Hello. This is my first post on here, so be gentle.

I have a project which has almost succeeded. It's an internet-enabled controller for a wet-electric storage heating system using a Raspberry Pi. But you don't need to know all the detail as the problem is in one area: using SSRs to control 3 immersion heaters.

The SSRs are Fotek SSR-25DA 25A /24-380VAC details here.

The GPIO pins of the Pi when high, provide 3.3v DC to contact 3 of the SSR with contact 4 connected to one of the Pi's gnd pins. When high, light is visible from the little LED window in the SSR though it is quite faint.

On contacts 1 and 2 of the SSRs, we have supply and load for the immersion heaters. (The neutral and earth are uninterrupted but the live is interrupted by the SSR).

With the immersions supply isolated and the controller running, I've checked the voltage across contacts 3&4. It is indeed 3.3v and rock-steady.

With the immersions supply live, I could hear sounds from only one of the immersion heaters and much quieter than normal. By the end of the night, there was hardly any increase in tank temperatures.

I have temporarily bypassed the SSRs so I now have the immersion heaters controlled only by their own thermostats (and they are working fine). The intention is to have a variable tank target temperature to minimise standing losses on days when not much heating will be required.

I just can't work out what is wrong but I suspect the SSRs may be faulty. I haven't checked any AC voltages, mainly because I prefer to keep my hands well away from live mains. I don't think it can be incorrect polarity on the DC side as, firstly I've checked and double-checked and secondly, the LEDs do light up.

Is there anything else that I can have got wrong?

 

Check the input side current, use a 6v or 9v battery if neccessary, if OK, use another AC load source to test or use a couple of 100w lamps and check current and illumination for visible effect of the problem?
Max.

 

Only thing I can come up with is input voltage measurement.

Seems you are close to the lower limit.

Can you connect a 9v battery to ssr input? See if the indicator is brighter.
A clamp on meter would be helpful and safe.

 

Thanks for the quick replies.

I was wondering if the control voltage might be the problem and as I've never knowingly seen an SSR before trying to do this, I've no idea how bright the LEDs are supposed to glow.

The other thing that I didn't mention is that the system is up in the wild bits of Scotland and I'm now back home in southern England. I'm hoping to go back to Scotland early January armed with all I need to solve the problem.

So: I can buy another of these SSRs and try things out with batteries. Or, I can buy different SSRs that have a better chance of working. I favour the second option as there's not much I can do about the 3.3 volts control level. If these SSRs don't switch on at 3.3 (even though the spec says 2.4) then I definitely need some different ones which do.

I have to admit, when shopping for the parts for this controller, I tended to go for cheap things on ebay from China. Is Fotek a known brand? If so, are there fake ones to watch out for? It's a bit of a scary thought. Any suggestions as to what kind of SSRs would be more likely to work in this situation?

 

I generally stick with Opto22, not sure if they are avail in the U.K. but are pretty popular.
They are also available on Ebay.
Storage heaters have a much lower resistance (current) at cold switch-on so your SSR should be rated on the high side.
Max.

 

Will a Raspberry Pi work on 5 volts?

 

The 2.4 is minimum which you can't use for design. The 3 should be ok though.

That's about it. Will either turn on or maybe skip phases at threshold.

 

I've had those exact same SSR's not trigger with a 3.3V board either but 5V worked just fine.. You need a higher voltage.. (relay driver)
Like others have said its as simple as trying a higher input voltage to make sure its working right..

"most" of the "low cost" SSRs Ive seen say 3-32V in the specs but if you did deeper its really 3.5V min and they just leave the extra .5V off the marking.

 

I've had those exact same SSR's not trigger with a 3.3V board either but 5V worked just fine.. You need a higher voltage.. (relay driver)
Like others have said its as simple as trying a higher input voltage to make sure its working right..

"most" of the "low cost" SSRs Ive seen say 3-32V in the specs but if you did deeper its really 3.5V min and they just leave the extra .5V off the marking.

Nothing beats experience!

ps.
Can you bias them on? And turn them off with controller.

 

Never had problems with these.
http://www.opto22.com/site/pr_details.aspx?cid=3&item=240A25
Max.

 

ps.
Can you bias them on? And turn them off with controller.

R.Pi only offers 3.3 volts on the true I/O pins, though there's also a constant 5 volts available which comes direct from the board's power supply.

By "bias them on", do you mean give them 5v from the start and use a transistor to switch off? Yes, I guess I could do that.

 

I know next to nothing about Raspberry Pi, but I have read that the I/Os can be configured to sink or source from 2 to 16mA. The SSR that you have seems to require at least 6mA. You say that the "little window" is quite dim. How much current is the Pi sourcing to the SSR?

 

I think we sneaked up on a good method in a roundabout way. Give the SSR +5V and sink the current through the Pi.

 

I can see a big price difference. The Opto site won't give me a price unless I register, I guess. But equivalent devices from RS Components (a UK major components supplier) are about 70 USD equivalent and I paid around 5 USD each. Perhaps you get what you pay for.

 

There is this seller that is cheaper ebay 370964044875
Max.

 

I think we sneaked up on a good method in a roundabout way. Give the SSR +5V and sink the current through the Pi.

Like he said.

 

Tracecom: I didn't measure current. The Pi can do 50mA total at 3.3 volts and only 2 of these SSRs would ever be ON at a time. There are other things attached to the Pi, but none that would take significant current on 3.3v. There's a 4-channel 5v relay module but that takes its power from the 5v pin and uses 3.3 just for signal. There are six 1-wire temperature sensors which use the 3.3 supply.

Max: Thanks - that one is a good deal (used - which is fine by me) but he only has one.

I'm coming round to the idea of keeping these cheap SSRs and giving them 5v via a transistor. I have a total of 2amps to play with at 5v (including power taken by the Pi and all USB devices) so I think it will be no problem.

The only remaining worry is the efficiency of the cheap SSRs. I gather that 98-99% is the usual but I don't know if cheaper ones are significantly worse at generating heat. Each immersion heater is 3KW, so I expect 30-60W warming each device. If there's much more than that, the running cost will make them more expensive in the long run. Does anyone know if there's much variability in efficiency?

 

If you feel ambitious you can make your own fairly easily and cheaply, opto isolator and Triac, you do not even need zero cross switching as it is not a phase angle controlled application.
Max.

 

I learned something today.

Testing Crydom 40 amp ssr's for turn on voltage. D1240-b
Had a couple new ones in a drawer.

The "B" means nc.

Did not know they were available.

Anyway, turn off is 1.8 volts. Don't know if that would follow standard ones.

Glad I tested them before using them.

Tested some 100 amp ones I have on ebay.
Indicator comes on dim at 2.4 volts. Turns on fully at 2.9

Pretty confident it's voltage issue with op.

 

Tracecom: I didn't measure current. The Pi can do 50mA total at 3.3 volts and only 2 of these SSRs would ever be ON at a time. There are other things attached to the Pi, but none that would take significant current on 3.3v. There's a 4-channel 5v relay module but that takes its power from the 5v pin and uses 3.3 just for signal. There are six 1-wire temperature sensors which use the 3.3 supply.

My question was intended to ask if you set the Pi I/O ports up to source at least 7.5mA, which is the minimum required by the SSR.