I more chuckled about that hack because my brain was already busy thinking about generating a clean 50Hz sine wave. Then she had a much simpler solution based on the knowledge of the inner workings of that chip.
not sure what technology will be like in 40 years, but some future nerd will respond to their audience ... i am using this vintage Arduino, because that is how we roll...
Yeah, I don't doubt it. XD You get that in several areas. I'm working on programming a SNES, which is 30 years old at this point, and for good measure an Atari 800 XL, which is from 1983, but is a revision of a 1979 microcomputer design. That's a different field, but you could (and do), run into people asking the exact same question. What's the point? Why not use a modern system instead? Well... I use these old systems because that's what I feel like doing. XD There's nothing else to it.
Just like me with my PDP-8/E -- it's about 42 years old now. I use it (and sometimes emulators/simulators when it breaks down) because that's what I feel like doing! and I plan to get my Teletype & paper tape working for programming & saving what I write. Eventually, I hope to get my RK05 (removable platter drive) working with it too. :)
I've just spent the last 4-5 years making a ZX Spectrum Laptop using original parts! (Apart from the screen and case). But it was fun finding the bits & getting it all working! (I had no internet access till 4 months ago). Putting a modern kit with emulator would be too quick, easy & boring!
LOL, I thought I was the only one who's bench looks the same as yours...and it carries over to the "Future" projects bench and the couch in my lab... Fran Always a treat watching your videos.
I'm glad to see that there is still someone out there who doesn't answer every situation with Arduino. It's got to the point where I have seen people use a microcontroller to get a LED to switch on when you press a button.
Very nice project! I always knew that the clock chips count the AC line frequency - but the method of muting certain pulses and having the chip work accurately even though it's not an even frequency is very good thinking!
No fun just throwing a cut and paste microcontroller. Awesome vintage projects Fran. Your channel is my favorite. I love learning about the vintage stuff it’s more of an art form designing a logic circuit versus slapping an Arduino in their. Keep it up!!!! 👍👍👍👍
Sometimes the simpler solutions rob you of a learning opportunity. Besides that, it's also nice to get things working in different ways. There's more of a talking point for that circuit than there is for saying I just used a micro-controller, that would make it kind of boring :-)
Agreed! Sometimes, we must work with what we have. I've done it many times. I applaud your efforts. I call this mind solutions. It's most certainly worth the effort. Great job!
We used to have a synchronous electric clock. I really liked it. It seemed to keep near perfect time from the 60Hz AC line frequency. Now everyone seems to use quartz for some reason.
Mixed feelings here. I used a 50252 chip to make an alarm clock in school, mostly as an exercise in making circuit boards. So some nostalgia that only nerds would understand, but I wouldn't be interested in working with it again. That clock got me to class on time many mornings. Thanks.
Fran, excellent solution. Thx for sharing. A PLL f multiplier was my first thought. Your use of a "Pulse Swallowing" method was genius. Some programmable counters use "Pulse Swallowing" methodologies. Your solution was simple, elegant, practical, accurate & functional. A monostable multivibrator using say a 555 could've been an alternative if you didn't have a 4017. That approach might've used more components though. I call it a junkbox solution, that is, whatever you've got available to avail a solution. Great job.
In my opinion that pulse dropping circuit is exceedingly clever. I confess I would've set about building a power inverter around a frequency standard, confident it was the only way to go.
What a fascinating design decision to make a display change based on current cycles. Thanks for following up with the information that they made the other two versions that were fixed display modes. I was immediately curious about that. I suppose most people ordered what they needed for their design. When you got to the end and asked yourself the rhetorical questions about why one would do it this way. I certainly wondered if anyone from that era was faced with the same problem of getting the cycle switching display to show 24-hour time with 60 Hz. Maybe that never actually happened, but it reminds me that for people in those situations, where you have a chip/thing and you have to "make it work", there are no rhetorical questions. Thanks for sharing your solution.
Sometimes we do the things we do because we CAN! My 13 year old as asked me why I weld stuff back together like a broken chair or a broken tool. Answer: because I can, and it’s fun to exercise your skills!
did it the exact opposite way to get a 60Hz clock chip to run on 50Hz mains. just looked in the spares box and found a NTSC colour burst crystal and made a single gate oscillator to get the output, then used some TTL dividers and counters to get a 120Hz clock out of it, then used a divide by 2 to get the final 60Hz clock to drive the unit. Stuck it all on a small piece of protoboard and crammed it in the unit, along with a small hole to get to the trimming capacitor for the oscillator. Worked out to drift around a minute a month, perfectly good enough for my application, and all made with parts salvaged from scrapped electronic parts. Did a similar thing to run a 400Hz 26VAC hour meter, just used a 555 timer, a 2N2219A and a small audio transformer, with a tuned secondary, to get around 22VAC to drive the counter. did around 20k hours before the hour meter itself wore out mechanically.
Nice elegant solution, Fran! I tried to guess at the solution before I watched the video, and the thumbnail offered up a clue, but I was slightly off. My idea was to use the 4017 as you have, but have Q0 to Q4 feeding the mostek chip via small signal diodes. Q6 would be tied to reset, giving the missing pulse every 6 pulses.
I've studied the circuit and finally I've understud how it Works. The counter pull down a pulse every 6 so the pulses are 50 in a second even if they are not equal.
The nice thing about a messy lab is every now and then you find something unexpected in the clutter which triggers you off at a tangent and a whole new universe opens
Woot, it's Fran! In the Lab! Here's a top tip for anyone worried that their lab is filthy: Call it a workshop instead. Problem here though, is that FranWorkshop wouldn't kinda scan the same. Hmm! Hoping that the next evolution of the circuit just uses transistors, no chips.
What would you do if it was the other way around (running a 60 Hz clock in Europe)? Build a crystal oscillator and devide the frequency down to 60? Build a PLL circuit? Use the 50 Hz to make a resonant circuit oscillate at 300 and devide that frequency by 5? All of that sounds more complex/difficult.
Nice trick to fool the Mostek! I've had to produce clock signals in some of my projects, and I always go straight to a relaxation oscillator. I imagine that would work about the same, but your trick to 'skip a beat' is pretty neat. Shows how dumb chips are that they don't really care the time between pulses. +1 on the quickies!
My alarm clock still is one that I built using a kit around 1980. Because its case is badly twisted plastic, held together with Scotch tape, I am reluctant to open it for checking the chip. but I think it was a Mostek. Unlike yours, it runs on a crystal - I remember I could tweak the frequency. And it is in 24 hour format. I made two modifications to the kit, I did not like the 35 V supply that I felt was scary. I reduced it to something between 20 and 25 V. On the other hand, I considered its bulk capacitor inadequate, so I added a bigger one, just hanging there behind the PCB. With that, my alarm clock survives nearly 10 second long power outages without losing its count. I was puzzled by the high operating voltage, but believe now that it was due to the chip being built on P-channel FETs. I’ll check it, if the clock ever fails so that can justify opening the warped case.
AC pwr line frequency in the US prior to 1935 wcould be 50 hertz or 60 depending on what part of the country you lived in. The grids were all synchronized after that date to keep the street clocks on time
I though of a cycle swallow, or maybe faking it with a crystal oscillator and divider - but on the other hand, the mains frequency is actually specified to be long term accurate …. The main problem with mains driven digital clocks as opposed to synchronous motor clocks is that interference can add extra clock pulses
Awesome workaround to get 50Hz. I can totally relate to wanting to use classic chips- in the late 90s I used to scour thrift shops for a certain 70s Radio Shack project kit that had a (very simple) integrated circuit on board, which I would cannibalize to build the light controlled oscillators I'd use in tandem with my circuit bent SK-1's... (SK-1 pulse code driving LEDs coupled with CdS cells modulating oscillators built with said chips) People just didn't "get" why that was the chip I wanted, they'd say "just use a 555!" But there was something about hunting down those 150-in-1 kits and extracting their precious parts... ahh, the memories. --I think I still have a couple of those kits floating around somewhere!
Hello, did you succeed?, btw i am in a simillar problem, i have an old general electric radio alarm clock from canada, but i live in europe(50hz), so time slows down, is there perhaps any solution for me? Thanks in advanced
Great solution, I think they call that a 'swallow counter' since it swallows a pulse every so often. Note that the 4000 series chips like your 4017 have a maximum supply voltage rating of 15v but 16 probably won't kill it!
That's a proper lab that is! It's not a proper lab unless you have piles of projects half way through!!! Also wouldn't a ULN2803 or similar have worked for the lowside transistors? As the ULN2803 has been around since Pontias was still a pilot it would not be breaking with the vintage concept.
when i started to watch the video i was thinking the ways i would consider changing the freq'y, there was 2 ways, one was using a 555 timer to power a fast switching high power transistor, the other was using a decade counter and taking the pulse off 5 of the 10 but that still required to be driven by a timer chip... the reason why i thought like this is because id more likely run it off a set of batterys and even with a 100LED if you drive them at 3mw they should be bright enough and that would be less than 300mw of power as not all the LEDs would be on! i didnt consider running it off mains voltage and tapping off the internal freq'y like you have ... you could use a 10mhz clocking xtal and a x5 multiplier, again good for battery use
It looks to me you might have saved some effort by running the "23" volt transformer as a bridge rectified center tapped for 11.5 volt (close enough to 12), or was 16 volts a requirement?
I'm often amused by the "youngins" who like to throw an arduino or pic at almost anything. I'll read what they're doing and think "you could do that with a 555..."
Ted Thompson: Spot on, but I can just see the responses from some people... "555? Pah! What's wrong with a combination of basic ICs?" "ICs? Pah! What's wrong with discrete transistors?" "Discrete transistors? Pah! What's wrong with good ole vacuum tubes?" "Vacuum tubes?" Pah!..." Etc. ad nauseum... #o)
To be fair, if you're just throwing some prototype together, an arduino is great. Of course, if you want to make more than just that, you can use something that fits better. Sometimes, a microcontroller IS what you want, though.
I think a 556 would be better for this. Make one half time out in about 95ms, and make the other oscillate at about 50Hz. Feed the output from the first half into the reset of the second and every fifth pulse will be lengthened or shortened so as to maintain an average speed of 50Hz.
If I just wanted to throw an Arduino in there... what would be the point, you know? There's a million clocks out there that would be made like that. No. I'm using a vintage.... ... that's how we roll in the Fran Lab" - and that's why I love this channel so much!
Well remember - a cluttered desk is a sign of a dynamic mind. And I can do the 24 hour times in my head. Right now it's 19:19 hours here. Reason is I worked for a company with clients all over the world. And it's much easier to use a 24 hour clock than the 12 hour clock.
Fran, if the PM is lit I would have added 12 to the hours digits? If PM High and D6 high add 1. If PM high and D5 high add 2. It would save messing around with the frequency. The 4017 and transistor is ingenious :D I hope that helps. Say Hi to Missy from England :D (Edit I forgot to say Hi to Missy)
The output of the clock chip isn't binary, it's already been decoded to 7 segments and enable outputs for the six digits, so it would have to be converted back to binary first to do the addition.
I once built clock wich needed one impulse per minute (to move the minute hand in a slave station clock). Did that in a very similar way by cascading two counters.. About 35 years back... LOL
Great video Fran, like the short videos. The timer solution is nice for the 50Hz but how do you get the 4017 to work? I have done short video on using this thing but for the life of me I can not get it to work, so a quikie on getting the CD4017 / HCF4017 working would be nice hopefully showing the pitfalls as I'm falling down all of them. Thanks.
Hi Fran, Hope you are still reading comments from this somewhat old video... Fran, I need your help. Being that I live in South America (Chile) where the AC mains standard is 220V @ 50Hz, I need a circuit the opposite of yours, with an output of 110V with a AC frequency of 60 cycles. Before I continue on this endeavor, your opinion on how to proceed would be very helpful. A) Is it possible to use your schematic as a road map for modifications? B) Bite the bullet and just purchase some commercial converter device? C) An Arduino (MCU) programming solution? (yuck) Regarding your comment about using a micro controller solution (I agree with you, it's like cheating) versus an old school component solution... your singing to the choir. I much prefer the latter. Thanks again, Ken
one problem i see is, her circuit drops pulses, while your circuit would need to *add* pulses. you could create a simple 555 timer oscillator that runs some safe margin higher than 60Hz, say, 100Hz (to leave plenty of room for that RC oscillator to drift). Have a counter circuit watching the input, activating the oscillator after 5 input pulses, and have a second counter circuit watch the output and deactivate the oscillator after 6 output pulses.
@@spambot7110 Thanks! I never considered adding pulses beyond spec and then correct "backwards" by dropping pulses as needed. Thanks for the idea (My reaction to your solution was similar to Scotty's reaction hear: ruclips.net/video/jxzYTDX8bIg/видео.html
Any suggestions for a circuit to generate 50Hz or 60Hz without access to mains, and ideally without a microcontroller or crystal, but still fairly accurate? 555 style timer circuit is probably not accurate enough. Doesn't need long term stability, but over a few hours use, should keep reasonable time.
The MM5369 was a great chip for this. Little 8 pin DIP that would generate accurate 60 Hz pulses using a cheap and easily available 3.579545 MHz TV colorburst crystal. Unfortunately discontinued, but you might find NOS on eBay or surplus sellers.
Dave Curran There have been clock chips using a RC oscillator as fallback in case of mains blackout; the Fairchild FM7010 as used in Grundig's Sonoclock 21 would come to my mind here. But in real life the accuracy was more than lousy; even after just minutes of blackout it would already deviate a minute or two - no matter however often you calibrated the osc on it. Change in ambient temperature alone would throw it off. An old school approach for a 50Hz timebase using still available standard components could include a CMOS 4060 4027 and a 3,2768MHz xtal.
Thanks for the replies, yes, that is what I had thought, just trying to find a solution for a client of mine who has some contradictory requirements for size / cost / accuracy.
So glad you did not use an Arduino or similar. Much more fun this way :-) Interesting chip. I have been thinking for a while now about putting together a clock using 50Hz power in the UK. I wanted to add the radio clock signal update as well. (en.wikipedia.org/wiki/Time_from_NPL_(MSF) ) . Assuming the frequency balancing in the UK is good then it should keep relatively good time, but thought a quick update from the radio signal would not go amiss - Besides - more fun. I love you 3d printed displays so was thinking to use those too. My clock might take a tad longer do though... I need to learn more.
Fran, how would a 50Hz vibrator valve (tube) go in being the clock driver for your display? I have no idea whether a vibrator is even appropriate in this case, but as a Just-For-Fun project it may be an idea for a future vid. ɐᴉlɐɹʇsn∀ ɯoɹɟ sɹǝǝɥƆ
I found an old clock chip mm5312 and I did the same thing using pic12f675 to generate 50hz or 60hz square wave, if you are interested i can send it to you the code.
Nice solution, but one thing to remember here; what OPAMP are u using? It surely is not an open collector output, and with some Opamps I would not feel really good shorting its output to gnd and just counting on its internal output current limiting to save it. Secondly as far as I see your approach only works 100% if the delay in the signal thru the buffer opamp is greater than the delay of the 4017 counter to turn on the transistor. If the signal thru the buffer opamp is faster than the blanking pulse (or it is a toss up) you may (occasionally) get glitches at the start of the blanking interval where a bit of the 6th pulse slips thru. Maybe too small a glitch for the slow Clockchip to act on it, but maybe so. Guess only long term stability would tell. Keep up the work.
Also the drawing of the schmatic seems to be a ltittle weird. The link between Pins 5 and 15 on the 4017 provides the reset pulse on reaching the 7th count (which seems counterintuitive at first but is correct). Still that pulse is only available for the few microseconds that the 4017 takes to reset count. The output to the green LED and the base of the blanking transistor must be tied to any regular count before that. And in the video at 5:05 the closeup of the PCB shows clearly that the green LED and 10k to the base of the blanking transistor are in reality tied to pin 10 of the 4017 (active on 5th pulse). But maybe that misdrawing was the intentional error aimed at chinese copycats - if they would ever design with obsolete MOS clock chips.
Old fashioned digital, I think I commented in G+. :) There should be some old style way to multiply 60Hz by 5, then divided it by 6, and end up with 50hz.
In 1974, at age 10, I built a Heathkit GC-1005 6-digit clock. I still have that clock and it works. This is based on the Mostek 5017 chip. It is switch configurable between 24/12 hour mode. It is 60Hz line frequency based. The power grid 60Hz does not seem as "on time" as it was 44 years ago. Back then, I could set this clock to the telephone "talking clock", or the CBS "Bing, da da da, da da" top of the hour jingle, and it would always be within a few seconds of real time. I built a few years ago a 6 digit Nixie Tube clock with the WiFi "NTP" clock sync. Now I easily compare "real time" to AC line time. I've seen AC time drift as much as 30 seconds. BTW: You can call WWV/WWVH on the phone: (303) 499-7111 / (808) 335-4363. Talking clock is not dead, yet. The Mostek 5017 has a silicon Easter egg on the die of Mickey Mouse pointing his arms to 7 o'clock. How cool is that? You can see an image of this here: www.flickr.com/photos/dr-engineer-001/52647186
What a novel 60 to 50Hz converter!, I would probably have used an Arduino nano to synthesize a continuous 50Hz (no gaps), totally overkill for your use. But hang on, I live in Australia, a 50Hz country, so going the other way, without a micro, could prove an interesting challenge.
That clever 'pulse loser' circuit makes for a nice elegant solution.
I more chuckled about that hack because my brain was already busy thinking about generating a clean 50Hz sine wave. Then she had a much simpler solution based on the knowledge of the inner workings of that chip.
It's clever, not elegant. Elegant would make the pulses equidistant, albeit with no functional gain.
This is the style I love about you. So happy to see it surface again!
not sure what technology will be like in 40 years, but some future nerd will respond to their audience ... i am using this vintage Arduino, because that is how we roll...
hahahah
Yeah, I don't doubt it. XD
You get that in several areas.
I'm working on programming a SNES, which is 30 years old at this point, and for good measure an Atari 800 XL, which is from 1983, but is a revision of a 1979 microcomputer design.
That's a different field, but you could (and do), run into people asking the exact same question.
What's the point? Why not use a modern system instead?
Well... I use these old systems because that's what I feel like doing. XD
There's nothing else to it.
Just like me with my PDP-8/E -- it's about 42 years old now. I use it (and sometimes emulators/simulators when it breaks down) because that's what I feel like doing! and I plan to get my Teletype & paper tape working for programming & saving what I write. Eventually, I hope to get my RK05 (removable platter drive) working with it too. :)
I've just spent the last 4-5 years making a ZX Spectrum Laptop using original parts! (Apart from the screen and case). But it was fun finding the bits & getting it all working! (I had no internet access till 4 months ago). Putting a modern kit with emulator would be too quick, easy & boring!
"That's how we roll in the FranLab"
Wouldn't have it any other way!
LOL, I thought I was the only one who's bench looks the same as yours...and it carries over to the "Future" projects bench and the couch in my lab... Fran Always a treat watching your videos.
I'm glad to see that there is still someone out there who doesn't answer every situation with Arduino.
It's got to the point where I have seen people use a microcontroller to get a LED to switch on when you press a button.
Nice to see you Your channel Fran, I like the vintage projects
Nice to see you again. I like the kind of challenges and the way you resolve these.
Very nice project! I always knew that the clock chips count the AC line frequency - but the method of muting certain pulses and having the chip work accurately even though it's not an even frequency is very good thinking!
I built a clock using this chip back in 1976 and its still working as my bedside clock today.
No fun just throwing a cut and paste microcontroller. Awesome vintage projects Fran. Your channel is my favorite. I love learning about the vintage stuff it’s more of an art form designing a logic circuit versus slapping an Arduino in their. Keep it up!!!! 👍👍👍👍
I enjoyed your quickie Fran, now I'm breathless. keep them comming.
The way you do things, the Frean Lab way, is why I subscribed to your channel. I love it.
That's pretty ingenious. What a completely wild solution. I was expecting a discrete PLL or something like that.
Sometimes the simpler solutions rob you of a learning opportunity. Besides that, it's also nice to get things working in different ways. There's more of a talking point for that circuit than there is for saying I just used a micro-controller, that would make it kind of boring :-)
Your bench looks like my computer desk.... yes, dig deep and find eras of past attempts and issues... well... at least they're still there. Fran!!!
Nice! I like the way Fran rolls.
I admire your clever solution to the frequency conversion. Kudos!
As for the messy lab, "Creative minds are seldom tidy".
Agreed! Sometimes, we must work with what we have. I've done it many times. I applaud your efforts. I call this mind solutions. It's most certainly worth the effort. Great job!
We used to have a synchronous electric clock. I really liked it. It seemed to keep near perfect time from the 60Hz AC line frequency. Now everyone seems to use quartz for some reason.
Sometimes the fun is as much in the journey as the destination, I totally get it!
Mixed feelings here. I used a 50252 chip to make an alarm clock in school, mostly as an exercise in making circuit boards. So some nostalgia that only nerds would understand, but I wouldn't be interested in working with it again.
That clock got me to class on time many mornings.
Thanks.
This was a great watch, thank you for uploading this!
Fran, excellent solution. Thx for sharing. A PLL f multiplier was my first thought. Your use of a "Pulse Swallowing" method was genius. Some programmable counters use "Pulse Swallowing" methodologies.
Your solution was simple, elegant, practical, accurate & functional.
A monostable multivibrator using say a 555 could've been an alternative if you didn't have a 4017. That approach might've used more components though. I call it a junkbox solution, that is, whatever you've got available to avail a solution. Great job.
Great solution Fran! Also, thanks for making me feel better about the random state of my own lab :)
In my opinion that pulse dropping circuit is exceedingly clever. I confess I would've set about building a power inverter around a frequency standard, confident it was the only way to go.
Awesome. Looking forward to seeing it completed Fran.
I really enjoy the jimmy rigging you had to do to work around that really strange chip!
What a fascinating design decision to make a display change based on current cycles. Thanks for following up with the information that they made the other two versions that were fixed display modes. I was immediately curious about that. I suppose most people ordered what they needed for their design. When you got to the end and asked yourself the rhetorical questions about why one would do it this way. I certainly wondered if anyone from that era was faced with the same problem of getting the cycle switching display to show 24-hour time with 60 Hz. Maybe that never actually happened, but it reminds me that for people in those situations, where you have a chip/thing and you have to "make it work", there are no rhetorical questions. Thanks for sharing your solution.
Sometimes we do the things we do because we CAN! My 13 year old as asked me why I weld stuff back together like a broken chair or a broken tool. Answer: because I can, and it’s fun to exercise your skills!
Neat solution! It reminds me of how drop-frame time code works in video.
Compliments for your technical knowledge, which makes you understandable.
did it the exact opposite way to get a 60Hz clock chip to run on 50Hz mains. just looked in the spares box and found a NTSC colour burst crystal and made a single gate oscillator to get the output, then used some TTL dividers and counters to get a 120Hz clock out of it, then used a divide by 2 to get the final 60Hz clock to drive the unit. Stuck it all on a small piece of protoboard and crammed it in the unit, along with a small hole to get to the trimming capacitor for the oscillator. Worked out to drift around a minute a month, perfectly good enough for my application, and all made with parts salvaged from scrapped electronic parts.
Did a similar thing to run a 400Hz 26VAC hour meter, just used a 555 timer, a 2N2219A and a small audio transformer, with a tuned secondary, to get around 22VAC to drive the counter. did around 20k hours before the hour meter itself wore out mechanically.
Nice elegant solution, Fran! I tried to guess at the solution before I watched the video, and the thumbnail offered up a clue, but I was slightly off.
My idea was to use the 4017 as you have, but have Q0 to Q4 feeding the mostek chip via small signal diodes. Q6 would be tied to reset, giving the missing pulse every 6 pulses.
I've studied the circuit and finally I've understud how it Works. The counter pull down a pulse every 6 so the pulses are 50 in a second even if they are not equal.
It’s reassuring to know your lab’s organised the same way my workspace is, Fran Blanche. ;)
A messy lab is a happy lab :)
The nice thing about a messy lab is every now and then you find something unexpected in the clutter which triggers you off at a tangent and a whole new universe opens
Great stuff always learning from the Fran Lab! please keep it up.
Way to go, Fran! Keep it up.
Woot, it's Fran! In the Lab!
Here's a top tip for anyone worried that their lab is filthy: Call it a workshop instead.
Problem here though, is that FranWorkshop wouldn't kinda scan the same. Hmm!
Hoping that the next evolution of the circuit just uses transistors, no chips.
More quickies please :-) I'm on a limited data plan, and can only watch so many longer videos. Great project. Creativity is a good habit.
Nice solution to the 50 Hz problem! I totally agree not to just start throwing uPs at it. Old style is good!
Been a while since I've seen any of your vids-then I stumbled across this one.......said it before and I'll say it again......I LUV you!!!! :)
Hi Fran, love the clock stuff, esp the vintage part of it. I am trying to learn the mm5314 chip.
Keep the vids coming!! Thanks, Jim
That's a really clever hack... food for thought; cheers! 😁
Thanks Fran. I enjoyed the quickie
Love these quickies Fran. Hope you're going to do some longer videos like before
What you do is art.
What would you do if it was the other way around (running a 60 Hz clock in Europe)?
Build a crystal oscillator and devide the frequency down to 60?
Build a PLL circuit?
Use the 50 Hz to make a resonant circuit oscillate at 300 and devide that frequency by 5?
All of that sounds more complex/difficult.
That's a nice classic way to do.
Love your work!
Nice trick to fool the Mostek! I've had to produce clock signals in some of my projects, and I always go straight to a relaxation oscillator. I imagine that would work about the same, but your trick to 'skip a beat' is pretty neat. Shows how dumb chips are that they don't really care the time between pulses.
+1 on the quickies!
My alarm clock still is one that I built using a kit around 1980. Because its case is badly twisted plastic, held together with Scotch tape, I am reluctant to open it for checking the chip. but I think it was a Mostek. Unlike yours, it runs on a crystal - I remember I could tweak the frequency. And it is in 24 hour format. I made two modifications to the kit, I did not like the 35 V supply that I felt was scary. I reduced it to something between 20 and 25 V. On the other hand, I considered its bulk capacitor inadequate, so I added a bigger one, just hanging there behind the PCB. With that, my alarm clock survives nearly 10 second long power outages without losing its count.
I was puzzled by the high operating voltage, but believe now that it was due to the chip being built on P-channel FETs. I’ll check it, if the clock ever fails so that can justify opening the warped case.
AC pwr line frequency in the US prior to 1935 wcould be 50 hertz or 60 depending on what part of the country you lived in. The grids were all synchronized after that date to keep the street clocks on time
I though of a cycle swallow, or maybe faking it with a crystal oscillator and divider - but on the other hand, the mains frequency is actually specified to be long term accurate …. The main problem with mains driven digital clocks as opposed to synchronous motor clocks is that interference can add extra clock pulses
It's either that 60 to 50Hz converter or a really long extension lead across the pond.
I want a year down and tutorial of your storage scope :)
Great solution !
Awesome workaround to get 50Hz. I can totally relate to wanting to use classic chips- in the late 90s I used to scour thrift shops for a certain 70s Radio Shack project kit that had a (very simple) integrated circuit on board, which I would cannibalize to build the light controlled oscillators I'd use in tandem with my circuit bent SK-1's... (SK-1 pulse code driving LEDs coupled with CdS cells modulating oscillators built with said chips) People just didn't "get" why that was the chip I wanted, they'd say "just use a 555!" But there was something about hunting down those 150-in-1 kits and extracting their precious parts... ahh, the memories. --I think I still have a couple of those kits floating around somewhere!
Awesome solution! I was wondering about filtering out power supply spikes, which would falsely trigger the logic. Then I noticed the 390pF cap.
I sent you something in the mail which echos that exact sentiment :-)
- How is your transformer?
- FUUUUUUULLY RECTIFIEEEEED!
These quicky videos are great and I hope you keep going. Especially with that intro 😛
Now that’s hard core. I love the design. Damn those who want an Arduindo or RTC.
I enjoy the way your projects lift the lid on how complicated the things we take for granted are.
Awesome work!
What is happening with the oscilloscope keeping the signal on the screen ? Very impressive
Storage Scope for the Win!
Fran Blanche indeed !
i'm watching out for them on ebay
That's not filthy Fran, just well used!
Hello, did you succeed?, btw i am in a simillar problem, i have an old general electric radio alarm clock from canada, but i live in europe(50hz), so time slows down, is there perhaps any solution for me? Thanks in advanced
Great solution, I think they call that a 'swallow counter' since it swallows a pulse every so often. Note that the 4000 series chips like your 4017 have a maximum supply voltage rating of 15v but 16 probably won't kill it!
Max Vcc is 18v... no worries.
That's a proper lab that is! It's not a proper lab unless you have piles of projects half way through!!! Also wouldn't a ULN2803 or similar have worked for the lowside transistors? As the ULN2803 has been around since Pontias was still a pilot it would not be breaking with the vintage concept.
when i started to watch the video i was thinking the ways i would consider changing the freq'y, there was 2 ways, one was using a 555 timer to power a fast switching high power transistor, the other was using a decade counter and taking the pulse off 5 of the 10 but that still required to be driven by a timer chip...
the reason why i thought like this is because id more likely run it off a set of batterys and even with a 100LED if you drive them at 3mw they should be bright enough and that would be less than 300mw of power as not all the LEDs would be on!
i didnt consider running it off mains voltage and tapping off the internal freq'y like you have ...
you could use a 10mhz clocking xtal and a x5 multiplier, again good for battery use
That good ol' reliable CD4017 again! Nice.
It looks to me you might have saved some effort by running the "23" volt transformer as a bridge rectified center tapped for 11.5 volt (close enough to 12), or was 16 volts a requirement?
I'm often amused by the "youngins" who like to throw an arduino or pic at almost anything. I'll read what they're doing and think "you could do that with a 555..."
Ted Thompson: Spot on, but I can just see the responses from some people...
"555? Pah! What's wrong with a combination of basic ICs?"
"ICs? Pah! What's wrong with discrete transistors?"
"Discrete transistors? Pah! What's wrong with good ole vacuum tubes?"
"Vacuum tubes?" Pah!..."
Etc. ad nauseum...
#o)
To be fair, if you're just throwing some prototype together, an arduino is great. Of course, if you want to make more than just that, you can use something that fits better. Sometimes, a microcontroller IS what you want, though.
>"Vacuum tubes?" Pah!..."
Galena crystals! :-P
Wayne Hall: Don't mention crystals!! Some eejit will start going on about dilithium and warp cores!
#o)
I think a 556 would be better for this. Make one half time out in about 95ms, and make the other oscillate at about 50Hz. Feed the output from the first half into the reset of the second and every fifth pulse will be lengthened or shortened so as to maintain an average speed of 50Hz.
Is that a SECO Model 76 on the bench?
One of my secret favorite channels... :-)
what about just using a 50-hertz free run sine wave oscillator?
If I just wanted to throw an Arduino in there... what would be the point, you know? There's a million clocks out there that would be made like that. No. I'm using a vintage.... ... that's how we roll in the Fran Lab" - and that's why I love this channel so much!
Well remember - a cluttered desk is a sign of a dynamic mind. And I can do the 24 hour times in my head. Right now it's 19:19 hours here. Reason is I worked for a company with clients all over the world. And it's much easier to use a 24 hour clock than the 12 hour clock.
Smart solution!
Fran, if the PM is lit I would have added 12 to the hours digits? If PM High and D6 high add 1. If PM high and D5 high add 2. It would save messing around with the frequency. The 4017 and transistor is ingenious :D I hope that helps. Say Hi to Missy from England :D (Edit I forgot to say Hi to Missy)
It's a good thought but I think it'd need more circuitry..
also 'proper' 24h mode starts at 00;00 (midnight) and not 24;00
The output of the clock chip isn't binary, it's already been decoded to 7 segments and enable outputs for the six digits, so it would have to be converted back to binary first to do the addition.
I once built clock wich needed one impulse per minute (to move the minute hand in a slave station clock). Did that in a very similar way by cascading two counters.. About 35 years back... LOL
Great video Fran, like the short videos. The timer solution is nice for the 50Hz but how do you get the 4017 to work? I have done short video on using this thing but for the life of me I can not get it to work, so a quikie on getting the CD4017 / HCF4017 working would be nice hopefully showing the pitfalls as I'm falling down all of them. Thanks.
They are very current sensitive and for that chip all of the input and control pins must be held at either V+ or Gnd for it to function.
Thanks for the reply, I will have to have another play around with this IC.
Also your circuit allows for extreme accuracy in timing unlike a 555, Arduino, or even a crystal oscillator.
Hi Fran,
Hope you are still reading comments from this somewhat old video... Fran, I need your help. Being that I live in South America (Chile) where the AC mains standard is 220V @ 50Hz, I need a circuit the opposite of yours, with an output of 110V with a AC frequency of 60 cycles.
Before I continue on this endeavor, your opinion on how to proceed would be very helpful.
A) Is it possible to use your schematic as a road map for modifications?
B) Bite the bullet and just purchase some commercial converter device?
C) An Arduino (MCU) programming solution? (yuck)
Regarding your comment about using a micro controller solution (I agree with you, it's like cheating) versus an old school component solution... your singing to the choir. I much prefer the latter.
Thanks again,
Ken
one problem i see is, her circuit drops pulses, while your circuit would need to *add* pulses. you could create a simple 555 timer oscillator that runs some safe margin higher than 60Hz, say, 100Hz (to leave plenty of room for that RC oscillator to drift). Have a counter circuit watching the input, activating the oscillator after 5 input pulses, and have a second counter circuit watch the output and deactivate the oscillator after 6 output pulses.
@@spambot7110 Thanks! I never considered adding pulses beyond spec and then correct "backwards" by dropping pulses as needed. Thanks for the idea (My reaction to your solution was similar to Scotty's reaction hear: ruclips.net/video/jxzYTDX8bIg/видео.html
Brilliant idea! Sadly it doesn't work the other way around, from 50 to 60Hz.
Again, for upcounting to 60hz you would try a similar circuit but instead of muting a count you'd add one short pulse every 5 counts.
Any suggestions for a circuit to generate 50Hz or 60Hz without access to mains, and ideally without a microcontroller or crystal, but still fairly accurate? 555 style timer circuit is probably not accurate enough. Doesn't need long term stability, but over a few hours use, should keep reasonable time.
Wild clocks are just that - you need a very regular oscillator which is usually a crystal.
The MM5369 was a great chip for this. Little 8 pin DIP that would generate accurate 60 Hz pulses using a cheap and easily available 3.579545 MHz TV colorburst crystal. Unfortunately discontinued, but you might find NOS on eBay or surplus sellers.
Dave Curran There have been clock chips using a RC oscillator as fallback in case of mains blackout; the Fairchild FM7010 as used in Grundig's Sonoclock 21 would come to my mind here. But in real life the accuracy was more than lousy; even after just minutes of blackout it would already deviate a minute or two - no matter however often you calibrated the osc on it. Change in ambient temperature alone would throw it off. An old school approach for a 50Hz timebase using still available standard components could include a CMOS 4060 4027 and a 3,2768MHz xtal.
Thanks for the replies, yes, that is what I had thought, just trying to find a solution for a client of mine who has some contradictory requirements for size / cost / accuracy.
I didnt get how the 60hz pulse converts to 50Hz.."Counter counts 1 less?" Cant get my head around it....
So glad you did not use an Arduino or similar. Much more fun this way :-) Interesting chip. I have been thinking for a while now about putting together a clock using 50Hz power in the UK. I wanted to add the radio clock signal update as well. (en.wikipedia.org/wiki/Time_from_NPL_(MSF) ) . Assuming the frequency balancing in the UK is good then it should keep relatively good time, but thought a quick update from the radio signal would not go amiss - Besides - more fun. I love you 3d printed displays so was thinking to use those too. My clock might take a tad longer do though... I need to learn more.
Fran, how would a 50Hz vibrator valve (tube) go in being the clock driver for your display?
I have no idea whether a vibrator is even appropriate in this case, but as a Just-For-Fun project it may be an idea for a future vid.
ɐᴉlɐɹʇsn∀ ɯoɹɟ sɹǝǝɥƆ
my lab looks the same!! woohoo!!
I like little quickies with Fran.
I found an old clock chip mm5312 and I did the same thing using pic12f675 to generate 50hz or 60hz square wave, if you are interested i can send it to you the code.
Nice solution, but one thing to remember here; what OPAMP are u using? It surely is not an open collector output, and with some Opamps I would not feel really good shorting its output to gnd and just counting on its internal output current limiting to save it. Secondly as far as I see your approach only works 100% if the delay in the signal thru the buffer opamp is greater than the delay of the 4017 counter to turn on the transistor. If the signal thru the buffer opamp is faster than the blanking pulse (or it is a toss up) you may (occasionally) get glitches at the start of the blanking interval where a bit of the 6th pulse slips thru. Maybe too small a glitch for the slow Clockchip to act on it, but maybe so. Guess only long term stability would tell. Keep up the work.
Also the drawing of the schmatic seems to be a ltittle weird. The link between Pins 5 and 15 on the 4017 provides the reset pulse on reaching the 7th count (which seems counterintuitive at first but is correct). Still that pulse is only available for the few microseconds that the 4017 takes to reset count. The output to the green LED and the base of the blanking transistor must be tied to any regular count before that. And in the video at 5:05 the closeup of the PCB shows clearly that the green LED and 10k to the base of the blanking transistor are in reality tied to pin 10 of the 4017 (active on 5th pulse). But maybe that misdrawing was the intentional error aimed at chinese copycats - if they would ever design with obsolete MOS clock chips.
Old fashioned digital, I think I commented in G+. :) There should be some old style way to multiply 60Hz by 5, then divided it by 6, and end up with 50hz.
Quite smart, I like! :)
- The guy making an arduino clock... lol
Clever!
In 1974, at age 10, I built a Heathkit GC-1005 6-digit clock. I still have that clock and it works. This is based on the Mostek 5017 chip. It is switch configurable between 24/12 hour mode. It is 60Hz line frequency based. The power grid 60Hz does not seem as "on time" as it was 44 years ago. Back then, I could set this clock to the telephone "talking clock", or the CBS "Bing, da da da, da da" top of the hour jingle, and it would always be within a few seconds of real time. I built a few years ago a 6 digit Nixie Tube clock with the WiFi "NTP" clock sync. Now I easily compare "real time" to AC line time. I've seen AC time drift as much as 30 seconds. BTW: You can call WWV/WWVH on the phone: (303) 499-7111 / (808) 335-4363. Talking clock is not dead, yet.
The Mostek 5017 has a silicon Easter egg on the die of Mickey Mouse pointing his arms to 7 o'clock. How cool is that? You can see an image of this here:
www.flickr.com/photos/dr-engineer-001/52647186
What a novel 60 to 50Hz converter!, I would probably have used an Arduino nano to synthesize a continuous 50Hz (no gaps), totally overkill for your use. But hang on, I live in Australia, a 50Hz country, so going the other way, without a micro, could prove an interesting challenge.
For upcounting to 60hz you would try a similar circuit but instead of muting a count you'd add one short pulse every 5 counts.