Charging "tall" battery stacks

If I'm designing for pharma, medical, industrial, military, etc. markets/clients, everything is easy-peasy. I can put test/validation procedures in writing and not worry if they are followed or not (the customer is "responsible" and capable of making his own determination as to how much a stickler he wants to be regarding "by the book"). He's probably budgeted for the added costs of my including those provisions in the design -- and for my verification of the design against his specifications. He's probably budgeted for any routine maintenance as well as some amount for out-of-warranty repairs -- as well as a contingency plan for how he's going to deal with that inevitable "down time" (hot/cold spare?). And, he's "mature" enough to know that when he cuts corners and gets bitten, it's PROBABLY *his* fault ("Hmmm... now I understand why the recommended best practices called for these components to be replaced BEFORE they fail...")

Designing for a consumer market, OTOH, is entirely different. These folks are the types who:

- had VCR's that displayed 12:00 continuously until the day they were discarded

- don't change default passwords on their "security related" products

- use same password for all their accounts (and account names like GWalpert)

- leave the cheap, noname batteries that came WITH their devices in them

- can't tell you, definitively, *if* they have a working flashlight, *NOW*

- buy the $19.99 version of a product as they can't perceive the difference between it and the $99.99 "version" of the "same" item

- "temporarily" disconnect their smoke detectors because they don't have spare batteries on hand when they start chirping -- and didn't proactively think to replace them BEFORE this happened

- don't backup their computer

- don't verify that they can RESTORE a backup (if they've actually done one)

- send money to Nigerian princes

- click on links in emails from "Mom"

- don't check the pressure in their tires

- return products because they couldn't figure out how to use them (and couldn't read the instructions that accompanied them)

Far fewer problems with computers when only the "elite" used them; the bar was set high enough that those that weren't qualified were kept away. Was software less buggy, back then?

The trick is to ACCEPT that your customers won't do as you tell them they *should* and design so that you can still give them the best user experience possible.

There are many ways of addressing these types of issues -- each with attendant costs. E.g., new smoke/CO detectors come with 10 year batteries;

10 years being the age at which you are expected to be REPLACING your smoke detectors! No need to replace batteries, EVER!

I can minimize the need for batteries to transit BRIEF outages (a few cycles) by reducing power consumption, arranging "shedable" loads to be on supplied that DON'T have additional bulk filtering, etc.

I can monitor the voltage across each of the N "batteries" in the stack to predict how the stack as a whole is likely to perform.

I can design to accept a much wider (low end) range of battery voltages without loss of significant functionality.

Etc.

Or, I can skip the battery issue entirely and tell the user to purchase a COTS UPS -- at many hundreds to kilobucks.

[I suspect the last option will quickly be criticized by potential customers: "Yeah, and you have to drop a grand on a UPS lest everything RESET each time the lights flicker!"]

The point of this post is to try to identify and quantify the benefits of various circuit/system design strategies. Not just generalizations that can't be evaluated as to their actual *value*.

"Use upper and lowercase in your password along with digits and punctuation" OK: Aa1@ Not enough characters? P@ssw0rd

In concrete terms, consider these sorts of experiments:

1 four brand new batteries, same date/lot code, 100% initial charge 2 four brand new batteries, different manufacturers 3 assortment of batteries in various states, one of which recently replaced etc.

Now, what circuit topology changes will make *noticeable* improvements in the "user experience" for each of those EXPECTED scenarios?

[A bad user experience is one where batteries fail often; or, where they have to be serviced often (replace one this week, another next week, etc.); or when the costs in time/money become "noticeable" to the user (if you had to recharge your iPhone every hour, you'd consider it a really crappy phone -- regardless of the features that it has!)]

If there is *nothing* that I, as a designer, can do to "save the user from his folly", then I should do the least costly thing and live with the consequences of the user's characteristics! And, "diligent" users (those who WOULD follow directions for a better designed product) end up seeing the same diminished product quality. :-(

ed

carded

rt)

*

ce

vely

ne)

s;

ke

on"

aced

);

m

Don you're going round in circles. If you think you can put numbers to each event, go for it, it's your product and it doesn't look like anyone else h as any exact numbers to offer.

You keep saying you want to do more than monitor each battery v and issue a n error when one fails, but what else can one do? That's it. I've already l isted the few things manufacturers do to encourage the user into good pract ice, and into accepting it's not your fault when batteries die, and enable you to deny fault or liability when the user tries to be a cheapskate and d oesn't get away with it.

Re numbers, the us gov pdf has some life numbers, but they don't line up at all with the realities of less than top notch hardware. So it's not like t here's one set of numbers, more a writhing mass of interconnected numbers & conditions.

NT

Can you show me that circuit?

You keep saying that, but it's not remotely correct. I'm tired of repeating myself. Reread what I wrote.

--

Rick C

comparator, transistor or relay, power taken from further up or further down the battery pile.

NT

I have no idea why you are calling this a "diode" circuit. It doesn't even have a diode in it necessarily. Whatever, we aren't talking about the same thing, so no point in continuing.

--

Rick C

It behaves like an ideal diode. (at-least at DC)

--
This email has not been checked by half-arsed antivirus software

Yup, rather than a Si diode which would let the cell voltage go to -1v or thereabouts. I'm not clear where the confusion was/is from.

NT