With only some minor comedic incidents
Last post on this topic, I explained the why I needed one and how I was going to build it. Go there for schematics and function details. Finally having gone electronics shopping last week, I was able to get the one critical part I needed for this project: a 100Ω trimpot. One measly part which took me a month to get off my ass and get. Such is life :). I also managed to grab some beefier crocodile clips to make my test leads that much more durable and awesome looking, not least because I could actually fit the rubber boot over the two wires to each clip!
Soldering was trivial and I managed not to burn myself or set anything on fire, or wreck the board. I did cut myself though when I jabbed my hand with a pair of needle nose pliers though the damage to my hand is so slight I’m only mentioning it as I take that as a sign I’m actually doing something useful if I am bleeding.
With the board now fully assembled I thought “what they heck” and decided to power it on to see if it blows up or something. To my delight, it did not blow up and worked as expected! Hooray! That never happens.
Calibration goof up
This is just funny. Remember in the last post how I thought I was being smart by setting the output current to 1A instead of 100mA so that mV = mΩ? Well that part works a treat. My multimeter does read directly in milliohms no problem. The difficulty is in calibrating the damn thing. Like most DMMs, mine comes with 2mA, 20mA, 200mA, and 10A current ranges. In order to calibrate this thing so it’s actually useful, I need to have the LM317 outputting a constant current of as close to 1A spot-on as possible to get an accurate reading. Since the output needs to be 1A, I have no choice but to use the 10A range which will only give me a reading of 1.00A, which it did. This is great at first glance, spot on 1A right? Well no. What about the third digit? Assuming it rounds it, that means that it could potentially be ±5mA and I wouldn’t even know it! Had I stuck to one original design, I could have had it set to 100mA which i could dial in to within a bee’s dick of 100.0mA (±50µA making it all the more accurate. For lack of one extra digit on my multimeter, I cannot calibrate it any better than I have it now. Unless I devise some clever plan to give it a known resistance… chicken and egg scenario, that’s what I built the damn thing for! Oh electronics will get you :). Anyone have a multimeter than can measure 1.000mA that I could borrow for 2 minutes?
These sorts of issues will plague me until I get a better multimeter, no doubt about it. When I build the power supply those extra digits will come in handy for calibration.
I measured a bunch of low-ohm resistors I had laying about and got reasonable results according to their stated values which is nice. I know it’s still a bit off, has to be, but it will do for now. A 1Ω 1% resistor I measured read 1.025Ω making it 2% off it’s stated value, so unless this resistor is out of spec, the thing needs some tweaking. I won’t know until I calibrate it properly.
The next part is almost trivial, almost. I just need to safely whack it in a case and call it a day. I did notice the thing heats up quite a bit and I measured 83°C on the case of the LM317 (well within it’s limit) so that’s not a big deal. I may drive it harder, by using a 10Ω load and have it up the voltage to push an amp through it and see what the temperature shoots to. The thing isn’t really required to be on for hours or anything, just enough time to test a resistance quick, so I may just try to get away with some vent holes or (depending on the temperature at 10Ω) I may add a little fan to cool the bugger.
Then it’s case drilling and mounting everything, whack in a power LED. Oh and add an appropriate power supply for it, which I still do not have. Still, hard part over.