Lessons from the trash can: Wrong power supply polarity.

In my previous post, Lessons from the trash can., I introduced the idea of creating the Planet Analog Lessons Learn Knowledge Collection . This time I will share a short story around what could probably be the most common mistake that waits out there for catching our fingers. This affair taught me more than one lesson, for this reason I have chosen to put it into context.

The wrong power supply polarity

This is as obvious, common, and easy to avoid as buffer-overruns in C language programs. But it is worth to mention it here, and the story behind it taught me several unforgettable lessons.

Two former co-workers were constructing one DC (12V) to AC (110V) converter. The original circuit was designed for using PNP germanium transistors and is shown in Figure 1.

Figure 1

DC/AC inverter [incomplete] schematic using PNP Germanium transistors.
Click here for larger image

DC/AC inverter [incomplete] schematic using PNP Germanium transistors.

My colleagues wanted to extract more output power and re-designed the circuit for using NPN transistor of type 2N3055. This transistor was very popular and easy to obtain at that time.

After powering the circuit, the result was no current, no buzz, no output voltage… and questions. I recommended this circuit to them based on my previous experiences. I had built one unit and it was working more than reasonably well, yet with very low efficiency.

All calculations they made were correct and the transformer construction was impeccable. The windings polarity and component values were also carefully checked. There was no apparent reason for it to fail to work.

At that point, my suggestion was simple: check the power supply polarity. But their answer was also straight and simple: the positive and negative cables from the power supply are connected to the nodes labeled +VCC and -VCC, respectively, in the circuit.

To their knowledge, it all was “undoubtedly” correct. My offer to serve as a third opinion and revise the assembled circuit was not exactly what they wanted to hear at that moment.

They quit after some time, checking and rechecking their notes and went for lunch. When they came back to the office, the inverter was working and powering one large fluorescent lamp. They asked me what was wrong with the circuit. “Nothing”, I said, “it was apparently related to some strange behavior of the electrons coming from the power supply, they were moving in the wrong direction with respect to the circuit.”

It is time now to speculate. Did you figure out why?

There are a few lessons we can extract from this affair. Failing to check the power supply polarity before powering our circuits can produce unpredictable results, or even smoke as it happens in most situations. We must assume checking our power supplies’ output voltages and polarities as an obligatory pre-flight task.

And probably of higher importance in this case is the need to be humble, and the absolute convenience of being open-minded.

Occasionally we overlook extremely simple details.

The top one is: Wrong power supply voltage

In the late 80’s we had one very skilled professor from the Electronics Department that was giving practices in the electronics lab using one big 0V-to-150V DC power supply. He configured the power supply to provide 5VDC for one unique and rare Zilog Z80 development kit. His very first comment before all his classes was check, check, check the power supply, its polarity and the rest of the wiring. But that day he was not aware that during the time elapsed between his first and second classes, another professor reconfigured and used the power supply. During the second class, during the first switch-on, the kit was destroyed instantly.

I thank this professor, and of course many others too, for the many lessons and advice given to us during our quest to become engineers. I also thank him, in particular, for his effort to educate us in one simple and essential principle that I have present every day, it is a Russian proverb that says:

“Measure seven times, cut once”

Once again, we must never fail to check the power supply voltage before proceeding to power our circuits. And furthermore, never take for granted that no one has altered the state of shared resources like power supplies, oscilloscopes and other tools and lab equipment in general.

Note: To be strictly honest, my former professors and colleagues also contributed to it, but I owe to my Father (“mi Viejo”), with no doubt of any kind about it, and up to its full extent, amongst many other things, every single bit of my passion for doing things and my love for my profession, and also one very important part of the person I am.

3 comments on “Lessons from the trash can: Wrong power supply polarity.

  1. antedeluvian
    September 15, 2017

    Victor, Back in the day the intel 8008 cost ~$400, when a dollar was worth a whole lot more. I remeber the lab assistant at university used to check every pin on the socket before inserting the micro. Later the Intel 8051 bondout chip would also set you back several hundred dollars if you powered the circuit up wrong, let alone applying the wrong voltage. Get burned once and you paid attention forever after. I still wish there were reasonably priced sockets to test voltages before inserting the micro.

  2. antedeluvian
    September 15, 2017

    Victor Sometimes checking is not enough- as you mention you need someone else to check. Looking at your own work, you are often blind to your mistakes. I have connected the mains to the output of a power supply on two separate occasions, once right before a demo at a customer.And I checked in both cases!

  3. Victor Lorenzo
    September 17, 2017

    Yes Aubrey, I agree with you. I think that during the first checks round we are very conscious of what we are doing and pay as much attention as we can. But as we look at the circuit for one second time, third time, and so on, we somehow start missing more and more details. This is the time at which we start needing some “fresh eyes”. Things going the wrong way during demonstrations and visits are also one of the classics. We used to call it “The visit effect”/”The visitors effect” at the CEETI, while I was working at the UCLV University. I've been in a couple embarrassing situations during visits to customers and one was my fault. We arrived late and in the hurry I forgot that I changed the board before leaving the office. I didn't upgrade the FPGA in which we had implemented the soft core processor. I spotted the situation almost inmediately at first system boot, but also the customer did, during this first boot the system was not doing what it was expected to do.

Leave a Reply

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