I know, I got a very non working Phillips power supply, and, after a bit of head scratching, I did eventually reverse engineer what it does, and how to fix what was wrong with it. Dead capacitors, dead diodes and totally missing power output stage, I got it back in working order. Nothing really spectacular, basically an emitter follower running off a potential divider, with current limiting shunting the zener diode to drop the voltage. Used a pair of other boards I got at the flea market, some old Phillips PMR unit power supply heatsinks, as the output stage, as the boards each had 2 power transistors mounted, along with standoffs, and emitter resistors for the transistors, along with some unused SCR devices, which turned into connection points. Basic Darlington connection, on the one board, and the other as power output stage, and the flea market also supplied some 2200uF 35V capacitors to replace the very unhappy MysterLee caps the board had, old repair. Some 1N5403 diodes for the power side (very overkill for a power supply of 24VDC and 500mA, but better than the very sad looking originals), and it works. current meter is a little sticky, but voltage does work, and now just have to find the missing hardware to close the case again. Not bad for something that in total cost me $2 for the non worker, $1 for the heatsinks with parts, and $1 again for a box of NOS Hitano 85C caps.
You are who you are and I learn sometimes from the "small talk"! Also, if I can't watch, I'll watch it later or skip, but bee yourself, just like you said! Thank you for your videos!
Driving the coil is quite easy, you just use a set of outputs, and a series resistor to limit current. both start off set low, and then you alternate driving each high for a short period, so that the coil sees an AC drive, but the outputs are basically shorted when low, removing the back EMF pulse. With the stepper motors you can reduce the current by using a series resistor in the common power line for 5 wire ones, or in one of the phase leads for 4 wire ones. You lose a lot of torque, but lower current, lower noise, and you have limits on the ramp up speed before it starts to lose steps, but for your use with only a low speed drive should work well. Plastic sleeve as gear will work, and will eventually become a gear where it contacts the steel gear. Code unroll loops, should be plenty of room in there to not have critical stuff in a loop, so you get rid of a lot of the overhead. Just be careful your compiler does not decide to roll them up again to save space.
Martin, if this is who I think it is, they should be open on Monday, as the Lockdown here in South Africa is being relaxed slightly this Friday, which is a public holiday, so they probably will at least be able to get into their offices.
I really enjoy your T4D's! If a viewer doesn't appreciate your time and effort, they should go somewhere else! And your "rant" about "doing it wrong" is SO true!!! After all, how does a person learn?!?! from their mistakes and experiences! And for you to share these "errors" makes your videos so much more educational and informative! It shows that you are willing to express your errors so that we can learn! I would imagine that those that claim "your doing it wrong" have made MANY mistakes themselves, but won't admit it because they are too "small" to do so! It would be so much better to suggest a method, or something to try, rather than just blatantly accuse you of "doing it wrong". After all, they could be alot more helpful, or just go away! Using a stepper motor to drive the clock is a great idea since the inductive kickback and polarity switching complexity with the original motor creates a set of problems. Just wondering how accurate the "friction" drive is going to be after a while though, and if the bearings can handle, long term, the side load from the friction drive. You stated that problem so it will be interesting to see! Thank you so much for taking the time to show your progress with the clock project! It is interesting!
Thank you for the video (including "rant" and "ramble"), very interesting project. I would definitely have to explore driving the original "stepper motor" as the "ticking" of your current set-up would drive me nuts - and I like mechanical clocks! Using direct port manipulation on Arduinos is one way to speed up execution time as a digitalWrite HIGH takes 3.95 uS and a digitalWrite LOW takes 4.55 uS
The noise of the "ticking" stepper was something I forgot to chat about. There's no isolation between the stepper motor and "speak box" / acrylic case of the clock, which amplifies the sound. I will correct this in the next iteration. Thanks for your post.
Martin I love your statement starting around 4:45. You and several other intelligent sharers have spurred me into constructing a few upcoming tutorials for blog and youtube, and the fear of making mistakes was definitely getting in my way, at first. Such an understated reminder.
