Hi there all,

I have a project for driving 40W 120V AC lamps with the means of a digital clock's LED voltages of 1.7V DC. I had Sharp S202S02 relays for this purpose. The problem is about the digital clock that I've used. It has a strobe effect in 7 segment LEDs, like multiplexed drive. I got my AC lamps driven like 20Hz or something flickering, that I discovered in the LED's voltage measurement. It was changing all the time when the lamps are to be turned on and off. I need 0V DC while the LEDs are turned off.

My question is: where can I find such a non-multiplexed digital clock? OR is there a way of getting enough trigger voltage like 1V DC to the relays by the means of using some transistors or infrared diode couples.. I tried different tabletop digital clock radios as well, but couldn't find the correct pin assignments neither.



Any help would be appreciated..

Thanks

Burak.

 

If you have a schematic of that clock it would help. Other than that you might think about building a demultiplexing circuit. I've been working something similar for a stadium clock (that will probably never be built).

 

I really appreciate your fast reply, this is turned out to be disaster for me, even I tried it with a digital thermometer, and it worked, just because I thought that clock would be the same, but it is not. Here is the schematic for the clock itself: http://www.vellemanusa.com/downloads/0/manual_k8009.pdf

on page 22 , you will see the diagram. Waiting for the comments...

 

I think I can help you with this one, I have several ideas that should be easy work arounds. This the kind of project I like, truth to tell. You built this from a kit, so you know how to solder, how comfortable are you breadboarding? I'm thinking of something like optoisolators across each segment of the LEDs, with a simple integrator. Should be a piece of cake, although it's easy to say when you're not doing the wireing. You could even keep the clock as a display where ever, just have a bundle of wires leading out to another black box that interfaces between your kit and the big display.

 

Ready to hear that; I thought the same, even bought these: http://www.radioshack.com/product/index.jsp?productId=2049723&cp

I have only one day left to finish this and really rely on your help Bill...

 

I hope radio shack isn't your only source of parts....

Give me an hour or so, and I'll draw up a concept drawing and post it on this post. The good news is those SSRs were exactly what I was thinking about using, different package style than I'm used to, but that doesn't matter.

If flicker is still an issue then up the capacitance. Hope this helps. Just tap off each segment with a pair of wires, you'll need a lot of relays though, that is what multiplexing is supposed to help with.

 

I am gonna give a try right now. let me check if I have the parts..I'll be back..

 

Bill, it didn't work out. I used exactly 10uF and 4.7KOhm, but doesn't trigger. I tried without them, the 40W lamp keeps blinking. With the capacitor and resistor, nothing happens. Should I increase / decrease values? I tried without the resistor, nothing triggers..Any comment?

 

Ta taa... Bill, I owe you one.. I increased the capacitor to 1000uF, and decreased the resistor to 100Ohm.. now, it works.. Do you think 100Ohm is neccesary, I mean the resistor?... or the 1000uF would be enough? Just asking because on most of the segments, I have 6 40W 120VAC Lamps to drive. That means, for every 3 LED like you draw, one relay I have for them. I don't use 24h digit format, so I need only 23 of them. I already soldered the relays on protoboard. Do you think the project would be reliable like this? I would like you to see the roof.jpg again, you gonna see how many lamps I have to drive for each segment, and decide whether it is gonna be uitable or not..THANKS again for the most helpful tip that I've ever had in last 4 weeks!!!

 

Do you think the project would be reliable like this?

If you're using relays, no. I hope you're not being paid to do it,
because then money moved from one person who doesn't know what they're doing,
to another person who doesn't know what they're doing.
That's what triacs are for.

 

The resistor keeps from loading the drive circuits too much, but with SSRs this isn't a problem. I suspect it will work fine from here. Yes, the resistor is necessary. Do you really have 28 separate SSRs? Those suckers are expesive. How much did you pay per unit, if you don't mind me asking? I buy mine used for $5-15 each.

The resistor seems a little low, but it should work. It is there to protect the LED driver circuitry from the capacitor. At 1000uF you are loading the drive circuit pretty hard, You might think about bumping the value of the resistor up just a little. If you can't then don't worry about it.

If you're using relays, no. I hope you're not being paid to do it,
because then money moved from one person who doesn't know what they're doing,
to another person who doesn't know what they're doing.
That's what triacs are for.

Art, you really ought to learn what Solid State Relays are before offering an opinion.

 

I doubt that it is "multiplexing" as you say, more likely the LED segment is driven with a pulsed waveform to reduce LED power consumption without affecting the apparent brightness.

To rectify this, you want a peak detector. Place a diode and resistor (to limit the current draw from the segment driver) in series and into a capacitor to gnd. You may recognize this a a typical half-wave rectifier/filter circuit (sans the resistor). The resistor value should be selected to limit the current well below what the driver can source.

Connect the + cap to a comparator, with the reference set a bit below the typical value of the cap charge, the comparator can then drive a relay (either directly or via a FET), which will turn on the bulb.

When a segment is turned on, the pulses will charge the cap to a certain value. When this value exceeds the threshold setting on the comparator, the comparator output will switch and turn on the bulb. When the segment turns off, the capacitor will bleed off until it is below the comparator threshold and turn off the bulb. You may need to place a resistor in parallel with the cap in order to facilitate an appropriate turn-off time.

 

Did you check the schematic? It was definitely multiplexed, to reduce parts count. The problem (I've run into it before) is the two frequencies strobe together to create the flicker he was complaining about, the multiplex freq and the 60 hertz from the AC on the light bulbs. He just needed to turn on the SSR full time to eliminate it, which the integrator did.