Fibaro single switch on daikin Inverter air condition

[Tundel 9]

Nice relay 

 

I tried to measure my Toshiba air heat pump power usage (I don´t want to switch on-off with this relay such appliance) with  

Please login or register to see this link.

 relay.It was not a good experience, because of constant overcurrent protection problem (Peak value seen on attached picture).

 

Then I  replaced this relay with FGWP-101 wall plug (older, z-wave model) and this is measuring power consumption just fine.

 

Please login or register to see this attachment.

[petergebruers 1,268]

On 1/3/2018 at 4:30 PM, robmac said:

finally installed that heavy duty switch so opened the cover and took a few pictures for those that were interested in switching heavy domestic loads.

 

Thanks! That's awesome, you even included the datasheet! Yes that is a nice relay. It is almost the same as the black one, on my desk. It is the biggest relay, not counting the SSR. Look how tiny that FGS-212 relay (smallest relay) is...

 

Please login or register to see this attachment.

 

It is a "latching" relay meaning you energize the coil to open it, then energize the coil with opposite polarity (single coil latching variant) or a second coil (double coil latching variant) to close it. So it stays in either on or off position, without current flowing through the coil. This also means in case of power loss, the relay does not return to "off" which makes it unsuitable for certain applications. They do not specify a (cage) motor load, but assuming it can handle 1/5 th of the maximum current, that is still about 12 A so 3 KW "nasty load". So plenty of LED power supplies, an induction motor, all your fluorescent lamps, a washing machine...

 

I can't see what kind of terminals those are, I think those are not "cage type". You almost certainly need a good quality cable shoe if you use stranded wire. Solid wire can be bent in a loop.

 

On 1/3/2018 at 5:57 PM, Tundel said:

I tried to measure my Toshiba air heat pump power usage (I don´t want to switch on-off with this relay such appliance) with  

Please login or register to see this link.

 relay.It was not a good experience, because of constant overcurrent protection problem (Peak value seen on attached picture).

 

Then I  replaced this relay with FGWP-101 wall plug (older, z-wave model) and this is measuring power consumption just fine.

 

Interesting experiment, although you did not measure peak current at all - but that is not your fault!  And we can learn a lot from this, so thanks for sharing!

 

Here are some random thoughts:

 

FGWP-102 is Z-Wave Plus and has better, stricter protection than the older FGWP-102. It will not allow you to exceed the capabilities of the relay, not even for a fraction of a second. This is good, this is to protect you! Official statement!

 

 

The main point is "The device has a protection that will turn the load off in the case of detecting current greater than 110% of rated current (>12A). It is a safety function and it cannot be turned off. After its activation the load can be turned on back again by pressing the B-button or sending a control frame. This function is independent of overload safety switch set in the parameter no. 3" Maybe the 12A should be 13A but anyway, exceeding the maximum turns of the wall plug.

 

The FGWP-101 does not have this "fast" protection. It does have overload and over-temp protection, so it will still turn off the load if you exceed the ratings for a longer period. Thermally, that might be minutes. Exceeding the max W setting is faster  (I think it takes a few seconds to trip, I have not measured it).

 

This is why your FGWP-101 can handle the load, but it won't be happy... The contacts will degrade much faster if you exceed the inrush capability of the relay.

 

Why do I claim you did not measre peak current? There are several reasons...

• The definition of peak current or inrush current varies. For motors, please check the link at the end of this post. Electronic components like TRIACS for instance have two parameters. The first one, "Non repetitive surge peak on-state current" is defined as the current, during one cycle (so 20 ms in Europe), a device can handle without self-destruction. . For a TRIAC this is (rule of thumb) 10 X rated currend. So a 40 A solid state relay can start something that causes 400 A to flow during 20 ms (so averaged over 20 ms). Exceeding this limit will cause the die to melt so the failure mode is mostly "fail short" like a welded relay. The second parameter is "I2t value for fusing". For a 40 A TRIAC this is 1000 A2s for impulses shorter than 10 ms. It is often said, this is a limit of the bonding wires, exceeding this current will destroy the connections So, If you have a current of 1000 A but only during the first 1 ms then the SSR should be able to handle that.
• All Z-Wave devices are too slow to record fast current changes and peak values... We are talking about "milliseconds" and "tens to hundreds of amperes". The converters on those modules simply do not allow this kind of measurement. But the FGWP-102 and the newer relays and also Dimmer 2 can *detect* high current. They will shut off and report "over current" to the controller. I did some measurements on the dimmer 2 and if I remember correctly, it detects an over current condition in less than 0.1 ms (it is a mosfet device, not a TRIAC, so it can turn off a current). Notice, they all can report "overload" which is not a peak current, but a longer, averaged current.
• Measuring peak current and inrush current can only be done with an oscilloscope, a non-inductive current shunt and a clean setup.

User @earto shared a picture of an oscilloscope measurement here.

 

 

Some clamp meters used in industrial settings can report inrush current fo motor loads. I chose a random article from Fluke to give you some background...

 

Please login or register to see this link.

 

Key points if you do not want to read that article:

• Only meters that offer an inrush button can measure inrush current.

• this inrush function takes approximately 400 samples over a 100-millisecond period and calculates the actual starting current.

• Inrush current can be 4 to 10 times greater than the normal running current, depending on the type of motor.

For sake of completeness, this is what a FGD-212 can report

 

4-2 Heat Alarm -> Overheat detected,Unknown Location

8-5 Power Management -> Voltage Drop/Drift

8-4 Power Management -> Surge detected

8-6 Power Management -> Over-current detected

8-8 Power Management -> Over-load detected

8-9 Power Management -> Load error

9-1 System System hardware failure

 

So 8-6 is inrush current, 8-8 is exceeding current or power over a longer period (longer than 20 ms).

 

 

 

Edited January 6, 2018 by petergebruers

[robmac 173]

4 hours ago, petergebruers said:

They do not specify a (cage) motor load, but assuming it can handle 1/5 th of the maximum current, that is still about 12 A so 3 KW "nasty load". So plenty of LED power supplies, an induction motor, all your fluorescent lamps, a washing machine...

I think Aeotec claim 5hp motor so 3 to 3.5 kw ish. I use on my washing machine and tumble dryer with no ill effects.

 

Also the device comes with eye connectors to fit to your cables.

 

While on the subject of relays and switching tricky loads, my brother pointed me to these which I have not tested yet but look promising.

 

Please login or register to see this link.

 

They have a web api so could be integrated to HC2 and @ 16amp * many relays could switch a few independent circuits of LEDs if you have a room that you want to light brightly.

 

 

Edited January 6, 2018 by robmac

[petergebruers 1,268]

On 1/6/2018 at 7:57 PM, robmac said:

While on the subject of relays and switching tricky loads, my brother pointed me to these which I have not tested yet but look promising.

 

Please login or register to see this link.

 

Interesting! I'm not familiar with the product. Here's my personal opinion, based on what I have read. Specs & pictures look good. They use decent relays. Many variants, some with both analog and digital I/O. Wifi or wired. But I see a few possible issues. It is kind of "integrator stuff" (not a finished product you can plug into your electrical system) so you might have a problem with regulations (electrical code) in your country. I checked a few of them and they do not seem to measure current or power so please double check that. It is not Z-Wave zo you'll have to use them in scenes or VDs, no nice device icons & device types...

[robmac 173]

Hi @petergebruers

 

It is integrator stuff a bit like your very nice SSR but using code rather than wires.

 

Also only good for tricky loads that can be split into many banks like many LED bulbs that would be too heavy to start for a single relay as a single load. 

 

It depends on the load a bit regarding regs. If on 230V side probably an issue but if on a low voltage side then not.

 

Power metering not sure about but here you can educate me. 

 

Is the welding of the relay caused because the inrush happens as the contacts come together? Possibly a bit of bounce also that allows arcing?

 

If it is, would a lesser z wave relay measure power safely as long as it was always closed as the other relays controlled the load? 

 

 

[petergebruers 1,268]

On 8-1-2018 at 8:05 PM, robmac said:

It is integrator stuff a bit like your very nice SSR but using code rather than wires.

Also only good for tricky loads that can be split into many banks like many LED bulbs that would be too heavy to start for a single relay as a single load.

Agreed!

 

On 8-1-2018 at 8:05 PM, robmac said:

Power metering not sure about but here you can educate me. 

I checked a few of the variants and I did not find a spec. Also, going by the pictures of the PCBs, they lack the necessary components. Just saying you might want to check if you want to detect (state of) your load.

 

On 8-1-2018 at 8:05 PM, robmac said:

Is the welding of the relay caused because the inrush happens as the contacts come together? Possibly a bit of bounce also that allows arcing?

Indeed, a high current flows through a very small contact area when the contacts "meet". It may seems strange, but all power relays "weld" to some extent by design... That is how they get such a low contact resistance (milliohm). But the pulling force of the spring should be able to pull the contacts apart, unless you exceed the inrush capability of the relay. High inrush current causes some metal to vaporize and cool and deposit somewhere else on the contact. So it might not stick but still the contact degrades (because it becomes uneven). Older relays contained cadmium, which is soft and makes good contact. Another strange thing... power relay specifies some *minimum* current too. If you go below this spec, the contact might not weld at all, leaving non-conductive material between the contacts and then the relay will not have a milliohm resistance (can be several ohms on relays if they rarely switch). When contacts open under conductive load, arcing occurs, which also causes localized heating and some contact material loss. A device to quench those arcs is a "snubber".

 

On 8-1-2018 at 8:05 PM, robmac said:

If it is, would a lesser z wave relay measure power safely as long as it was always closed as the other relays controlled the load?

 

I like the way you think

 

Yes... and no... Yes, the relay may weld and stick but if you do not care that is OK. Technically, the inrush current might blow the current measuring device, that is most likely some sort of resistor (but it could also be a magnetic Hall sensor). Some resistors are designed to withstand overload conditions. The Aeon "heavy duty" switch specifies inrush current, so they should have the correct type of resistor. I am not saying the shunts in other relays are bad, but rather "unspecified". I think you should not exceed the spec and Fibaro says "resistive load" in many of their manuals.  This means "no inrush". Also, the second generation Fibaro devices will still detect overload and over-current and then probably turn of their relay.