Woe is me! my project has had a slight problem.

Hi,

I have a problem with my caving torch which worked fine last night but we have a small/big problem this morning.
Last nights post on the caving forum:

Oldham conversion complete!

I have finished my second light project!
(my first is still in prototype stage due to irritating delaying electronics suppliers).

The light is based on an old Olham 'type o' light.

The original battery pack was thrown straight in the bin**

I have made a custom pack from sub-c cells set in thermo plastic.
It is light enough to wear on the back of the helmet and moulded to the shape.

The head unit is the same from the outside but when you look through the glass it looks like something out of a low buget sci-fi movie; Loose wires, circuitboards, optics, a friking expensive LED and most importantly duck tape and meccano!

It still charges through the original charger and should be completely cave waterproof*

All together it seems at this stage a total success.***





Some stats for those science inclined:

Oldham:
Maximum output: ~42lm
battery life: ~8 hours (yeah right)
bulb life: ~80hrs

Newham:
Maximum output: 150+ lm (on high, 5 'log based' light settings in total)
battery life: over 2 days (on high, over a month on low)
bulb life: ~10000hrs

The problem was that a magnet on my desk attracted both cables on the batterypack and caused a short.
I don't know how long it went on for but I only spotted when the pool f water under the batteries was 6 inches across.

my circuit won't light from this pack that still reads 6v but will light from four normal aa nimh. which is a little strange.
So there is a serious problem somewhere.

What is a safe way to check max cell current?

my concern is any idea how much damage to the capacity this will have caused?
Are these cells still safe to use?

does anyone need 5 cylindrical rust covered acid soaked paper weights?

Also does anyone have any good advice?

please rescue my project!

If I get some decent responces I may post a few good photo's of the torch and burst battery pack.

If you leave a voltmeter attached while trying to draw power from your over-exercised pack, what does the voltage drop to?
What voltage does your circuitry/LED need to turn on?
What was your battery pack - 5x NiMH cells?
Can you get to the individual cells to measure their voltage when attempting to draw current from them?

I'd be interested in seeing pictures and hearing more about your project - I also make caving light units to fit in Oldham headsets.

I am using 5 sub-c cells, nimh.
It's quite a heavy pack but you can't complain when you can run a torch like this for a full weekend without needing charging.

I have dimantled the battery packs and am running discharge cycles through the charger (it was fun trying to fit sub-c cells into a AA battery analyser).

The circuit takes 1.1v max and the LED on max about 3.5V.

The first 3 cells seem to perform OK in the analyser but i won't know for certain until tonight, the fault is either with the other two (couldn't fit more than three in the analyser at one time) or somewhere else.

A lot of the wires were melted through in the battery pack and just the rust dust was conducting which could be a good reason for failure.

I am thinking about putting a fuse in the pack over each cell.
I dread to think what it would feel like getting a short off a 300w battery pack while swimming if there was a leak.


What are you using for your projects and where are you based?
I am happy to cave anywhere in England as long as I can buy breakfast at Bernie's first.

My main caving light is a bit more exciting, we have on max around 1000lm for iluminating huge chambers but a few day runtime on low (which is bright enough to cave on easy)

I use 3x 18670 or similar (7/5AF, 4/3A, etc) NiMH cells to get a ~4Ah pack, which lasts me a minimum of 8hrs on full output with slowly declining light
after that, or ~150hrs on mimimum setting, which is still usable for moving around.
The packs are small enough to make carrying a spare in a pocket easy, and I can run the light off 4.5V flatpacks if I'm away from electricity for a long time. Currently, all my NiMH packs are made from salvaged cells from medical equipment backup batteries, which have to be replaced every 1 or 2 years even if the cells are perfect.

My lights have [selectable] 2 or 3 power levels per LED, with the two highest powers using very low dropout (~70mV) linear regulation, and the low setting using PWM.
Linear does involve some efficiency losses, but with the pack voltages being a fairly good match for the LEDs, the efficiency still starts off at >75% and just goes on improving as the battery discharges. It also means that there's an automatic graceful tail-off as the battery approaches exhaustion, and it allows easy building-in of redundancy into the circuitry.
In practice, it seems like ~85% of 3xNiMH battery capacity is used before the circuit starts to slide out of regulation on its high setting, which seems about right.

I run at a max of ~1Watt per LED, which actually seemed perfectly adequate with Lux IIIs and Is, but which is much nicer now I'm using SSC P4s and Cree XR-Es. 1 Watt also means dumping heat doesn't require any fancy solutions - just a heat spreading plate to mount the LEDs on.

Most of my caving is Yorkshire, with some in N. Wales, though that's mainly digging.

Have you done any heat tests on the XR-e running inside a oldham?

I have mine maxing at 500mA but generally using 250 or 125.
The headset has the biggest heatsink that could fit in it, but no venting to the outside. I was considering using the charge path as a thermal channel conecting it to the blade but I reasoned that Oldham isolated it for a reason (even if the reason was mine gas =-0)

Do you think this can be run to safe temps on 500ma or should I knock it even lower?

Digging seems a nice idea but I prefer my caves pre-built =-)

PeterScowcroft said:

I am using 5 sub-c cells, nimh.
....
I am thinking about putting a fuse in the pack over each cell....

Click to expand...

If you have the cells in series, that's useless at best - the fuse will react on the current, and as long as the internal cabling is ok, the current through all the cells is identical (if the internal cabling melts, the efficiency of the fuses will be strongly depending on the exact failure mode - and if you get dirty water inside the battery pack, the fuses will simply be bypassed).

But: for your application, you might want to look into 'Polyswitch' self-resetting fuses (actually, they're PTCs, but they're sold as and act like fuses). I suspect if the fuse blows in the cave, the number of spares might be somewhat limited - with the polyswitch, you just wait for a few minutes after clearing up the cables for it to cool down, then you can go again. I'm using them on my bicycle battery packs for ~25 years (after a similar event demonstrated that even an old NiCd pack can deliver enough current to melt 1 mm^2 wires...).

Bye
Markus

I did some rough heat teats in an Oldham headset, with a random thermistor, roughly calibrated by dunking in various mixes of fresh (cold) and boiling water.

With a Luxeon-based lamp going flat out on 3xNiMH cells, drawing ~450mA, the ~1.5 sq. in. aluminium- plate inside seemed to get to about 30-40C above ambient. One thing I was surprised about was the speed of response to changing outside temperature - running the lamp in a warm room until it stabilised, if it was then placed on a cold windowsill in a draught, the temperature inside reacted within only a few seconds.

I tend to run my own light at full power (total ~450mA) much of the time, and a fair number of other users seem to do the same, and I haven't noticed any heat-related problems - year-old much-used lights seeem as bright as newly-made ones with similar LEDs.

With a P4 XR-E, somewhat more electricity gets turned into light than with a ~40lm/W Luxeon, so maybe there's roughly 25% less waste heat to dump?

I have been told for a white source that 250lm/w is perfect efficiency so this at 100lm/w compared to 40lm/w should dump 40% less heat. (based on potentionally horrendously flawed physics)

So using my specious analysis of thermodynamics meets electroluminescence I could push it to 750lm safely?

I have seen the polyswitch resetting fuses on Maplins and thought they looked great but spending £4 p&p for a £1 item seemed steep.
Anyone know where else you can get them?

Also while I am here is there anything like an 'anti-fuse' that isolates the batterypack if the voltage gets too low?
I.e. if the pack reads <5v it gets isolated until the voltage increases (when you stick a charger on it)? I wanted something like this to protect my cells from over discharge.

Hmm while I am on the subject of electronics!

Remembering from ages ago that this might be what a transistor does.
I am also looking for something that turns of a switch on a seperate circuit when one circuit has a voltage across it.

basically something that will isolate my computer circuit when I plug my charger in. on paper my circuit is designed to take the charge fine, but then again on paper oldham lights can run 8hrs!

PeterScowcroft said:

I have been told for a white source that 250lm/w is perfect efficiency so this at 100lm/w compared to 40lm/w should dump 40% less heat. (based on potentionally horrendously flawed physics)

Click to expand...

Well, simplifying the maths, if we take 240lm/W as full conversion of electricity to light, a 40lm/W is wasting 5/6 (10/12) of the input energy, an 80lm/W 4/6 (8/12), and a 100lm/W 7/12.
For the same input power, compared to a 40lm/W, an 80lm/W will generate 4/5 as much waste heat, and a 100lm/W will generate 7/10 as much

PeterScowcroft said:

Also while I am here is there anything like an 'anti-fuse' that isolates the batterypack if the voltage gets too low?
I.e. if the pack reads <5v it gets isolated until the voltage increases (when you stick a charger on it)? I wanted something like this to protect my cells from over discharge.

Click to expand...

Can you set your power controller to just not turn on when the battery voltage is too low?
I guess one advantage of using a 3-cell pack is that by the time the voltage gets seriously low (<0V9/cell) the light is seriously dim.

PeterScowcroft said:

basically something that will isolate my computer circuit when I plug my charger in. on paper my circuit is designed to take the charge fine, but then again on paper oldham lights can run 8hrs!

Click to expand...

Plus, I guess you have to wonder what might happen if a breaker in your battery pack decided to open - what is the charger's open-circuit voltage, and could your circuit handle that?
However, if the charging socket connects directly to the battery, and you had some circuit disconnecting the lamp circuit if the voltage got a little too high, you'd need to make sure that the battery voltage itself could never get high enough to trigger the circuit, otherwise you could have a lamp that didn't light when disconnected from the charger, at least until the battery voltage dropped sufficiently.
What kind of circuit do you use in your lamp to regulate the light levels - how much voltage can it handle?

I think there is a slight problem on the converter. I wouldn't be surprised as I haveshorted contacts on the boad before, melted part of it with a jet flam and most recently run 300w through the board!

When the battery drops to 5v the main beam starts flashing giving you plenty of warning. When the board is off it takes such a small amount of current the batteries should never go flat....but they do!

I was thinking of something that fully isolates the cells incase of an over-drain as an emegency backup.

I have found the problem with the current battery pack. Athough the pack as a whole reads normal. All the cells bar one are reading 1.2-3v the remaining reads a pitiful .6v.
When I plugged this battery into the charger the solder tags got so hot they started melting the charger in less than 15s. I should point out they were only being charged at 200mA as I had not programed the charger yet!

I use a buck converter, not as simple as direct drive but you can't complain about the efficiency! In theory the circuit can take 24v (just in case I decided to strap two lead acid to the back of my head =-) but I only use 4-5 cells NiMH so the charger voltage is never that close.

It's not hard to have a circuit that disconnects your main circuitry if the battery voltage drops too low, but it is one more potential point of failure, which is always something to consider with caving lights.

When the board is off it takes such a small amount of current the batteries should never go flat....but they do!

Click to expand...

That seems like a hint that either the current isn't always as low as it's supposed to be (or measured to be) or that the pack manages to fade faster than it should given the low drain. It could be useful working out which is the case.

It's always possible that the duff cell in your pack has been on its last legs for some time, and self-discharging rather faster than the rest.