EEVblog #801 - How To Design A Digital Clock

[zapta]

Does the 4026 sink the current as well when its low on those pins or does the total path have too much resistance/voltage drop to light the segments?

The voltage at the right side of the diodes can be at most ~1.4v (voltage drop of two diodes, one of them in the transistor between B and E) so the pullup resistor can induce at most ~0.7v on the three inputs (1.4 - one diode drop) which is insufficient to light the LEDs.

[zapta]

fun stuff to watch as usual. Just wondering about the carry from the seconds to the minute, and the minutes to the hour.
Shouldn't it be the reset there in stead of the carry ? the carry will only pulse on a 9->0 transition i presume.

No, the carry goes low once it hits 5 (see timing diagram) so the reset then switches it back high. So it works for any count >=5
http://www.ti.com/lit/ds/symlink/cd4026b.pdf

What prevents it from starting at 7 (invalid value) on power up? 

[Jr460]

Let me know, I'll be glad to answer anything I can remember.

I have many questions but they will be out of topic here. ;-)

Just a quick one, when you work in a power station, are you busy all time or just letting the computers do their job and waiting for the next incident?

Just to keep it on topic.  I did build a digital clock when I was about 11-12 years old, mid 1970s.  Based on one of those fancy all-in-one clock chips. 

Computers, even in the newer units of the time did nothing in terms of running a unit.  The computer downtown, a PDP/11, sent commands to the tribune control to bump up or down the power set point.  But one push of a button on the controls and the turbine was all local, and would either stay at the load set, or follow the load.  The control systems took care of fuel feed rates air dampers, valve settings, etc.  I suspect that even new units, the control systems might get some "assist" from a computer, but can keep things running just fine without one. 

(Sorry to ruin it for all folks that think a hacker on the inter-webs can take over power plants.)

If all the computers were dead everywhere, then the units would run, and follow the load.  You wouldn't have adjustments for 24 hour freq. control, and you wouldn't get the best mix of loads across units for fuel usage.

The plant I worked at started with two units in 1917, those where gone by the time I got there, and that space was used by one of the newer units.  Two more units were mothballed.  Ripping out all the asbestos would have cost a lot, and the space wasn't needed at that time.  The next newest unit was built in the 1950, the next in the 1960, the next in the early 70s, and the newest in the late 70s.  That plant is still running.  I can tell from the outside that have added to the two largest units SOx scrubbers.  I suspect the oldest unit that ran when I was there has been shutdown.  It was designed for lower operating pressures, thus was not very efficient.

The oldest running unit had a very basic control system.  Lines from the control room to the plant were all pneumatic.  The two newest ones each had a room with about 24 racks full of analog controls for the turbine and boiler.  They may have had some 7400 type logic chips.  No one ever touched that stuff expect for once a year when the unit was down, techs from the respective manufactures would come in and help tweak and align the feedback controls.

Back to your basic question.  The "operations" staff were the guys that sat in the control rooms.  If everything was fine, they sat around and waited for something to show up on the panel.  If a unit's load went way down, they might have to shutdown a coal mill, and then later start one back up.  That required someone to follow a procedure to start motors and change valves, and light off the burners from that mill.  Sometimes going out into the plant and cranking something open or closed.  Full unit startup and shutdown, they were busy as hell and you stayed away from the control room and out of the way.  I think it was requirement that you were a member of the local biker club to be in operations.

"coal yard" was busy all the time, got to keep getting coal out of the barges.  The "maint" department always had someone in the plant, with most of them being in on the day shift.  Something always needed fixing.  I think they had the most people of any department. 

"Instruments and Controls" was always busy with scheduled work.  They would pull sensors on a schedule and bring them back in and calibrate them.  Service all the chart recorders,  file/catalog the charts.

I was mainly in "Tech Srvcs"  We had the lab.  Coal was analyzed, boiler water chemistry was monitored and adjusted.  We did the engineering for new projects, did all the efficiency studies.  Adjusted the feed water heaters and the electrostatic ash precips.

The precips might be fun for Dave to see.  Big hockey puck SCRs and control that are fast to detect and ramp back power before an arc can jump between the wires and plates.

Other than that it was few office people, management, cleaning staff, etc.

[drws]

Does the 4026 sink the current as well when its low on those pins or does the total path have too much resistance/voltage drop to light the segments?

The voltage at the right side of the diodes can be at most ~1.4v (voltage drop of two diodes, one of them in the transistor between B and E) so the pullup resistor can induce at most ~0.7v on the three inputs (1.4 - one diode drop) which is insufficient to light the LEDs.

of course! I'd completely forgot to consider a voltage drop to ground via the base-emitter of the transistor.
All makes sense now. Really appreciate you taking the time to explain it for me zapta. 

[NiHaoMike]

I was just idly thinking what a waste of energy bench electronic loads are, chucking it away as heat, surely it could just be fed back into the mains? Not necessarily back into the grid (especially considering the pathetic amount paid for it these days), but just to take the burden from all the other lab gear, washing machine, microwave, lights, etc. on the mains from the electricity company and your bill.

Extra advantage - no need to dissipate all that heat. Run your load cool.

I found what look like some expensive commercial versions of this. Any DIY?

Most electronic loads are not used enough to be worth doing that. It's mostly the problem of efficiently accepting power over a very wide voltage range.

[zapta]

Really appreciate you taking the time to explain it for me zapta. 

That stretched my electronic knowledge to its limit.  ;-)

[EEVblog]

I am done you and everyone else get the last word.

Good, because you couldn't back up a single claim you made.
Now, how about that 4000 series logic huh?

[sbprojects]

I love Dave's videos, but I've never got myself to join the forum. Especially for this topic I created an account to showcase my own series 4000 CMOS clock, which I have finished just a couple of months ago. http://www.sbprojects.net/projects/cmosclock/



Unfortunately I was brutally reminded why I'm not a forum person. I thought this forum, being an electronics engineering forum, would be different. I really can't understand why all these useless flames are going on here about whether Ahmed is a dangerous bomber or not. These discussions are pointless and leading us nowhere. There are people who believe every Muslim is a dangerous killer, while (fortunately) most other free thinking spirits know better than that. There's no way any of us can make either side think differently. There is simply no reasoning between the two camps. And the rest of us don't even care.
I appreciate that Dave doesn't sensor any of the messages. But perhaps he could flag some of the replies as “flame” so those of us who don't care can filter them out.

[EEVblog]

What prevents it from starting at 7 (invalid value) on power up?

Nothing. I showed that in the video. Just forward to fix.

[AF6LJ]

I am done you and everyone else get the last word.

Good, because you couldn't back up a single claim you made.
Now, how about that 4000 series logic huh?

4000 series is fine, I have seen it do some real great things back in 1974 like a 2-12 MHz PLL synthesizer in a man pack radio. The stuff wasn't real fast back then. I was just out of high school and a friend was a project engineer for a local communications equipment manufacturer. Learned a lot back in those days.

Besides;
The only person I have anything to prove to is Me.

[linux-works]

I appreciate that Dave doesn't sensor any of the messages.

his lm75 is just stuck, is all.

(sorry, bad joke.)

[Macbeth]

I love Dave's videos, but I've never got myself to join the forum. Especially for this topic I created an account to showcase my own series 4000 CMOS clock, which I have finished just a couple of months ago. http://www.sbprojects.net/projects/cmosclock/

Very nice, I like your breakdown of the circuit. Unadblocked, and ahem, interesting links clicked too... Apparently I have >1000 Russian or Ukranian beauties looking for a guy just like me 

I appreciate that Dave doesn't sensor any of the messages. But perhaps he could flag some of the replies as “flame” so those of us who don't care can filter them out.

Seriously, I think you are asking too much. He can't be a mindreader on what people find offensive. Or watch every freakin' thread for it either 

Also why post in a thread that would have this flame moniker in the first place?

[zapta]

Now, how about that 4000 series logic huh?

It's awesome. I remember when I switched from TTL to 4000. I measured the supply current of a 4xNAND and the needle didn't move. I remember that moment, I was impressed.

It opened the door for new possibilities, with a little bit of creativity you can simplify the design. For example, I built once a phone pules dialer with only two 4000 ICs. I don't remember the details but one had inverters or gates and the other was a counter with 10 decoded outputs. The 'buttons' where an aluminum plate with 10 holes, each has a screw head isolated from the plate. You dialed by touching a screw head and the plate around it and the body resistance activated it.

At the time I was not ware of ESD so had no issues with it.

Also built a clock with TTls. Some binary counter IC and a high voltage decoder that drove nixies. High voltage came from rectified mains.

I like it much more today today those small and cheap MCUs, Digikey online ordering, internet communities, affordable oscilloscopes, and datasheets that are one click away. Life is good.

[zapta]

...I appreciate that Dave doesn't sensor any of the messages. But perhaps he could flag some of the replies as “flame” so those of us who don't care can filter them out.

Like your own message?

;-)

[Tac Eht Xilef]

You are not alone.  I am sure there are a great many people who come here would like to find an oasis where some respite can be found from such things.

^This^.

[cimmo]

Blimey, where do you live? The only time we had a power cut since the glory days of trade unions in the 70's was a year ago when I noticed my server powering down and rebooting and then the house lights dimming randomly.

We rarely get power cuts or brownouts here.
At home, maybe once a year at worst, and it's usually just for a few seconds. Maybe one major blackout in the 10+ years we've been there that lasted long enough for us to go "we'll, what do we do now?".
At the lab I've never had a single blackout (don't know about nights when not here), just maybe 3-4 glitches in the 4 years I've been here. Coincidently I got one this morning actually. Didn't reboot the computer, but dimmed the lights.

This weather event caused a seven day blackout (screenshot taken literally seconds before we went back to medieval living):

[boffin]

There is no need for opinion, this is purely a technical question. Please explain how a commercial desk clock chip can be modified into a countdown timer.

I think FOX News have a tutorial on their website.

[cimmo]

Did someone mention old-school clocks?

I do believe this one is based on Strowger Relays:
https://en.wikipedia.org/wiki/Strowger_switch


And here's one with Dekatrons:


http://wv7u.com/cwc/hourglass.html
Transistor free - but still digital!

[sbprojects]

However all hope is not lost. You can add the people whose posts you would prefer to not see to your personal "ignore" list by going to your Profile --> Forum Profile --> Modify Profile --> Buddies/Ignore List.

Thanks for the tip. It might help me stick around for a while longer.
Hmmm, but I can't block Dave, can I?
So I will still see part of the flames.
We'll see.

[sbprojects]

Very nice, I like your breakdown of the circuit. Unadblocked, and ahem, interesting links clicked too...

Thnax man, now I can buy myself a beer from today's earnings. Problem is, I don't drink beer though. Well at least you have some nice company in the near future. Fortunately for you the language of love is universal.

[SeanB]

Just saw this "interesting" clock on Reddit, so thought the humour might be appreciated here.

http://imgur.com/a/HOSEC

Luckily living in South Africa I do have the ability to have off grid power, and have enough fuel to be able to survive a few days without power, though the longest I have had to do so was 2 days after a drunk gentleman parked his car in the minisub and they had to dig up the road to get to the 11kV cabling to rejoin them, then place a new unit on the new slab they brought in by crane. The other was a shade over 5 hours after the other minisub did an impressive flash and burn after the 11kV incoming cable arced over, probably from a rat, though there was not enough left around the big charred hole to be sure. 3 20 hour outages after the cable thieves broke in and stole the neutrals and earth bonds.

Last night was watching a crew at 10PM up a wooden pole replacing an insulator on the one 230V phase, as it was arcing ( again) to the pole. The whole street is scheduled for a rewire soon, going from bare 25mm copper cable to 35mm aluminium Aerial bundle cable. The copper scrap value pays for the upgrade.

[bartm]

No, the carry goes low once it hits 5 (see timing diagram) so the reset then switches it back high. So it works for any count >=5

Ah i see... Asumption is the mother of all f**kup's..

[SeanB]

Same as the 4017 carry out, which is not surprising as they are basically the same chip, with one having 10 decoded outputs and the other a 7 segment decode.

The question of the DTL pulling the digits to light is simple, all CMOS chips, aside from a few designated as having open drain outputs, are push pull outputs, having an active device to pull up or pull down the output.

[timb]

We riddin' spinnaz, also clocks.


Sent from my Tablet

[RedOwen1177]

Hello Everyone!

I'm a new person here, but I love the EEVBlog. Dave, keep up the good work! I have a question about the clock project(I'm going to make one). With the "6 Decoder Reset", where do the outputs(?) go from the transistors and the diodes? Do they go to the reset pin? To ground? To Vcc? I really want to make a clock, so please help!