Cutting All Ties, Or How The DIY Isolation Transformer Enables Oscilloscope Probing of Bridge Diode And Other Rectifiers (This One Is A Doozy!!!)

Having accomplished my stated goal of eating my way through Thanksgiving, now thoroughly satiated, I am finding myself eager to get back to this blog post. One of the first things that comes to mind is why am I bothering with any of this? Oh, don’t get me wrong, I immensely enjoy the subject matter which I gather should constitute enough reason to continue with this focused study of radio and electronics, but…

I am sitting on the couch with a postprandial skyrocketing glucose level, having devoured a lovingly made smorgasbord starring corn beef and cabbage (yes, boiled dinner for Thanksgiving, a personal favorite). The day’s festivities have wound down, and I am now free to crawl back into my own mindscape. If there were only a way to combine roller skating, crocheting, and electronics, I would be all set. I even picked my ukulele back up this week (well, ukuleles…well…because…it would be odd if I had a hobby where I stopped at just one of something…) I am a hobby person. And a rather committed hobby person at that. But unlike going in circles with wheels strapped on my feet, tying yarn into seemingly endless strings of knots, or pretending to be a rock star whiling strumming three rhythmless chords on a tiny guitar, it is this electronics hobby of mine that I cannot figure out the “why?” of. I simply have no true worldly need to understand how an isolation transformer works or how to make one. Yet here I am.

I am working my way through “Laboratory Exercises for Floyd’s 10th Edition of Electronic Devices” by David Buchla and Steven Wetterling (Pearson, New York, 2018). The second chapter of both the main textbook and the lab companion have considerable focus on the basics of using diodes as rectifiers. You might recall that I somewhat tackled this topic already in this post. Buchla wants me to use a commercially built isolation transformer plugged into mains power in order to use an oscilloscope to measure the output of a full-wave bridge rectifier. If you read my post, you will see that an oscilloscope cannot actually be used directly to measure the output voltage of a bridge rectifier circuit that is fed with a signal generator, because the oscilloscope and signal generator (assuming both devices share an earth ground return) inappropriately short the rectifier circuit to earth ground, causing only a half-wave rectified output. Buchla’s solution is to utilize a 60Hz AC source from mains power to electrify the circuit, but through an isolation transformer that will in essence “float” the bridge rectifier. Once the bridge rectifier circuit is floating (i.e., no longer has a connection to a common ground reference, in this case earth ground), then it can be accurately probed with an oscilloscope because there will no longer be a pathway created between it and the signal source that shorts the circuit to earth ground. Whew, that’s a lot to wrap my head around, but I am getting there.

And along those lines, I am too ignorant of an electronics hobbyist to go connecting anything into a wall outlet directly. Fortunately, I do not have a mains power isolation transformer because I do not want to be even partially tempted to try it.

But I do have a FT240-31 toroid and 14awg magnet wire…

…And I do know a thing or two about transformers. Meaning I do not know three things about transformers…

---

Building the Isolation Transformer…

I decided to wrap 10 turns of wire around each half of the toroid. Arbitrarily, well, because I thought 10 turns would fit nicely on the pretty ferrite ring. The primary was just a coil of 10 windings. The secondary was also a coil of 10 windings, however, I additionally twisted the wires together at the 5th turn in order to create a center tap; I had noticed that Buchla’s isolation transformer had one. I scraped off the enamel from the ends of the primary coil, secondary coil, and the secondary center tap.

I tested my new isolation transformer by running a 60Hz 12V_PP sinusoidal signal through the primary coil from my generator. I measured the output of the secondary coil with my oscilloscope, and saw…an unaltered, horizontally-oriented, flat line at ground potential. In essence the same exact trace that I see when the oscilloscope probes exactly nothing. My isolation transformer did not work.

So, like any other incompetent, pretend design engineer, I twisted some knobs on the signal generator to see if anything would happen. Changing the voltage amplitude did nothing. I messed with the frequency instead. By the time I got into the kilohertz range, I started to see a bit of a wiggle show up in the output trace on the o-scope. By 50KHz, the output voltage was now showing as a fully recognizable sine wave. And by the time I got to a megahertz, the transformer output signal voltage began to match the input signal voltage. Now was I actually surprised that this transformer had a frequency dependent response? No, not really*. What did surprise me though is that the transformer seemed like it was going to be usable…even if at 1000 times the cycles per second that the lab instructions suggested.

[*In fact, Mix 31 type ferrite material has an operating frequency range of 1MHz to 300MHz. What this actually means for creating toroidal transformers out of it, I really cannot speak well to at this point. Perhaps a future blog…]

Below you can see the 50KHz 4.25V_pp signal as it is inputted into the primary windings of the isolation transformer on the top part of the screen. Below it is the inverted 50KHz 5.40V_pp signal outputted from the secondary windings of the transformer. Despite my not-quite best efforts, I have a slightly step-up votlage transformer (rather than a pure 1:1) due to poor construction. But the isolation component is effective, so I do not particularly care if the voltage is off for this use case.

I test the center tap of the transformer secondary windings (the red alligator clip on the twisted pair of wires).

And find that indeed, the voltage across one end of the transformer and the center tap is approximately half (2.85V_pp, bottom trace) that of the full output of the transformer (5.4V_pp, not shown). The top trace is still the 4.25V_pp input signal.

You can see my test equipment rack below. The function generator reads “50.0” designating 50kHz. Unfortunately, this device does not display voltage, which needs to be read off of the oscilloscope (i.e., the top trace which shows 4.25 V_PP as discussed earlier). The output of the secondary windings of the transformer is 5.4 V_PP , as shown in the bottom trace on the oscilloscope. Additionally, I measured an in-circuit current (I_RMS) of 395µA in series with a glowing red LED and a 470Ω resistor (circuit diagram below). Mathematically or practically I am not sure this low current measurement makes sense but it is what I measured.

The simulated circuit is shown below, along with the much higher predicted current of 1.629mA. Which interestingly is not enough current to light the LED in the model…

…unlike my identical IRL circuit below which clearly shows a lit LED. The centrally located black and red alligator clips are the in-series multimeter leads which displays 395µA (or 0.395mA) as shown above.

So, what the heck was the point of this entire exercise again?

---

DIODE RECTIFIER CIRCUITS

To study diode rectifier circuits! And be able to see the effects of these circuits with an oscilloscope. Remember the earth grounds of the oscilloscope and function generator are going to short circuit the diode rectifier circuit when they are used together without isolating (“floating”) the rectifier (i.e., circuit under test). This will cause full-wave rectifier circuits to look like half-wave rectifier circuits on the scope. Adding the isolation transformer means that none of the wires of the circuit under test are touching the wires fed by the signal generator, therefore the diode rectifier is not grounded back to earth. I now can just probe away with the oscilloscope! Yay!

---

Half-Wave Rectifier Circuit

Now the diode rectifiers! Starting with the half-wave rectifier, composed of a forward-biased diode in series with a load resistor. Also shown below is a capacitor in parallel with the load resistor in order to provide filtration to (i.e., “smooth out”) the output voltage of the rectifier.

Below is the unfiltered (i.e., no parallel capacitor) half-wave voltage rectifier circuit. The bottom trace is the output of the secondary windings of the transformer. The top trace is the output of the half-wave rectifier, changing all of the voltage to positive. Note that the frequency is the same for both traces. The capacitor would be used to “smooth out” the positive half-waves, so the voltage normalizes around a single positive value. An example of this is coming up.

The Multisim simulation of the half-wave rectifier circuit is shown below. Ignore the different values and pay attention to the waveforms. I moved settings on my function generator around to optimize the visualization of the waveforms, so the voltages may no longer match (for you very astute oscilloscope operators…)

---

Full-Wave Rectifier Circuit

Below is a breadboard version of a a full-wave rectifier composed of two diodes each connected to the ends of the secondary winding of the transformer. A load resistor spans between the node created by the negative terminals of diodes, and the center tap of the transformer.

An overhead view shows the full rectifier circuit and its placement with respect to the transformer. The oscilloscope probes are also both reference grounded through the transformer center tap.

The results of the full-wave rectifier with the oscilloscope probe across the output of the secondary winding of the isolation transformer to its center tap is shown on the bottom trace. The oscilloscope probe across the load resistor is shown on the top trace. There is now full wave rectification of the input signal through the rectifier. Note the characteristic doubling of frequency for a full-wave rectifier. All of the voltage is now positive even though this is difficult to ascertain from an old oscilloscope trace.

The simulation circuit is shown below, along with its model oscilloscope readout. It appears identical to my real world results.

Of course, a voltage that is direct current, meaning does not change from positive to negative unlike alternating current, is a useful goal of diode rectification. However, it is also not particularly helpful to have a voltage source with such wide undulations in amplitude particularly when the goal is to power up an electronic device. For this reason a capacitor is added to filter, or “smooth out” the resultant waveform so that a direct current with a more uniform amplitude is achieved. I demonstrate this with the 0.1µF capacitor added in parallel to the load resistor in the full-wave rectifier circuit shown below. (The circuit is on the left side…ignore the “spare” components on the right side.)

In the oscilloscope readout below, I have adjusted the display to only show a portion of the positive peaks of the waveforms associated with the secondary output of the transformer. I have then adjusted to the display to show the voltage across the load resistor which is now filtered by the additional capacitor. Ignore the horizontal dotted and dashed lines, they are there to help determine the voltage (in this case 70mV) of the “ripple” as the product of the filtered rectifier circuit is called. The ripple voltage is the sawtooth line with a frequency that is twice that of output of the secondary windings of the transformer (as would be expected since it is the product of a full rectifier). The capacitor changes the waveforms from having large fluctuations in positive voltage (as shown in the unfiltered trace above), towards a more consistent voltage. The higher the capacitance, the flatter the rectified voltage becomes. Any increase in load resistance, however, increases the amplitude of ripple voltage. I placed these traces on top of one another simply for convenience and ease of viewing (and hopefully understanding).

Indeed, if the same circuit is properly established in Multisim, the virtual oscilloscope can be adjusted to show an identical set of waveforms. I would just note that each trace below (unlike my real o-scope shown above) does utilize differing voltage scales simply for the purpose of being able to recreate a pattern similar to that of the breadboard circuit).

---

Full-Wave Bridge Rectifier Circuit

And finally the last of our full-wave circuits – the bridge rectifier. This circuit is a cleverly arranged square of diodes that forces alternating current to become direct current similar to the others I have shown.

Do you recall one of my recent posts on bridge rectifiers? And how I had to use the math function on the o-scope to piece together a waveform that resembled the output of a bridge rectifier? Well, now with the isolation transformer in place, this is no longer necessary! The input voltage of the bridge rectifier is shown below at 3.54 VAC_PP. Note, that this voltage is actually what is coming off of the function generator BEFORE it enters into the isolation transformer. I am not certain, and I did not test it, if you can actually use two probes at a time in the bridge generator without causing a short to ground.

However, the top trace shows the output voltage of the bridge circuit across the load resistor — this time directly probed from the rectifier circuit-under-test — and it spans an amplitude of 1.2V.

The simulation of the circuit is shown below. And in this case, Multisim does allow its virtual oscilloscope to take measurements across the secondary windings of the isolation transformer. I should have tried it with the real circuit when I had the chance…

Wooooowee! That was a LONNNGGGGGGGggggg one! But we have indeed reached the end of the line, and I need to head off to the roller rink. Although I am certain you are sick of reading about diode rectification at this point, for me, I have a really solid understanding of how a voltage regulator works starting to brew. Those ubiquitous bricks, and dongles, and wall warts–those black box devices that change magic outlet fairy dust into captured sub-microscopic Oompa-Loompas working in the charging factory inside my noise-cancelling headphones (my roller rink favorite!)–are actually starting to make a bit of sense to me.

Okay, hanging it up here, because unfortunately roller-skating sessions do not run on KM1NDY-time.

Toodle-Loo!
KM1NDY