Handy Circuit Boards Save Time in the Lab

I do a lot in prototyping as part of my repair work on old circuit boards. Over the years, I have developed an array of circuit boards that come in handy when performing tests, whether I need to measure something or output a signal. Below are a few of my favorites.

Most of my testing is at low frequencies, so my primary connections are done using banana plug-to-banana plug test leads. Now, some of these may look crude, but the usefulness and the quick proto-typing is what were needed at the time.

The basic device is a simple trim pot. Nothing exciting on this, but it allows me to have a multi-turn pot put where I need one quickly:

The next circuit is a simple amplifier with an offset trim adjustment using the ADA4627. This circuit allows me to configure the amplifier as an inverter or non-inverter. Usage is up to ±15V supply:

Complexity gets stepped up a notch with this board. I am able to configure the SOT23-5 op-amp in any manner, plus have a trim-pot to help set the gain:

The next circuit is a bit short of a component, but was used to drive an input with a variable waveform at a known current value from 1μA to 100μA. This required a high-impedance differential op-amp set to measure the voltage across a known resistor value, R5. The missing component is a power amplifier, but in this case I was able to use a signal generator. Initial tests using a scope probe affected the current significantly, thus using the ADA4627 op-amps minimized any current being fed into the measuring circuit:

One program required measurement of MOSFET leakage current. This circuit allowed me to test either a SOT23, or a SC70 packaged device by applying a voltage and monitoring the output for abnormal drift that would indicate a high leakage current. I did find some bad MOSFETs, but after putting them through an oven for a few hours, the baking took out the moisture and the parts fell back into spec for the circuit:

One current project is working with a syncho/resolver measurement system that requires a high voltage reference voltage and a couple of other boards to clock in digital signals and read the digital result. Here is the current bench set-up:

The board on the right allows me to clock in digital signals where I can set the output logic level voltage from 3.3 Volts to 10 Volts (yes, I do work with some old logic boards at times). There are LED on the board to indicate a high or low rating. And the board allows me to drive up to 17 data lines.

The board on the left is a generic break-out board for a particular 70-pin connector. Several projects require this connector, but I have removed the polarizing pins to ensure it will work with all connector polarization variations.

The board in the upper left is a quad power-amplifier board (only one is stuffed in this set up). This board allows me to configure an LM3886 to accept any function generator input and then amplify the signal and provide up to a 1 Ampere current drive with output voltages swinging ±40 Volts. This acts as a basic linear amplifier, but is very useful when doing these quick proto tests.

I have others, but these are my mostly used boards. I may share others later. In the meantime, what type of useful test boards and test jigs do you use?

Related posts:

19 comments on “Handy Circuit Boards Save Time in the Lab

  1. eafpres
    June 7, 2013

    Hi Derek–did you fabricate the bare boards yourself that are shown in the first few pictures?  They look like DIY boards.

  2. Davidled
    June 7, 2013

    It looks like diameter is about 0.25 inch located each corner. Test point is very much impressive.  Is there any reason for a big testing point in board?

  3. Vishal Prajapati
    June 8, 2013

    Break out boards are very useful when prototyping the conceptual design.  This is one example of SMT to through hole board.

    Proto Boards are also useful for prototyping which contains only one device on board.

  4. Brad Albing
    June 10, 2013

    Derek – good idea – making your own boards to facilitate testing. I have made various test boards or test jigs to facilitate testing when I'm doing my own prototyping work. You've motivated me to write about it.

  5. Scott Elder
    June 10, 2013

    3-D Printing PCBs.  That seems like a very viable product rather than 3-d printing custom analog ICs.

    Have a bowl of molten solder that is written with a hot solder pencil.  Glue the parts on the board first so alignment problem is removed, take a picture, and then draw with a pen on the computer screen the wires.  Turn the solder writing pencil loose to lay down the molten solder wires.

    No mess with milling.

    Would this work?

  6. eafpres
    June 10, 2013

    Hi Scott–I'll try to refine your idea.  A spool of very fine solder wire with an automated feeder, a solder pencil on a robotic arm, drawing the traces and pads etc.

    The issue is bonding to the substrate–you would need something that solder wets just enough to get good consistency and reliability, but not too much that it spreads every where.

    Another way that has been done and I think fell out of favor once the little milling machines became widely available, was to use a sheet of copper backed with pressure sensitive adhensive (i.e., like copper tape but in a sheet), put into a machine that has a knife blade and cuts the pattern.  You then peel off the part you don't want and you have a circuit.

  7. Scott Elder
    June 10, 2013

    Perhaps just dispense solder paste and then a layer of sealant or glue on top of the board.  And then a heat blower to melt the paste under the sealer.  Could also do multi-layer.

    June 10, 2013

    The boards that have no solder mask were all made in my office. I used a 2D mill at one job, and that was very nice to have to cut the copper away. However, with some of the new packages, I have not seen it possible to get down to the super fine pitch required. The board pictured were done using toner transfer method. I use 15 mil pitch/spacing rule to get decent results. I have gone down to 10 mil a couple of times.

    June 10, 2013

    I would be willing to use it. Just looking at cost and pitch/spacing requirements.

    I have done the Xacto knife cutting on copper-clad board as well, but only for stuff with 100 mil leads.

    Ideally, getting to a print-type PCB would be better. There is work going on using ink-jet printers and conductive ink. There is some chemistry mixing required. Plus, the conduction of the traces is not as good as copper.

    But, if printer can be improved and then one can print one layer over another, the next challenge is just getting the alignment down. Run a piece of paper through a printer with the same print and notice the shift of the image. Not good enough for 25 mil pitched parts.

    June 10, 2013

    2nd board has no test point in the middle, it is a through-hole since a couple traces crossed. Trick here is layout to use a single-sided board. I could have used 0 Ohm resistors as jumpers, but too many traces to cross – thus a small wire on the back that is barely visible.

    4th board has an LM3886 not installed. The LM device was to be configured as a current driver to help set the current signal that was needed if higher than a few 10's of ma was needed, plus it would be able to drive beyond the voltage limitations of a basic function generator. Things were not working so well, so I ended up simplifiying the proto board. to have limitations from a function generator.

    My general test points in the corners are for banana connections and clip/scope leads. Keeping my options open as to how to interface to the rest of the world.

  11. Scott Elder
    June 10, 2013

    Derek, Could you please comment on the economics of prototype PCB work.  In other words, there are multiple ways to do this entirely by one person on the cheap.  Then there are milling machines.  And then there are overnight PCB services for $200 or so.  Where do you draw the line between each of these appoaches?

    And then assembly.  Build inside with stencils or simply subcon the three or four board run?

    As the saying goes, time is money.  So what level of sophistication pushes one outside versus DIY.



    June 10, 2013

    My metrics involve several factors – cost, time, free-time, etc.

    To start, milling would be a great idea, if I had one in the lab. It would be my preferred method in many cases.

    I would typically do a quick-turn, but as you noted, cost can be high, but the time is still 3 days minimum in my view – sending files, a day to manufacture, a day to ship, third day receive.

    My last option is the cheapest, but involves my time to make the board, but by far the fastest. Between toner transfer – getting the print correct. Then making the transfer look good. Then touch-up before etching. Etch time (about 15 min). Then plating time (about 10 min). The whole cycle can be from 2 hours out to 3 hours – depending if I need to redo the toner transfer portion because it did not look good before etching.

    I use either the second or third method. If I need to use more than 2 layers, or if I need solder mask, or if the two-layer requires many jumpers, I would use a fab house and bite the bullet for time and money.

    If I need only one layer (up to a few jumper wires), then third method is usually the clear winner. Only issue is that there is no solder mask involved.

    Another reason I would go with the fab house would be if the pitch is too fine, or if there are many parts with fine pitch – both of which would make toner transfer more challenging than I want to deal with.

  13. Brad Albing
    June 10, 2013

    @eafpres >> use a sheet of copper backed with pressure sensitive adhensive … I used to use various versions of this stuff – one version was marketed as CircuitStix. Very convenient – tho' moreso in the days of DIP packages.

  14. David Maciel Silva
    June 11, 2013

    It sure is a good technique because a module can be reused several times, on different projects.

    In many applications, knowing the risks utilize simulation tools, so I end up relying on a little faith that everything to be integrated on a printed circuit board, not a malfunction.

    For those who can invest a little more in rapid prototyping, a good option:

  15. RedDerek
    June 11, 2013

    When I was in the semiconductor industry, my company had an LPKF system. Very nice to use for the prototyping. But one would need to generate several boards to make it pay off; which we did as proto app boards before going to production.

  16. Scott Elder
    June 11, 2013

    Those PCB Milling machines seem like tweener products.  Either one needs a board cheap (DIY) or plan 1-2 days in advance.  If one ever had a hot-hot project, then there are always board shops that will jump immediately on your project for a high premium–but still less cost than a machine.

  17. Brad Albing
    June 11, 2013

    Still, it would be great to have one of the milling machines for those basement lab, home hobbyist projects. Maybe we can get Derek to buy one for us and then pay him a few bucks every time we want a board made.

  18. Brad Albing
    June 24, 2013

    Seems like the molten-solder-writing-pen would work in general, but you'd need a way to make sure that the solder stayed where you put it. That's a double problem – got to make sure it sticks to the board somehow; and got to make sure it doesn't flow away to places where it should not be.

  19. PCR
    June 30, 2013

    Thanks for the links Vishal, It was very informative. 

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.