Lab equipment: To build or not to build...
I remember way back in the 1970s when we needed some quick and dirty tester in the lab we would start with hand building the circuit on some double-sided copper clad of the appropriate size, go to the shop, get some more copper clad PCB material, shear it into strips and solder it all together to make a box. Instant instrument!
I wish I had a picture of one of these as they were sometimes quite intricate and some of the engineers were quite good at making them. It was also a form of occupational therapy as it took your mind off perhaps more pressing matters and gave you a sense of accomplishment when you got done. Besides, it's always fun to make "Finished Products" as it seems to be in an engineer's blood to do so.
Today there are thousands of really nice project boxes just an Internet click away, and they may even arrive overnight; but what about specialized testers that won't fit into one of these nice pre-made box formats?
A favorite method of mine is to make an open-air chassis out of PCBs. The PCBs are cheap enough, the circuit can certainly be built on them, and with copper lettering or silk screening replacing Dymo labels you can get a nice finished front panel also.
All of the software that is needed for this is available online for free, as almost all PCB vendors have software for this that they offer if you make the PCBs with them. Or you can use the PCB design tool that you use all the time.
One example of mine that has lived through a couple of iterations is the "Constant Current Load." This started out with a really nice hand-sized heat sink and I thought, "Wow, I could make a nice little constant-current load with this for testing power supplies." 
Through the years I added a BNC on the back so that I could "whack" the load current setting with a function generator and use the load to do stepped response testing as well.
The Mark II Constant Current Load
Nearly 20 years later I discovered really nice dual-readout DVM modules made by Lascar Electronics that were perfect for this use, as the voltage and current could be read out at the same time .
So I decided to do the "Mark II" version of the load. This one incorporated dual settings, A and B, that can be toggled between with an adjustable-rate pulse generator made out of a CD4047 astable multivibrator. The control circuit used is basically the same as implemented in this Design Idea.
Figure 1: Front view of the Mark II load. These quick-turn PCBs make decent and cost effective front panels and you can use either copper or silkscreen for the lettering. I built a 1-amp model and a 10-amp model by changing the decimal point on the display and changing the sense resistor by a factor of 10. The 1-amp model allows for finer control of low current power supplies.
Figure 2: A side view shows the simple construction of the Mark II Constant Current Load. The Power MOSFET and sense resistor are mounted on a small PCB on the heatsink itself, then the drive electronics are on another PCB and finally the controls and display are mounted on a PCB that serves as the front panel. All simply screwed together with standoffs. Truly a quick and fun way to build a useful piece of lab equipment.
The Mark II version used a really huge 180-A MOSFET in a bolt-on package, simply because it will survive nearly any abuse thrown at it. Of course, on this big MOSFET the input capacitance is > 10,000 pF, so driving it took some trial and error in gain shaping to keep the control loop stable even when using an unlimited capacitive load op amp (Figure 3).
Figure 3: The essential innards of how to make a MOSFET behave like a constant current source. R1 is a power sense resistor that is amplified by IOP2 and fed to the main loop amplifier IOP1. Components R3, R4 and C2 shape the loop gain, reducing the high frequency AC gain to insure stability of the overall loop. This compensation is needed because of the > 10,000 pF input capacitance of the Power MOSFET.
This project actually used three PCBs to make the load:
- The MOSFET and Power Sense Resistor were mounted on one small PCB down in the heat sink channel
- The rest of the electronics and 9V battery were mounted on a PCB that fit the top of the heatsink
- The front panel was another PCB that had the holes for the pots and held the meter, etc.
Stepped Load Testing Power Supplies
Using a loop gain tester is the best way to make sure that our power supply loops are stable, but this is not always possible due to equipment constraints or if the power supply is a closed box. What is always possible however, is to attach a constant-current load to any circuit and sometimes this is the most useful in debugging a running system. In a running system we can see the stability of the power supply when it is driving all the normal inductances and load capacitances that affect its stability in operation.
What you are looking for on every edge of the load is a nice low impulse and a quick recovery without excessive ringing.
In just seconds one can get a good feel for a system's overall stability without the need to grab the loop gain analyzer and start breaking into the control loops.
For further reading on stepped load testing and some alternative constant-current circuits see Jim Williams' article on the technique .
Your favorite methods...
I've shared some of the fun ways that I have made useful lab equipment. What are some of your favorite ways to not only get "occupational therapy," but to make useful test equipment?
 Hageman, Steve "Constant Current Power Load"
 Lascar Electronics, Dual readout DPM: DPM702S
 Hageman, S., "A few added components make a self contained controller for a 100A load."
 Williams, Jim "Load-transient-response testing for voltage regulators", EDN September 28, 2006.