Run-Time Plot - MadMax Sandwitch

Here is the first of several runtime plots of modified and unmodified flashlights that I'll be doing. This one is for a AA Minimag with a Madmax Sandwich. the test was done with fresh Everactive Alkaline batteries.

Equipment used is a photo cell and Radio Shack 22-805 mulitmeter with a PC interface that is using ScopeView software (V 1.08).

Edited: New plot on 09-12-02



The vertical axis is in Ma and the Horizontal axis is 15 min per division.

Edited:

It was pointed out that to use the photocell properly, measurments of ma and not mv were needed. Thus a new chart.

Added 09-12-02

Runtime plot using Lithium batteries

Nice graph Roy, except the scope view representation reminds me too much of work! I was under the impression that the badboy yellow is more tightly regulated then the madmax but the madmax does pretty well, especially considering you're using alkalines.

As I understand it, Madmax sucks a battery down to zero where the Badboys leave enough voltage so as not to hurt rechargable batteries.

Nice graph.

That looks very typical for a good set of Alkaline batteries.

Intermittent usage the Madmax really shines. I think if you let the batteries rest they would provide mamy more minutes of reduced light.

Our nightime minimag with a Madmax has been run every night for the last two weeks up to 1-2 hours several times. It's been dim but it's still putzin along. Last time I took the batteries out they were both around 1V each. That was last week.

WayneY

careful with making the batteries zero...

i had a set of duracrapcells pop while getting them at less than 1 volt. i had to REALLY twist hard to get the tailcap off. clean up the dripping battery juice, and got the madmax module out clean and safe, then took the AA tube and soaked it and scrubbed the inside with metal polish. DANG! but that aluminum "scores" quickly. it's permanently etch. shee-it :>

btw folx, maglite has a new (to me) color out in targets: jade - a darn luscious green.

Roy, a little off topic but how expensive is it to create your test setup? I guess that would include a multimeter with pc option, applicable com cable to pc unless not supplied, a photo cell and software unless the software came with the pc option.

Roy, I note that on the vertical axis you are measuring voltage. Did you have the photocell shunted with a resistor? If not, it would appear that you are measuring the Voc of the photocell [loaded only by the input resistance of the meter]. Photocell Voc is not proportional to illumination. Photocells are current sources with current proportional to illumination [assuming spectral distribution remains constant].

I'm using a Radio Shack #22-805multimeter with a pc interface and software (p236 2002 catalog). This is no longer carried in the catalog but can still be found in local stock for a sale price of $39.95. My local store did not have any, but the manager checked his computer and found a store that had some in stock and ordered one from that store. The photo cell came from a $12.00 Radio Shack kit that demonstrated solar power.

Yes, I realise my measurments don't directly relate to lux or candlepower. That's why I only report mv output of the photo cell. No, I don't have a shunt on the meter. I do keep the light to photocell distance constant so the relative output between flashlights can be conpaired. The 2002 Radio Shack catalog has a foot - candle light meter (p239) but it doesn't have comupter logging.

Originally posted by Roy:
Yes, I realise my measurments don't directly relate to lux or candlepower. That's why I only report mv output of the photo cell. No, I don't have a shunt on the meter. I do keep the light to photocell distance constant so the relative output between flashlights can be conpaired. The 2002 Radio Shack catalog has a foot - candle light meter (p239) but it doesn't have comupter logging.

Click to expand...

<font size="2" face="Verdana, Arial">Roy, you missed my point. Photocell Voc doesn't even allow relative comparisons. Try this experiment. Find a light source and distance that gives a measurement of around 300 mV with your technique. Now decrease the source to photocell distance by one-half [this increases the illumination by 4X]. Note that your reading hasn't changed much. Repeat experiment measuring photocell output current on uA scale.

I was just going to say something about this, but I got beaten to it.

I connect my photocell to the meter and set the meter to a low milliamp range. The I butt the business end of the light directly against the cell, throw a black shirt over the whole mess, and then fire up the logging program.

An example (using a test circuit with some regulation) of the output would then look like this:



The program used to create this chart is homebrew. It's auto-ranging (places the brightest recorded event at the top automatically, regardless of how bright or how dim a light I test), and can create a graph for a test up to 364 hours in length; using a sample rate of 1 reading every 10 minutes.

So, I need to be measuring ma not mv. Ok to use the shunts in the multimeter?

Originally posted by Roy:
So, I need to be measuring ma not mv. Ok to use the shunts in the multimeter?

Click to expand...

<font size="2" face="Verdana, Arial">Yes, and generally yes. You want the shunt to have at most 250mV across it. If you check your meter manual, it will give the FS voltage "burden" for the various mA scales. Generally these are 250mV or less.

new chart in the first posting.

Originally posted by Roy:
new chart in the first posting.

Click to expand...

<font size="2" face="Verdana, Arial">Nice job, Roy! I suspect that your future runtime chart postings will be very popular and appreciated. One note of caution, your current test setup gives very good relative light measurements for a given LED light under test since the spectrum doesn't change very much as the light dims. This will not be true for incandescent lights because the spectral response of a silicon solar cell is very different than that of the human eye. Solar cells are much more sensitive in the red and infrared portion of the spectrum.

So it's my turn to step in and whine right

How ya getting the graphs, screenshot or that? It's a little, uhh, blurry but I am sure it's an easy fix.

If it IS a screenshot, what program are you using to resize it etc.

I'm using a program called ScopeView that came with the multimeter. It prints a black and white version of the color plots on the screen. I then scann the B&W plot and then resize the results. The scanner creates a plot that is 2000+ pixels wide and I reduce it down to 600, which is what you see on CPF. I'm using ADCSee to resize the scanned material (came with the HP 4400c scanner).

Hmm... The graph image seems quite legible to me. It looks pretty good, and I want to thank Roy for taking the time and trouble to make it. Hope to see many more in the future.

Roy, if you want to try something different, here is a technique that has worked well for me in Windows:
1) Get the graph window to display on-screen.
2) Press Alt+PrtScn to place the window's image into the clipboard buffer.
3) Fire up your favorite graphics editor. (Even Windows "Paint" will do.)
4) Paste into the graphics program.
You should now see the captured window image in your graphics editor.
5) Crop/resize as desired.
6) "Save As" a .gif or .jpg.



--Steve

There is a very simple, stable freeware for sceen-capture called Printkey. It allows cropping and color depth modifications as well as compression choice easily in one app. The current version is no longer freeware, but the older versions are still available at the ftp servers. Google search for Printkey. I'll email it if anyone wants it, only 282kb.

Found freeware version 5.05 on an FTP site via google search like you said. Handy. Seems to work fine.

Also, if any of you own or have JASC's Paint shop pro, from version 3.xx on, screen capture is built in and it automatically pastes your selected area into a new document window, and after you're done editing colors, sizes and whatever, you then can save it in any image format that it supports. (it supports many). Years ago it was shareware, but it never disabled itself, just had one annoying nag screen at the beginning. I don't know if the current version is offered this way though. Although it might be. After years of guilt, I finally broke down and purchased version 6 about 2 years ago for around 60 bucks.