, 28 Sep 2009 09:53:48 -0400, Phil Hobbs >> <pcdhSpamMeSenseless(a)electrooptical.net> wrote: >> >>> John Larkin wrote: >>>> Last time a car went dead in the garage, my wife's Fit, I hacked up a >>>> charger from an old DSL wall-wart and a sabre saw as a series current >>>> limiter. The garage geometry makes it essentially impossible for us to >>>> push a car uphill to the street to jump it. Now The Brat left her Echo >>>> in the garege for a month or so and it went dead, too. So I figure >>>> it's time to buy a real charger. Went to Kragen Auto Parts and bought >>>> two (one for here, one for Truckee) chargers. They are all "smart >>>> chargers", namely switchers with electronics, these days. >>>> >>>> The battery is really dead, 1.8 volts. The first charger hums and >>>> outputs nothing. Tried the next one: it hummed for maybe 3 seconds >>>> then sparked and smoked inside. >>>> >>>> Went back to Kragen and traded up, two better chargers. Neither >>>> charges... no current, battery steady at 1.8 volts. Both have their >>>> "charging" LEDs off and "charge complete" LEDs lit. >>>> >>>> Back to Kragen, 3rd time, got all my money back. Passed by Bob Pease's >>>> place all three trips, same collection of rusty VWs everywhere. >>>> >>>> A charger that puts zero amps into a dead battery does that by design, >>>> and there's only one reason to do that: to convince people they need a >>>> new battery. Kragen's sales pitch was exactly along those lines; "Tt >>>> won't charge, so all the cells are shorted." >>>> >>>> So I went to work and nabbed a cute little Lascar bench power supply. >>>> It current limits at 1.2 amps, so I just cranked it up. The battery >>>> went instantly to 16.5 volts, then settled down to 12 or so in a few >>>> minutes, and is creeping back up. >>>> >>>> Interesting. >>>> >>>> So I guess I'll buy a couple of 3 amp or so lab supplies, with nice >>>> volt and amp meters, instead of battery chargers. They're handier to >>>> have around anyhow, cost about the same as a "good" charger, and >>>> aren't booby trapped. >>>> >>>> What Kragen is doing is fraud. >>>> >>>> John >>>> >>> Or just get a nasty big iron Craftsman charger off eBay. Two >>> transformer taps, rectifier, thermal cutout. >>> >>> Cheers >>> >>> Phil Hobbs >> >> Many of the the newer battery chargers seem to do this.. they refuse >> to charge if the battery is very dead. It's in the preferred algorithm >> description for NiMH batteries too-- I just had to modify a charging >> algorithm because people make mistakes and occasionally kill batteries >> below the minimum charge voltage per cell and you need to be able to >> override the (usually microcontroller-based) smart charger. And had >> the same problem recently with a commercial charger for a stacker >> (sort of a little fork lift). The deep discharge Marine duty battery >> was too dead for the dumbf*ck little LM339 circuit to turn the SCR on. >> At first we thought it had failed.. > >One of the problems with our software-intensive era is that the >behaviour of ordinary objects has become arbitrary, like software, >instead of merely ornery, like physics. Software-based products are also opportunities to add self-serving gotchas, like this battery charger thing maybe. Or products that read an encrypted code that allows only "genuine" ink refills or batteries or whatever, and tries to invoke DCMA to protect their monopoly. I have an Agilent DVM that has a serious hardware problem that they obviously hid with firmware. > >It's nice to warn people, but building something that refuses to do the >job it's built for is poor engineering. (Of course our current plague >of tort lawyers is partly responsible, but only partly.) The charger has no warnings anywhere. The "charged" LED comes on. The implication to common folk is that a discharged battery can't be charged and has to be replaced. That's the only kind that Kragen seems to sell. John
From: John Larkin on 28 Sep 2009 15:07 On Mon, 28 Sep 2009 13:34:44 -0400, PeterD <peter2(a)hipson.net> wrote: >On Mon, 28 Sep 2009 07:09:34 -0700, John Larkin ><jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: > > >>> >>>John, you are skilled enough to build a charger! Just do it... <g> >> >>I'm skilled enough to buy a B&K or whatever bench supply from Digikey, >>too. >> >>John > >But where's the fun in that? What I ended up doing was finding a bunch >of old linear supplies in my junk box. Each was adjustable from 12 to >15 volts, so I set the regulators to 13.8 volts, and they work well as >trickle chargers. (IIRC they are rated at 2 or 3 amps). I have plenty of other stuff to design. I did hack a charger from an old DSL wall wart and a sabre saw; isn't that enough? John
From: Jon Kirwan on 28 Sep 2009 15:54 On Mon, 28 Sep 2009 10:45:51 -0500, Tim Wescott <tim(a)seemywebsite.com> wrote: >On Mon, 28 Sep 2009 10:08:54 -0400, Phil Hobbs wrote: > >> Spehro Pefhany wrote: >>> On Mon, 28 Sep 2009 09:53:48 -0400, Phil Hobbs >>> <pcdhSpamMeSenseless(a)electrooptical.net> wrote: >>> >>>> John Larkin wrote: >>>>> Last time a car went dead in the garage, my wife's Fit, I hacked up a >>>>> charger from an old DSL wall-wart and a sabre saw as a series current >>>>> limiter. The garage geometry makes it essentially impossible for us >>>>> to push a car uphill to the street to jump it. Now The Brat left her >>>>> Echo in the garege for a month or so and it went dead, too. So I >>>>> figure it's time to buy a real charger. Went to Kragen Auto Parts and >>>>> bought two (one for here, one for Truckee) chargers. They are all >>>>> "smart chargers", namely switchers with electronics, these days. >>>>> >>>>> The battery is really dead, 1.8 volts. The first charger hums and >>>>> outputs nothing. Tried the next one: it hummed for maybe 3 seconds >>>>> then sparked and smoked inside. >>>>> >>>>> Went back to Kragen and traded up, two better chargers. Neither >>>>> charges... no current, battery steady at 1.8 volts. Both have their >>>>> "charging" LEDs off and "charge complete" LEDs lit. >>>>> >>>>> Back to Kragen, 3rd time, got all my money back. Passed by Bob >>>>> Pease's place all three trips, same collection of rusty VWs >>>>> everywhere. >>>>> >>>>> A charger that puts zero amps into a dead battery does that by >>>>> design, and there's only one reason to do that: to convince people >>>>> they need a new battery. Kragen's sales pitch was exactly along those >>>>> lines; "Tt won't charge, so all the cells are shorted." >>>>> >>>>> So I went to work and nabbed a cute little Lascar bench power supply. >>>>> It current limits at 1.2 amps, so I just cranked it up. The battery >>>>> went instantly to 16.5 volts, then settled down to 12 or so in a few >>>>> minutes, and is creeping back up. >>>>> >>>>> Interesting. >>>>> >>>>> So I guess I'll buy a couple of 3 amp or so lab supplies, with nice >>>>> volt and amp meters, instead of battery chargers. They're handier to >>>>> have around anyhow, cost about the same as a "good" charger, and >>>>> aren't booby trapped. >>>>> >>>>> What Kragen is doing is fraud. >>>>> >>>>> John >>>>> >>>> Or just get a nasty big iron Craftsman charger off eBay. Two >>>> transformer taps, rectifier, thermal cutout. >>>> >>>> Cheers >>>> >>>> Phil Hobbs >>> >>> Many of the the newer battery chargers seem to do this.. they refuse to >>> charge if the battery is very dead. It's in the preferred algorithm >>> description for NiMH batteries too-- I just had to modify a charging >>> algorithm because people make mistakes and occasionally kill batteries >>> below the minimum charge voltage per cell and you need to be able to >>> override the (usually microcontroller-based) smart charger. And had the >>> same problem recently with a commercial charger for a stacker (sort of >>> a little fork lift). The deep discharge Marine duty battery was too >>> dead for the dumbf*ck little LM339 circuit to turn the SCR on. At first >>> we thought it had failed.. >> >> One of the problems with our software-intensive era is that the >> behaviour of ordinary objects has become arbitrary, like software, >> instead of merely ornery, like physics. >> >> It's nice to warn people, but building something that refuses to do the >> job it's built for is poor engineering. (Of course our current plague >> of tort lawyers is partly responsible, but only partly.) >> >> Cheers >> >> Phil Hobbs > >OTOH, physics combined with real parts can get pretty ornery by itself; >there are decision trees that would take a square foot of board to do >with diodes and op-amps that can be done with an 8-pin PIC and a less >than 256 words of program memory. I think Phil's point (or the way I personally took it, anyway, as I really shouldn't attempt to speak for him) is that software allows behaviors that aren't at all similar to how nature itself operates. Nature is very, very consistent and once one learns it well, those lessons can be applied in new situations. When software is inserted into the mix, that can be quite a different story where no possible prior experience elsewhere may act to inform expectations. Prior digital states matter in bizarre ways, sometimes. Despite some possible advantages, I'll take an analog, physical connection between my steering column and the wheels over a digital, software controlled one almost every time. At least I know how to interpret some of the feedback I get that way based on other experiences in life. What would be nice is if software people would know enough physics and mathematics to carefully design and thoroughly test algorithms that will behave more closely to their physical analogs found in nature -- or else develop new models that are "readily learnable" from natural experiences elsewhere, so that whatever the behavior happens to be it is close enough to what one might find at our macro scale of life to be considered congruent to natural laws. But that requires a solid knowledge of those laws in order to know which end is up. >I can't agree with you more that building something that can't do it's >job is poor engineering. I think much of the reason that software >becomes a problem is when people with only software expertise are called >on to write software to embody things like charge algorithms. Software people involved in embedded work that interacts with nature (as opposed to embedded work, for example, that is about iPods or cell phones, where the nature-part of the system may be limited to knowing the human eye and hand well), should know some electronics, physics, numerical methods, signal processing, and mathematics. In the end, it is the embedded software developer who is bringing everything together into a successful system (or one that isn't.) They must be able to interpret the work of physicists, electronics designers, and so on and be involved early so that issues related to software that others may not be as well informed about can be brought forward earlier and considered. This can include a wide spate, from numerical issuess few physicists may be aware of when they write things down to issues that others should have been aware of, but missed, such as z-space vs s-space. It's always good to have someone else similarly informed looking over the work to help catch things before they are baked into the system. Of course, if it is a hand-held iPod, you might want someone versed in various media formats and play methods. It boils down to having an embedded software engineer familiar with the application space, I suppose. And not merely "a programmer." >Consequently one ends up with a whole bunch of guesses embodied in the >software, rather than a bunch of (or a few) correct conclusions. Indeed. When ignorant, a software programmer may not even realize they are making flawed assumptions on their own or implementing the flawed assumptions of others (such as a physicist's natural assumption that operations and constants are infinitely precise.) Jon
From: Joerg on 28 Sep 2009 17:45
John Larkin wrote: > Last time a car went dead in the garage, my wife's Fit, I hacked up a > charger from an old DSL wall-wart and a sabre saw as a series current > limiter. The garage geometry makes it essentially impossible for us to > push a car uphill to the street to jump it. Now The Brat left her Echo > in the garege for a month or so and it went dead, too. ... It is not normal for a car battery to discharge so far over just a month, not in the climate zone your are living in. Unless the battery was shot to begin with or something in the car was left on. > ... So I figure > it's time to buy a real charger. Went to Kragen Auto Parts and bought > two (one for here, one for Truckee) chargers. They are all "smart > chargers", namely switchers with electronics, these days. > Big mistake :-) Electronics <-> cars, IME not a good combination. > The battery is really dead, 1.8 volts. The first charger hums and > outputs nothing. Tried the next one: it hummed for maybe 3 seconds > then sparked and smoked inside. > > Went back to Kragen and traded up, two better chargers. Neither > charges... no current, battery steady at 1.8 volts. Both have their > "charging" LEDs off and "charge complete" LEDs lit. > > Back to Kragen, 3rd time, got all my money back. Passed by Bob Pease's > place all three trips, same collection of rusty VWs everywhere. > Why didn't you stop at Bob's place for a beer? He'd have given you a real charger, meaning a transformer, a rectifier, a power cord with plug (or maybe without plug), some old meter, probably all neatly assembled on a piece of plywood with some cinge marks on there. And it sure would have worked. > A charger that puts zero amps into a dead battery does that by design, > and there's only one reason to do that: to convince people they need a > new battery. Kragen's sales pitch was exactly along those lines; "Tt > won't charge, so all the cells are shorted." > > So I went to work and nabbed a cute little Lascar bench power supply. > It current limits at 1.2 amps, so I just cranked it up. The battery > went instantly to 16.5 volts, then settled down to 12 or so in a few > minutes, and is creeping back up. > > Interesting. > > So I guess I'll buy a couple of 3 amp or so lab supplies, with nice > volt and amp meters, instead of battery chargers. They're handier to > have around anyhow, cost about the same as a "good" charger, and > aren't booby trapped. > Until one fine day ... *FOOMP* ... and there goes the "Panazonc Special" output electrolytic. > What Kragen is doing is fraud. > It certainly was junk if it didn't even have any status display. A good charger must be able to start from zilch. Interestingly, I found out that many designers of SMPS do not have foldback limit, they just error. Maybe they don't know how to do that anymore these days. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM. |