ION's Tips, Tricks, and Techniques

[ion]

I'll be posting Tips, Tricks, and Techniques for electronic circuits, power measurement and breadboarding, since these handy aids seem to get buried in regular threads. It will be a locked thread, any questions PM me and I'll post them here.

[ion]

Here is a little setup I've been using for years to determine power to any device that would normally require one or more 12V batteries. The Kill-A-Watt also has an elapsed time power monitor for long haul testing.

At $17 from newegg, this little device boasting 0.2% accuracy (True RMS voltage) is the best buy on the market for a researcher on a budget. Not a perfect solution, but beats the pants off trying to guess with batteries. At the very least, you will know if you are in the ballpark.

It can save hours of guesswork and give true power readout. Components should be chosen according to expected power use e.g under 3 amps, use a small Variac to keep excitation current to a minimum.

The excitation current for the Variac will be very low if it is a small unit, and it's losses and those of the other components can always be factored out by a test with a known resistive load.

I used a smoothing choke after the bridge rectifier to prevent high peak charging currents into the capacitor. It is optional. An isolation transformer can be added between the mains supply and the Kill-A-Watt for safety.

Another simple option would be to plug a regulated switchmode type 12 Volt wall wart into the Kill-A-Watt. I'll discuss optimizing this in a separate post.

[ion]

As I stated previously, I would expand on the use of a KILL-A-WATT / Wall Power supply combination

I tested a DVE Switching Power Supply #3872A127 with a KILL-A-WATT power monitor.

This wall wart is found on many scanners and other devices. It is rated at 12 VDC and 1.25 Amps.

This combination would be an ideal replacement for a small 12 volt battery source, but now you will be able to read actual power drawn by your device under test.

Unloaded, the device yielded the following:        Amps 0.01, VA 0.01, Watts 00, PF 0.42
Fully loaded,(1.25 A) these were the readings    Amps 0.23  VA 28    Watts 17, PF 0.60

As you can see, the unloaded power drain was negligible, and the loaded 15 Watts power drain created a 17 Watt reflection in the readings. Therefore there are about 2 Watts lost in the power supply.

Since we have only integer readout of Watts, we can get a better idea of the losses inflicted by the power supply by using the elapsed power feature of the power meter, and running it for identical time periods, loaded and unloaded.

I recommend the regulated switchmode supplies, as they dissipate less power than linear regulated types under medium to high load conditions. Your power supply can be characterized at several values of load, and now you will have a calibrated system where losses are known.

For long term testing, the elapsed power meter is very useful.

[ion]

Before I look to exotic failure mechanisms for semiconductors, I try to exercise good handling techniques and avoid ESD damage modes, which usually can create a tiny puncture wound of the mosfet or other semiconductor.

These tiny punctures do not always immediately prevent the device from operating, but degrade operation over time in the circuit leading to catastrophic failure, and usually when operated at a higher voltage.

I use these precautions to prevent ESD and other types of damage.

1) Use a grounding wrist strap before handling semiconductors, especially in winter or other low humidity environments.

2) Wrap the leads close to the device with a small strip of aluminum foil, which can be removed after it is soldered into the circuit.

3) For mosfet's, the first device I solder into the circuit is a protection Zener from gate to source BEFORE REMOVING THE FOIL. Alternately, the last device you should solder into your circuit is the mosfet.

4) Be sure to use a grounded soldering iron and disconnect your circuit from any ground paths before soldering.

5) If you use an ungrounded iron that has not been used in a while, the ceramic element can absorb moisture creating a leakage path to the AC line. Be sure to leave it on for a while to bake out the moisture before soldering or ground it.

6) Use an MOV or other clamping device to protect the drain from high peak voltages.

7) Add some current limiting such as an incandescent lamp in the voltage source or use the current limit feature of your power supply. This will not help if you have a large capacitor downstream as the peak current can rapidly take out a semiconductor.

If you must operate a Tesla coil or other such device near your inventory, be sure your stock is totally wrapped in aluminum foil. Best to store devices at some distant location.

I have had an extremely low failure of devices when these precautions are used.

[ion]

Per POYNT's request I have archived the subject matter here:

CONTROLLING THE DECAY TIME OR BEMF OF AN INDUCTOR.

Consider this: A lot of designers throw a diode across a relay coil in an effort to snub the BEMF spike and keep it from destroying the drive transistor.

But this is a rather poor approach, especially if you want the relay contacts to quickly open, as the diode keeps current circulating in the coil and hence the magnetic structure releases slowly. If you have an AC load on the contacts, severe arcing will occur due to the sluggish opening of the contacts.

A better approach is to use a series R-C across the coil. In this way the circulating current will snap off much quicker, the drive transistor will be protected, and the contacts will open much faster.

Alternately you could use a diode in series with a parallel R-C across the coil typical of a flyback snubber.

The R-C creates a slight delay in drop off of the coil current. Appropriately chosen you may get the delay required.

Maybe this technique can be used to provide the timed delay of the BEMF for the the RomeroUK Muller device

Another idea would be to switch off the drive current, but keep the current circulating in the coil with another transistor across the coil until you desire to turn that second transistor off.

Study electronic ignition systems as used on small engines (magneto's). Some of these have automatic timing advance.

The switch in the synchronous method is representative of a transistor and is opened at the moment the BEMF is required. Note that the drive will have been turned off some time earlier and the circulating current will maintain the field until you desire to release it.

It would be simple to create a timing diagram and the appropriate drive signals for the mosfet and the transistor switch.

Mosfets are shown as the main drive but these could also be BJT's

Others have suggested diodes such as TVS devices, Zeners, or MOV's could also possibly be used.