SHARP S202S02F SSR Relay Problem

[burbing]

Hi there, we have a project for driving 40W 120V AC lamps with SSR relays. The idea is representing a digital clock's leds with AC lamps that we mounted into a big frame for the roof-top.

You may get the idea of what we like to do from here:

http://www.openmaterial.org/digitalclock/clock.jpg

http://www.openmaterial.org/digitalclock/roof.jpg

http://www.openmaterial.org/digitalclock/relays.jpg

But we couldn't find a way to trigger the relays. We applied 1.7V DC first to the 3rd and 4th legs, coming from digital clock's led pins and tried to get continuity from legs 1 & 2..but there was not.. Increased the voltage from 1.7V to 3.5V, and 5V DC.. The relay doesn't seem to be switching. Even tried a real life simulation by connecting the light bulb with the relay...but no results..

We appreciate a guidance for this Input voltage / trigger problem. The Relay's datasheet is here:

http://www.openmaterial.org/digitalclock/S202S02.pdf

There is also an application guide that shows exactly what we want to do at page 13, Fig.8-15. You may find it here:

http://www.openmaterial.org/digitalclock/SSR_usage.pdf

**In this guide, at page 13, Fig.8-15 shows a diagram to trigger the relay, but I am not sure about the CD4050 usage. Is it an IC? diode..Here is the link:

http://www.fairchildsemi.com/pf/CD/CD4050BC.html

If so which legs should I use to trigger, the input voltage that I have to use comes directly from digital clock's led pins approximately 1.7V DC. I may use triple connections as once for gaining 5.1V DC..

That's all we know. Really appreciate any help.

Thanks.

Burak

 

[ScottyUK]

Pin 3 needs to be positive relative to Pin 4. You need to drive a current of about 20mA through the relay input. 20mA is more than the CD4050BC can drive, so you need a buffer of some type. Something like a ULN2003 would be an option, or use a discrete transistor for each relay. You'll need to calculate a series resistor value to limit the current through the SSR to 20mA or less.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[burbing]

Thanks for your reply, I am related with electronics, but not that deep as you, so I need more detail if you don't mind ScottyUK..

As far as I understood:

If I am gonna feed the relay input with 1.7V DC, then I am gonna need 85 Ohms of resistor, or for 3.4V DC, the resistor should be 170 Ohms, due to keep the ampere about 0.02A regarding to the application sheet page 13, fig.8-15. That is a must to trigger with ampere, not with the voltage?

And the ULN2003, which legs should I use.. a sketch would be perfect if you have time for it...

Thanks..

 

[itsmoked]

Sorry. You have a serious problem.

You don't realize a considerable design obstacle you have.

Those LEDs are NOT just turned ON until the numeral needs to change. They are actually scanned rapidly. Very rapidly!

I can see by the hardware on the board that that is the topology used in your clock board's design(most LED clocks, in fact, use this technique.

What does this mean to you? It means even if you get the drive voltage and currents right the SSRs are not going to function. They have a response time for both turn ON and turn OFF. This means that they are going to get an ON signal, but before the SSR can even respond, the ON signal will be back OFF again.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ScottyUK]

Oops, trust me not to look at all the attachments!

Good point Keith.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[burbing]

Hmm, what should do now then? Any ideas?..Using a non-SLR Relay would solve the problem?..

 

[itsmoked]

Scotty; You kind of have to know about these type devices. I don't think any of the documents state it anyway.

burbing.... Man that is a tough problem really.

A lot of possibilities but none of them, ah, um, pleasant?

Probably the best would be to skip the incandescent lamps and go with HPLEDs. Then you just use a MOSFET circuit for each 'bulb' (now HPLED) and have them be the same as the board LEDS just more powerful.

Otherwise you need to come up with a per channel way to latch each SSR on. A retriggerable time delay relay. Essentially once it sees the ON pulse it will automatically trigger ON and stay that way until two times the scan update rate. That way the SSR stays on until that LED is truly OFF, which would be after a few scans go by with no ON for a particular lamp. But this scheme means 43 of these little circuits.

Might be easier to build it from scratch using a microcontroller to just do your clock in a non strobed manner. That's what I'd do.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[burbing]

Thanks guys..this is the clock that I am using now. and you are sure that it uses strobe to light up?

http://www.vellemanusa.com/us/enu/product/view/?id=350610#

than what if I use this clock instead?..

http://www.canakit.net/Default.asp?...tp://www.canakit.net/Contents/Items/CK253.asp

or this one

http://www.canakit.net/Default.asp?...tp://www.canakit.net/Contents/Items/CK118.asp

even if I overcome this strobing effect, I still need help..
Actually for this time adjustment circuits are only 23 items, not 43..cause I am in US, and we don't use 24h format for the first digit...

 

[Skogsgurra]

All those are multiplexed.

There are so-called RTC (Real Time Clock) chips available and they interface easily with a microprocessor. Translating time information to output pattern for your relays is trivial.

That's a much better way than trying to utilize a complete multiplexed clock board. Gives you lots of options. A good consultant with a suitable I/O board could do this in a few days.

If you want to hack an existing board, like the one you show, you can add latches and let each 7-segment pattern lock into a lathed driver as the different digits are strobed. That is something you could do on your own even if your exposure to electronics is minimal.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[burbing]

Appreciate the help. Can't come to a conclusion now. The question is:
1) If I use a non-strobe digital clock, or try some other alternatives to feed the relay with continuous current, the CD4050 or ULN2003 would help me to trigger the relay, as it's shown in the sharp's SSR_usage sheet page 13, Fig.8-15?

2) What if I use standart 5V DC relays, instead of SSR, would they notice this strobe effect? Probably they treat it as continuous 1.7V DC voltage supply?..

Thanks

 

[itsmoked]

Hi burbing.

I agree with Skogs those are all strobed too.

It's very hard to tell what would work not having your board here under an oscilloscope. If I knew the scanning rate, for instance, we could make a stab at simpler alternatives that otherwise may not work at all, or could work with some slight modifications.

I suspect that you would not need the CD4050 or ULN2003 if you are not driving the board's original LEDs also, as the drive for those LEDs is the same current as the Sharp SSR requires.

Keith Cress
kcress -

http://www.flaminsystems.com