Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror
×
User Journal

Journal Alioth's Journal: Success and failure

The last few days has been marked by success and failure in various projects.

Fail: Spectrum ethernet project. I had some PCBs made, and I've not had much luck with them. I found one minor bug (and the reason I didn't catch it is it's a simple jumper to enable/disable all execution traps, and it's wrong in the schematic - so when I checked the PCB against the schematic, it looked right but was in fact wrong). However, I can just change the CPLD configuration temporarily to work around that. No big deal.

However, the first one I started assembling all went wrong when I put the memory on and tried to test it. So I thought before getting the hot air gun out, I'd assemble the next one bit by bit and see if it failed when the components for a key circuit were added, but I never got that far. The tip of my soldering iron crumbled, leaving me with the next chip to test only half soldered. The whole sorry story is at http://spectrum.alioth.net/doc/index.php/Current_events. The annoying thing about the PCB that doesn't work at all is that I can't find any electrical faults, and the chip pinouts match their datasheets. So I hot-air'd off the flash chip and it stopped crashing the machine, and the static RAM could be paged, read, written... so a duff flash chip. The trouble is the datasheet is also ambiguous about its TTL-compatibility: it gives two sets of DC characteristics, one for TTL compatible, and one for not - but there's no clue as to the package markings that differentiates them. I may end up having to make a breakout board for the flash chip and testing it separately.

I have decided to get a syringe of solder paste and fine (0.25mm) applicator needles, hand soldering fine pitch SMD components gets old quite fast.

Mixture of success/failure: My RC helicopter. I've spent all weekend crashing. I have the soft plastic rotor blades on at the moment which while making it fly like a blancmange, do mean that crashing seldom hurts anything. Which is a good thing. I'm trying to learn flying nose in, and numerous times I've pushed the cyclic the wrong way and had to jump out of the way while the heli smashed into the wall or other object. This morning I didn't quite get out the way in time and I hit myself (and drew blood). However, I'm getting much better at flying nose in - I had a couple of batteries worth of cruising around the lounge (it's very gusty outside at the moment) where I didn't crash while flying nose in. The trouble with nose in is that your brain has to be working 180 degrees to what you're looking at...and with a helicopter when things go wrong they tend to go wrong in random directions. But I'm getting there.

Success: Since my soldering iron tip crumbled away, and I found I had no spare bits left, I decided to test a small project for the garden. I have a solar power system on the shed roof which I built a couple of years ago to power stuff like garden lights and other things like that. I have a 240 volt pond pump at the moment, which runs off this system via an inverter, but the pump is far too powerful and in any case doesn't really get along all that well with a modified sine wave inverter (once it's running it's fine, but it'll often sit there with the motor jiggling, very little throughput, and huge current draw until you lift it out the pond and swish it around in the water a bit). So I thought I'd get a DC pump, but they are all 24 volt. My solar power system is 12 volts. So I built a little 555-based switch mode boost converter, to boost 12v to 24v. I've over built it on the current side to make sure it's never under any stress, so I wound my own inductor (I got the magnet wire for this when I visited Houston and dropped into EPO) since I only have fairly puny inductors that are factory made to hand. The 555 circuit's pretty neat, it's the one I use for nixie tubes. To change the voltage, just use a different FET (of the appropriate rating) and change the resistor values on the regulation circuit. It's fairly efficient too, close to 90%, but without requiring any uncommon parts (like purpose made SMPS chips). It doesn't regulate as well as a purpose made SMPS chip, but for motors and nixie tubes, it's no big deal.

This discussion has been archived. No new comments can be posted.

Success and failure

Comments Filter:

Nothing makes a person more productive than the last minute.

Working...