Bad Design And Patterns

One trap that a designer can get into is the one where even when you are working with a new technology you do things the way the previous technology did things.  You can see this in early internal combustion engines which looked and were run a lot like stationary steam engines.

Then there’s this case.  How batteries are charged and how you handle battery life in cell packs.

 have a diving scooter that I’m selling.  DPVs are crazy-convenient if you own a boat or have a willing operator (there aren’t many, unfortunately) and want to do wreck dives.  For longer cave-dives they can be essential.  They can also be dangerous as hell if you don’t think, because it’s quite easy to find yourself 5,000′ back in a cave and then have the scooter fail and if you are ever dumb enough to do a powered ascent with one in open water you’re at an extremely high risk of decompression sickness or outright death due to embolism.  Nonetheless I love the damn things and the only reason I’m selling mine is that I’m getting out of all but the most-mundane diving stuff.

There are two ways scooters fail, basically, although there are other less-serious problems that they can develop: One is catastrophically; they have a rotating shaft with a seal that had damn well better remain watertight, and not just because water and electricity don’t get along. If the unit floods at depth you’re double-****ed because it will instantly become 30+ lbs negatively buoyant and you will be forced to abandon it.  There’s not really much you can do about this risk other than pay damn good attention to any hint of water inside when you get back from a dive, and if you find it change the shaft seal immediately.  Fortunately the seals are not hard to change and are pretty cheap; it’s the same basic design as the seal used in your pool pump.

The other problem comes from the battery packs.  They’re very high-energy and come in three forms — SLA (basically UPS batteries), NiMH (usually made out of “F” cells) and Lithium-chemistry of various sorts.  SLAs give you a fair bit of warning before they die as their discharge curve is more or less a slope, so if you’re paying attention you will notice the power level goes down — assuming a brushed motor.  Both NiMH and Lithium chemistry doesn’t work that way — you have a quick drop to a flat voltage, then basically no change until the battery is essentially dead — at which point it goes from producing plenty of current to almost-none in a very short period of time.  The controllers shut down the motor when that happens (and they need to, as otherwise you won’t get anywhere but you WILL destroy the battery pack.)

Of course having that happen 5,000′ back in a cave sucks; you either can swim out on the gas you have left or you’re dead.  If you haven’t planned for this to happen it is simply a matter of WHEN, not if, you will die because you were stupid.

Anyway, the underlying problems with multi-cell high-capacity battery packs are well-known.  Add to this that people are lazy; they want their pack recharged and they want it now.

NiMH and Lithium chemistry packs can be charged at very high rates with a 4-hour charge or faster being pretty common.  An “F” cell NiMH pack is made up of 13,000 mah (nominal) cells of 1.3V each, 20 of them.  It has a “working” voltage of 24V and is pretty-much a direct replacement for 2 SLAs, which internally are 2V per cell and six cells each, or 12 cells (instead of 20.)

The problem is that batteries are not all exactly the same.  That is, they have slightly different capacities.  Therefore, over time what happens is this:

You charge them and you have something like this:


…. and so on.

Note that you can have a roughly 250mah capacity difference with batteries that are within 10% of spec right out of the box.  This isn’t unexpected.

Unfortunately since wattage (power) is volts x amps when you discharge them the power that comes out isn’t exactly equal.  What’s worse is that when you recharge them you put in 24V @ 4A, for example, but the charge acceptance isn’t exactly equal either and since it is energy that is stored (Watt-hours or Joules) charging a string of cells inherently leads to either some being overcharged or some being undercharged.

Overcharging when you charge at high rates causes the cell to vent (it produces gas and a relief valve opens) and that destroys it.  Therefore chargers are designed to avoid this by detecting the characteristics of the charge, including in some cases temperature rise, particularly with NiMH as that’s a very reliable indicator that the cell is full and you’re going to vent it if you don’t stop pumping energy into it.

But the fact that all cells in a pack are never equal means some cells will always be undercharged when a pack is fast-charged if it is appropriately protected against any of the cells being overcharged.

Over time this will eventually lead you to run the pack down with one or more cells that have no energy in them which reverse-charges it, and that instantly destroys the cell involved.

This is how multi-cell packs that are fast-charged die most of the time.

To avoid this you can slow charge the pack at a rate that won’t heat it up enough to cause anything to vent, but will overcharge (intentionally) most of the cells once in a while.  That “picks up” the weaker cells and fixes the problem — for a while — provided you catch it before any of them are reverse-charged and short.  That does cost you some of the cycles that the battery could otherwise produce (in other words it will wear out for real more-quickly) but in most cases a pack doesn’t die because the cells wear out — they die due to inadvertent abuse caused by the above situation.

It gets materially more-dangerous with many Lithium chemistry batteries because a reverse-charge of one of those will frequently cause the cell to explode and the contents to catch fire when next charged, especially if rapid-charged, as they essentially short out internally.  NiMH and NiCad batteries will burst and make a mess but at least they don’t burn down your house.  For this reason commercially-produced lithium packs usually have circuitry in them that detects this condition and fails “hard open”, disabling the pack entirely, since the manufacturers don’t want to get sued when your house catches on fire (but do note that this protection isn’t entirely effective either!)

Manufacturers could avoid this by instead of supplying a simple series charger providing you with a fast balance charger that individually charged each cell.  This would require a connector block to handle that but the circuitry involved on the charger end isn’t really much more complex and it would completely eliminate the problem.  Yet in more than 30 years of working with various equipment that have packs like this in them including UPS gear, consumer stuff, construction equipment and similar I’ve never once seen any manufacturer do that.

Eventually all batteries wear out, but what we have here is a case of planned obsolescence, basically.  Oh, and did I mention that these packs have a habit of costing upwards of $500 — in some cases as much as a grand?

I have here one that, it appears, has one bad cell.  Of course the factory charger that came with the scooter is a 4A unit, which is guaranteed to cause problems over time.  Worse, it’s hard to find something with (1) adjustable output and (2) the ability to handle a 20-cell NiMH pack.  The manufacturer wants $800 for a new pack; I will observe that the cells themselves are available from Tenergy for $12 each, approximately.

That’s a nice racket you have there…..

So what I’m doing is replacing the blown cell and adding a pigtail so you now have access for slow-charging each of two strings of 10 cells.  Chargers that are programmable and can handle 10-15 cells are easily found and inexpensive (about $50), thanks to the RC vehicle market.

Unlike Mr. Denninger I’m not going to say that what’s happened is some kind of racket.  A racket is a an easy thing to solve. You just prove criminality, arrest some people and have an nice trial.  This isn’t a racket though. It’s worse.  It’s technical inertia. If you know about technologies  You see this happen all the time.  What happens is this.  A technology starts out fairly simple and low power.  For instance the first rechargables in tools and whatnot were eight or twelve volts and didn’t have huge banks of cells.  And the resulting battery packs were pretty cheap.  So using a series charger saved a few bucks and nobody cared too much about battery life.  Since the series charge was the technology everybody used, as time went on everybody continued to use that technology.  The chargers were readily available, the technology was sound and because everybody used the same technology, it was cheap.

So, going forward, things went from 6v, to 12v, to 18, to parallel cells for high current applications and on and on with nobody really looking at what happens during charging.  I’m going to guess that a big part of that is that battery pack designers aren’t battery pack users. That is that while they are great electrical  engineers who care about their products, they don’t have the kinds of hobbies that use those high energy battery packs.   In most things, due to the nature of the product, the product development people may not be users of the products that they develop.     Along with the problem that in most products the user feedback, for many  consumer products, essentially nonexistent.  For the needs of failure analysis, what feedback that does get back is more than likely useless.  So there is no understanding that builds up to how the batteries fail and what the issues are.

The issues that Karl raises are the thing that one finds when you use a product intensely over a long time. Which is exactly the environment that engineers do not, for the most part, exist. When you are working on a product, you aren’t concerned with how well it’s going to work five or ten years down the road, you’re trying to make it work at all.  Which is usually no small thing. Then there’s the fact that the product is on a timeline and a budget, with things like battery chargers left to last and more than likely. purchased off the shelf as “good enough.” Going a quick google search for DPVs shows prices all over the map, with what suspect the higher prices DPV’s having better battery life and able to go deeper depths.

As for Mr. Denninger’s example of getting 5000 feet back into a cave on a dive, if you didn’t plan for some degree of equipment failure, having spare air and the other things that go with cave diving, well you are asking for a Darwin award.

Still, batteries should have better, and more reliable charging options. It seems to me that the engineering would not be that complicated and with the ongoing pressure to go to ever higher power levels in batteries that last longer, with the increased cost of the battery packs, paying attention to how many cells charge of one input is going to be a concern that the companies making the tools that use the batteries will have to consider.  It’s not a good idea to let inertia lead you down the golden path forever. Because that way always leads to a wall.


One comment

  1. penneyvanderbilt · May 8, 2016

    Reblogged this on Ancien Hippie.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s