in the Noise Floor. behind the youtube curtain

Howdy folks. Here we are nearly 2 years since I uploaded my first video - Happy Birthday! I started this YouTube channel as an experiment:

  1. To see how many people I could reach and share some knowledge
  2. To see if I could overcome my introversion, and if I could speak to an audience effectively
  3. To keep my tools sharpened by using them for each video
My 1st solo cake project. Coraline.

My 1st solo cake project. Coraline.

My wife had a custom cake business several years ago and she was so insanely busy, she barely slept. Ultimately she had to shut it down because there was little time left for anything else. I wondered at the start of TCS, would this be the same scenario? I set an arbitrary goal to see if it would be worth it. The goal was, reach thousands (really 2k) of people within two years and/or make enough money from the videos that the operation would be at the least, self-sustaining. So let's take a look at some metrics and see what we've drummed up for the past two years.

First and easiest to dissect are the financials or revenue. Let's look at the heaviest hitter when it comes to revenue. That would be #18 Unregulated Power Supply Design which I released on Oct. 19th 2016. As of today, it has 15,891 views and the primary reason that its one of the most viewed videos on TCS is because Hackaday picked it up as an article last year. So it's pretty easy to see that YouTube doesn't pay all that well when it comes to ad revenue, or I should probably say that what I expected is orders of magnitude less than reality. Other than those spikes, the remainder are video releases, activity usually lasting for only a day and then radio silence afterwards.

Best revenue, Hackaday article on video #18

Best revenue, Hackaday article on video #18

So before I bitch too much about YouTube, there has been steady growth, but you wouldn't notice with all of the ups and downs within the graph above. So here's a little chart showing monthly growth. You can see it peaked at a little over $12 in August!

Monthly Growth

Monthly Growth

Overall earnings in two years

Overall earnings in two years

 

The above earnings look solely at YouTube and do not consider those fine folks supporting my channel through Patreon (if you're curious, that produces $21/video at this time). So what about viewership? Well, let's look at the number of views, which other than subscribers, is the primary metric. That beautiful spike below is of course the Hackaday article. A few blips to the left is a mention on The Amp Hour, episode #300 so thanks Chris. You'll notice that a slight upward trend is going on around 1/29/17 and that is the point where I was releasing a video every two weeks. Notice that it starts to sag around 3/14/17? Thats about the time I started to release videos every month. YouTube channel Veritasium did a little video on this which is eye opening and I hope you give it a look. I am definitely not following this advice. I am seeing the same behavior though, interest on the day of release, but then... crickets. WTF YouTube?

Screen Shot 2017-09-20 at 11.17.54 AM.png

 

Now let's have a peek at average view duration or viewer retention. You can see from the start that some people watched the videos in their entirety, some just a little poke. As time goes on over the months, we're hitting an average view time of 4 minutes. Huh? My videos typically range from 5 minutes to 20 minutes. That tells me that folks are generally skipping over the majority of the content. I have to infer from this data that I must be doing something wrong. Either I babble on too much, the presentation or personality leaves something to be desired, or the content itself just doesn't hold anyone's interest. I have to ponder this one, but if you have an opinion, good or bad, leave it in the comments.

Screen Shot 2017-09-20 at 12.46.29 PM.png

 

Feedback, comments & thumbs. This has generally been positive. Being a small channel i don't see a lot of negative feedback or trolling. Of course I make mistakes. If you look hard enough at each video you're bound to see something amiss. For example, using darlington transistors in the "Driving Inductive Loads" video, I note Vbe as 0.7 volts though it should be 1.4 volts because of the series base-emitter junctions. It's bound to happen, and I would correct items like this, but alas, YouTube has removed annotations after the video has been uploaded. WTF YouTube? Anyway, you can't please them all and I'll admit a couple of videos are stinkers, but I keep them up because I won't sweep things under the rug.. as a result I get some dislikes. Not a ton of feedback from viewers, but enough to make some inferences - see below:

Screen Shot 2017-09-20 at 1.04.15 PM.png

I'm not fishing for compliments or attempting to stroke my ego, but I am confused why the positive comments and poor view duration. Personally if I like a video, I watch it all the way through. At least thats what I used to do before I started this gig. Now I'm lucky if I have enough time to watch anything at all. Yes, I'd like some cheese with my whine.

Screen Shot 2017-09-20 at 12.53.34 PM.png

 

To conclude, I love making these videos. It's fun because I get to tell a story, I get to keep my tools sharp and learn new things about shooting video, storytelling, continuity, editing. The list goes on and on. But it seems like the current format just isn't cutting the mustard. The experiment over these past two years is failing. I've reached some people and I am very grateful for that! However, the channel isn't self-sustaining in its current format. The revenue isn't "in the black", meaning more is going out than coming in. I was cool with that for a long time, but at a certain point where do you draw the line? That is to be determined I suppose. At this point, it seems that more technical videos and less builds is the roadmap. I don't want it to go that route, but YouTube has really hamstrung creators.

Beware newbie content creators.

Four Wire Kelvin Resistance Measurement

I can't afford a fancy benchtop multimeter with 4-wire measurement capability, so I made a current source and will show you how to make an accurate low-resistance measurement with this setup. There's a cool water-circuit that shows how current flows in a 4-wire system which turned out a lot cooler than I thought it would. Hope you guys enjoy! Here's a link to the Kelvin connection clips I mentioned in the video: http://www.mouser.com/ProductDetail/M... Hop on over to Teespring to support the channel!: https://teespring.com/stores/the-curr... You can send swag here! : The Current Source PO Box 620241 Oviedo, FL 32762-241 Follow more stuff here: @tcurrentsource www.thecurrentsource.com

Trigger a Migraine with Strobing Glasses

I get migraines quite often. In this video I try triggering my own migraine based on three things: 1. Dehydration, 2. Sleep Deprivation and 3. Flashing Light of a particular frequency. I show you how I made these flashing glasses and I run them at 60Hz, 200Hz and 20kHz to try and see if flickering lights trigger a migraine. Sound like fun?! Uh, nope.

Driving Inductive Loads via RPi or Microcontroller

Sometimes ya gotta drive a something that requires a little more current than the microprocessor can source. I made this silly device to notify me when my long-compiling code completes operation, and also to provide notification when new folks subscribe to my youtube channel. I’ve seen a bunch of tutorials that show how to protect driver transistors from high voltage when inductive loads are used, but sometimes they’re short on details or don’t hook it up to the scope. I built this nutty thing, so maybe it’s a good opportunity to show how all that stuff works.
BTW, this thing is totally not safe! Kids don’t make this at home.

Linear Variable Differential Transformer (LVDT) & Coefficient of Coupling

Howdy folks. I've been playing around with the idea of doing a video on LVDT's for a while now. I finally have the lock-in amplifier repaired, needed to measure the phase shift, so we're go for launch. I take a ball point pen and wind some coils around it (with the microprocessor controlled coil winder, a few vids back), use a ferrite core mounted on the ink shaft and it works like a charm. I also cover how to measure the coefficient of coupling with my crappy LCR meter. Worked out ok. Anyway, I hope you guys learn something or can use and LVDT for one of your projects. If you do, let me know. I'd love to hear about it!!!

Class D Guitar Amplifier (Battery Powered)

What's more fun than combining music & electronics? Not much to say here; I made a custom class-d amplifier for guitar. It has built-in distortion similar to the circuit we did for the Bass Ackwards guitar pedal. I also make a custom speaker/enclosure. Came out ok, but I still need to do some tweaking. What a fun project!
Should I make a kit out of it?

Using an I/O Expander IC

I've been working on a project which has been eating up all of my time and I thought I'd share part of it with you. In this video I'll show you how I connected an IO expander to a bluetooth module and use I2C to communicate with it. There's a lot of footage that didn't make the cut, mostly me banging my head on the desk trying to get I2C to work. There was some pretty delicate work soldering magnet wire to the IO expander chip, but unfortunately I don't have a good microscope solution and that footage didn't make it either. We'll have to fix that.

Microcontroller Based Coil Winder

So I need to wind some coils for an upcoming video and they need to be fairly closely matched in impedance and number of turns. In order to do that I rigged up this microcontroller based coil winder. I go through the overall theory of operation, a close look at the hardware and firmware involved. At the end you can watch the rotting fruits of our labor.

Quick Announcement - Germany + New PO Box

Just a quick announcement - traveling to Germany next month to work on a project and I've got a new PO box, so send your stickers in for the whiteboard! I'll give you a shout out!
 

DIY Stomp Build + Analysis

Howdy folks. Today I pretend like I'm an analog expert! I went and built a DIY distortion pedal or stomp box as it may be more widely known. I thought it would make a good video, and a little more fun than the F5 bias downer from last month. In this video we inject a test signal into the pedal and we look at how it distorts on the oscilloscope & look at harmonics in the FFT plot. Being based on the ProCo Rat pedal from the 80's, I was curious to see what the filter response was so we do some Bode plots with the dynamic signal analyzer. Good Fun!
(I accidentally swapped the in/out jacks - usually input is on the right with guitar pedals. This is where the name "Bass Ackwards" comes from.)

Class-A Amplifier Build & Fake JFETs

Unfortunately I wasn't able to accomplish the goal of completing this amplifier in the time I had alotted. My plan was to build it up and take some THD measurements and show how to bias this (The F5 Class-A) amplifier designed by Nelson Pass. I ordered some sketchy JFETs from fleaBay and yup, sure enough they were fake, couldn't bias a damn thing. So I'll have to make a follow up video once I get the real JFETs in. Sorry folks, but at least it should be an interesting build!

Unregulated Power Supply Design

I thought it would be interesting to do a theory video on linear power supply design. I build up a simple low voltage supply to setup the framework for the larger and more powerful split rail linear power supply used by the (Nelson Pass) PassLabs F5 amplifier that’ll show up over the next couple of videos. I begin building up the F5 supply and we compare ripple voltage between a switch mode supply and linear supplies. I also cover how to select the proper filter capacitor with some basic algebra.
I went a little nuts with GarageBand an made some techno music to go along with the time lapse build. Not my cup of tea, but it was fun to make anyway! Hope you guys enjoy this LONGER video.

Biasing Transistor Amplifiers

I thought it'd be fun to do some theory on my method of biasing a class A BJT transistor amplifier. I also show classes A, AB, C & D. Most of the time is spent on class A, but it gives you an idea on how to calculate a reasonable quiescent point with a predetermined load. All circuits are bread-boarded with the exception of a 2.5W class D amplifier which I demonstrate with an Adafruit PWM breakout board. 

Doppler Shift Experiment

This time we perform a little experiment using a switched quarter wavelength antenna array to create our own doppler shift. We then use a spectrum analyzer in zero span FM detect mode to listen to this oscillating doppler shift. I show how it was constructed with some lovely background music - loosely based on Francisco Tárrega's "Maria". I'm sure he's rolling over in his grave. 

Hall Effect Experiment

I decided it would be interesting to cover the hall effect. So in this video I show you a failed hall effect experiment, then we build a working test rig using gold foil - just as Mr. Edwin Hall did over at Johns Hopkins back in the day. We use a power supply to push current through the foil and measure the hall voltage with a Fluke 883A differential voltmeter. I also demonstrate an off the shelf hall effect sensor.

Measure an Unknown Capacitor

In today's video we take a quick look at what a capacitor is, how it is constructed and how to measure an unknown capacitor value. I ran across a double sided copper clad blank PCB and was curious about its capacitance value. We take a look at four different methods to determine an unknown value: 1. By measuring its time constant with a function generator and oscilloscope, 2. Configured as a low pass filter and measuring its 3dB down point with a dynamic signal analyzer, 3. Calculating  the value based on physical dimensions, and 4. By using a low cost LCR meter. At the end we blow apart a small length of solder with a 75uF capacitor charged up to 1.5kV!

Silicon Controlled Rectifier & the Crowbar Circuit

In this video we take a look at the basics of the Silicon Controlled Rectifier, or SCR. There’s also a little history about the Shockley diode and its evolution into the four layer device with a gate terminal, better known as the SCR. Different methods of latching-on the device are covered, and methods to turn the device off. We build a little tabletop demonstration to show how to actually turn on and off.
This is also the second video in which I recorded my own music - so please give a thumbs up if you enjoyed it. Thanks for watching and please subscribe! 

Headphone / Speaker AB Switch Aliens Style!

Today is the easiest build yet that anyone can do with a handful of parts and a soldering iron. It is a Sci-Fi themed AB switch that gives you the ability to switch between your headphones and external speakers, useful for music enthusiasts, gaming rigs, recording studios and work-from-home parents who stay up until 2am working on editing videos and don't want to wake their kids up with the subwoofer. All joking aside, I love this thing. No more dropping cables behind my desk as I switch plugs on the rear of my iMac - it stays plugged in all the time!

Stepper Motor Intro and Trapeziodal Velocity Profile

In today's video I give a basic introduction to stepper motors, what's inside and what makes them tick. If you've ever used a stepper motor and want to control it in a more advanced fashion than simply pulsing on and off, we take a look at a trapezoidal velocity profile to smoothly accelerate and decelerate from start to endpoint. Some discussion is also provided about why one would want to do such a thing. A stepper motor is connected to a linear actuator and we go over the basic calculations on how to determine the number of steps to travel a specific distance.
This  project is available on OshPark and source code can be downloaded from GitHub.

Surface Mount Soldering Walkthrough

This is kind of a part 2 soldering video covering surface mount technology. The main idea is that it's actually easier than through hole soldering - really, it is! As long as you have a steady hand, a good temperature controlled soldering iron and thin enough solder, all you need is a little practice and perseverance and those pesky QA folks won't be able to discern between you and a pick and place machine. Good luck!