Relay EMI is triggering sensors on custom Arduino PCB 2

[SolarTrap]

Hi All,

I have a problem with the relay on a custom PCB I designed. On my breadboard prototype the parts where somehow spaced out enough to not show the problem.
The relay I am using is a DPDT 120VAC (Mouser: 653-LY2-DC24). In 50% of the switching operation, a button is triggered as a side-affect. It is 'Push_button1' or 'Push_button2' that get triggered - luckily it does not also trigger the reset-button. I found some resources talking about arc suppression at relays that are RC elements. But I dont know how to use that with a DPDT relay. I use the relay to switch a load between 2 power sources.

Any suggestions?

Markus

 

[Skogsgurra]

Since the relay contacts still produce interference, but not enough to disturb the inputs any more, it seems to be a good thing to have. Or, if you are absolutely sure that the load will be the same all the time, you can go without them.

A side note: The internal pull-up probably brings the voltage to Vcc (5 V), but your volt-meter loads the high-resistance internal pull-up so you measure 4.75 V. That would mean that the internal pull-ups are around multimeter Ri/40, does that make sense?

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

Correction: Ri/20.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[OperaHouse]

I don't see why the relay is switching any appreciable current when it closes. I took efforts to prevent hot starts with the compressor in my fridge program. They are designed to handle that on occasion, but as a regular occurance it will cause failures. Good to see you tried the resistor idea. I have nothing against internal pullups. Saves adding a resistor pack, but it is about the same as a 10K resistor. Under the wrong conditions, that is like a floating high imp input. LED lamps without an internal ballast have to have an internal parallel load resistor to prevent them ghosting just due to wire capacitance. Load it enough and the noise just doesn't have enough power.

 

[MacGyverS2000]

Internal pull-up/-down current is typically on the order of 1-10uA for a good design, give or take. A 470ohm pullup for a 5V supply is needlessly wasteful for a FET-based input... well, it's wasteful for about ANY supply voltage. At the very least, make it a 47k.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[OperaHouse]

The resistor solution was only to test a theory and it did. The terminating resistor was to place such a high load to noise that it would kill it. Since the switch current is only for a fraction of a second it really doesn't matter. The problem is in the design of this product. The relay should never be switching a load and the inverter should not be turned on by switching the power to it. That is placing 120A on the board and just inviting a disaster. I know he wanted to make this plug and play but nothing about connecting up a fridge will be easy. Most converters of the size needed to run a fridge will have a remote power switch. He should be doing everything possible to keep AC off the board. Let that be the installers responsibility. Technical issues aside, this board just has a lot of fluff. I just don't see this as being a successful product.

 

[SolarTrap]

Technical issues aside, this board just has a lot of fluff

Can you elaborate what you mean with "fluff"?

 

[OperaHouse]

I admit I'm only getting gleaning of the operation of the board. It's a refrigerator control. I have been running one for four years and have experienced things that I never thought of at first and I still don't get this board. All I care about is my beer is cold and the milk doesn't smell. Who needs a display. Not a housewife in Sheboygan. This should collect dust and never be seen from again. On mine I have a LED that blinks every 6 seconds. From across the room I can see if there is a problem with the battery or the fridge. The FET's on the board should be lost and have just a dry relay contact. Enough inverters have a remote switch and they are easy enough to add. That is just adding more length of wire, potential connection problems, noise, and limits where you can place the box. At first I thought you had built in a controller. What function are the current sensors? You have external controller for that. I see this little odd transformer on the edge. Hope that is that an AC current sense for the fridge. You shouldn't be switching power in a cycle and you should be verifying that you are actually powering the fridge. I power mine directly from the inverter terminals and sense voltage there. I added a slump detector to the program. That gives the best test of the battery system. In four years that has alerted me to battery and connection problems. What does the transformer do? That only adds problems, cost. Just like having the relay on the board. It is all just too gadgetry. At some point you will have to have peer review ao at least face the market.

 

[SolarTrap]

Thanks a lot for your explanations!
First of all: it is not a refrigerator control - it is a modular UPS controller. I am using it to take my fridge off the grid during sunshine but it also powers my stereo, TV, phone chargers, laptop and internet modem. I need the transformers to monitor the phase of the AC inputs so switching the relay is only happening when in sync. The bigger transformer also powers the relay and provides power to the inverter when the UPS is in standby. The sensors are monitoring the energy flow. Main aspect is to know the SOC of the battery and switch back to mains power if a certain threshold is passed.

-'The FET's on the board should be lost and have just a dry relay contact'
You can use a relay and my original design indeed used a relay but that actually destroyed some inverters I tested with. Especially the pure sine inverters are very sensitive to that. With mosfets you can control the inrush and avoid another spark.

-'Enough inverters have a remote switch and they are easy enough to add'
this is a problem because in-fact very little inverters have remote switching. I looked at dozens of models and only the very high-end have them.

What is a "slump detector"?

 

[OperaHouse]

Think of it as a battery tester. Many that put a load on the battery of 60-100A. A refrigerator or other motor device through an inverter places that same kind of load on a battery each time the device turns on for a second or two. My detector software looks for this drop in voltage, stores it, and compares it to a set value. With my fridge, 120A is a good stress on the wiring and battery. Over several years it has found a weak battery and loose connections. Just a good overall test that is done several times a day. Beats waiting for the inverter to go into low voltage lockout.

 

[SolarTrap]

I have measured the same turn on surge but I measured 0.5s for its durations (I do 30 samples per second).

Can you tell more about which measurements point to a weak battery or loose connection?

 

[OperaHouse]

That would depend on each setup. Many inverters select 10.5V because they anticipate very large losses in cables. My system uses one good car battery from the pickup I leave at home for the summer. I have two other batteries that should be able to run it when I use the main on the boat. Last year they were all identical so I put tags on them (after forgetting which was which). Ran the system on each one individually for three surges and marked down the slump voltage. Think some got to 11.18V when charger removed.

"monitor the phase of the AC inputs so switching the relay is only happening when in sync.
Many would think that is reasonable." Some do not think I am reasonable.

"original design indeed used a relay but that actually destroyed some inverters I tested with."
"Sometimes people blow up from natural causes." Repo Man.

"Seeing and believing are often both wrong." FOW.