Why did my relay burn out?

I’ve built a simple Arduino controlled dry box for drying out filaments for my 3D printer. It uses a peltier element which is turned on and off by a relay, controlled by an Arduino Uno. The system is intended to run 24/7 and periodically turn itself on and off. Initially it worked great, but after the first day of leaving the box running unattended, I came home to an apartment smelling like burnt plastic. And the relay was so warm that I would burn myself touching it. Needless to say this scared me pretty badly since it easily could have started a fire, so I would prefer to understand what went wrong rather than simply replacing the now broken relay. The relay also had visible burn damages on it (see first attached photo). I also peeled off the plastic case to look for damages internally, but to my untrained eyes it looks fine inside (see second attached photo).

It says on the top cover of the relay that it is rated for 10A 30V DC. The peltier element (TEX1-12715), which it switches on and off, runs on 12V and according to the specification uses a maximum current of 15A.

Could someone please advise me on what could have gone wrong here? Shouldn’t 15A at 12V be well within the capacity of the relay if it can handle 10A at 30V? Are there even simple Arduino compatible relays rated for higher loads than this? In the electronic stores where I live I have not been able to find any.

How were you driving the coil? Describe and/or draw the circuit.

Can we see a clear shot of the markings on the top of the relay?

petters:
Shouldn’t 15A at 12V be well within the capacity of the relay if it can handle 10A at 30V?

Sorry but no. You have to consider both limits separately. Your relay burned because 15A is outside the limit.

Blackfin:
How were you driving the coil? Describe and/or draw the circuit.

Can we see a clear shot of the markings on the top of the relay?

I’ve taken a new image now of the top – see Above.jpg.

I’m not sure what you mean by “the coil”. Do you mean how the relay was connected? The positive wire from my power supply goes into the relay, and the wire coming out from the relay goes to the positive lead of the peltier element. The negative lead of the peltier is always connected to the power supply. The Arduino is then connected with 5V, GND and signal pin to the relay and controls its switching.

GypsumFantastic:
Sorry but no. You have to consider both limits separately. Your relay burned because 15A is outside the limit.

Really? So even if I run something on 3.3 V, 10 A is the limit? I was always under the impression that voltage and amps were deeply connected when it comes to limits.

Is it not possible then to control my peltier module using a relay? Because none of the DC relays I’ve seen has a rating greater than 10.

petters:
Really? So even if I run something on 3.3 V, 10 A is the limit? I was always under the impression that voltage and amps were deeply connected when it comes to limits.

Is it not possible then to control my peltier module using a relay? Because none of the DC relays I’ve seen has a rating greater than 10.

They aren't in that case.
Taking to the limit. If I run 10mA I can't apply 10000V and vice versa. At 10kV it is very probable that the arcs created by the high voltage destroy the contacts.

petters:
Really? So even if I run something on 3.3 V, 10 A is the limit? I was always under the impression that voltage and amps were deeply connected when it comes to limits.

Yeah, the current limit is to do with the heat generated in the conductors inside the relay. Of course current won't flow without voltage, but if you get more than 10A going through those conductors, things will start getting hot.

The voltage limit is to do with how well the relay contacts can handle arcing, and how well isolated the conductors are from one another, and so on.

Automotive relays rated for around 30 or 40 amps (with a 12V coil voltage) are pretty common - they don't come in convenient arduino-friendly modules but they'll work if you connect things up right.

And don't trust Chinese rating. If it marked 10A do not let more than 5 amps flow through it.

The power dissipated by the relay contacts is I * I *R ( I squared R ) where R is the resistance of contacts and I is the current. So voltage here is not so relevant except for contact arcing. Some types of load increase contact arcing (inductive).

You could look at a logic level power mosfet instead of a relay. Important is that the gate voltage is high enough to fully switch it on or you’ll get the same problem as with the relay. The important characteristic is RDS(on) , the lower this resistance is the better, but you have to drive the gate at at least the voltage specified. Obviously observe the general maximum ratings for the device.

For example, this Power mosfet: IRLZ34PBF only costs a couple of dollars and can handle 30A at 60V

GypsumFantastic:
Yeah, the current limit is to do with the heat generated in the conductors inside the relay. Of course current won't flow without voltage, but if you get more than 10A going through those conductors, things will start getting hot.

The voltage limit is to do with how well the relay contacts can handle arcing, and how well isolated the conductors are from one another, and so on.

Automotive relays rated for around 30 or 40 amps (with a 12V coil voltage) are pretty common - they don't come in convenient arduino-friendly modules but they'll work if you connect things up right.

Ok, thanks for the help. I ordered this now which is rated at 30A at 30V.

Should that be enough for it to remain cool and safe even when used in a plastic enclosure running 24/7?

6v6gt:
The power dissipated by the relay contacts is I * I *R ( I squared R ) where R is the resistance of contacts and I is the current. So voltage here is not so relevant except for contact arcing. Some types of load increase contact arcing (inductive).

You could look at a logic level power mosfet instead of a relay. Important is that the gate voltage is high enough to fully switch it on or you’ll get the same problem as with the relay. The important characteristic is RDS(on) , the lower this resistance is the better, but you have to drive the gate at at least the voltage specified. Obviously observe the general maximum ratings for the device.

Is this disconnect between voltage and current when it comes to ratings specific to relays or does it always apply? In wires for example: I’m using 2.5 mm² cables in my system intended for strong AC voltage where the voltage is up towards maybe 230V and the amp rating is perhaps 10 A. I’m using this thinking that 15A@12V is far far below the limit. Is this assumption also wrong and this might also be a safety concern?

petters:
Is this disconnect between voltage and current when it comes to ratings specific to relays or does it always apply? In wires for example: I’m using 2.5 mm² cables in my system intended for strong AC voltage where the voltage is up towards maybe 230V and the amp rating is perhaps 10 A. I’m using this thinking that 15A@12V is far far below the limit. Is this assumption also wrong and this might also be a safety concern?

It applies to anything with electrical resistance and electricity flowing through it. That is, for example, why bulk electricity transmission is done at high voltages. For the same power, a higher voltage means a lower current and the lower current means less heat losses due cable resistance so relatively thinner wires can be used.

Power = volts * amps ( P = V * I )

Power = I² * R (substituting I * R for volts as in ohms law )

The power is dissipated as heat.

2.5mm² cable will be fine at 15 amps assuming you are not using vast lengths of it. Have you seen 15 amp fuse wire ?

Hi;
The burn pattern follows the bar that runs up the inside of the relay to the cross bar as well as the back of the terminals.
Simply not heavy enough, everything just got hot.
The contacts look burnt, so it is not fine.
What does the other side of the module, the track side, look?

I would not even put 5A through those terminals on the side of the module.

Have you go any temperature monitoring on EACH side of the Peltier?
You need to monitor how efficient your Peltier module is performing.

Tom..... :slight_smile:

TomGeorge:
Hi;
The burn pattern follows the bar that runs up the inside of the relay to the cross bar as well as the back of the terminals.
Simply not heavy enough, everything just got hot.
The contacts look burnt, so it is not fine.
What does the other side of the module, the track side, look?

I would not even put 5A through those terminals on the side of the module.

Have you go any temperature monitoring on EACH side of the Peltier?
You need to monitor how efficient your Peltier module is performing.

Tom..... :slight_smile:

Hi,
I don’t see that the contacts/terminals where the peltier wires come in look burned.

Unfortunately I have no way of monitoring the temperature of both sides of the peltier. It has a heat sink on both sides, but it situated in a 45-50 °C ambient temperature, so the heatsink on the warm side is always very warm because of this, and I don’t think its very efficient. But wouldn’t that just lead to a less than optimal temperature difference between warm and cold side? Are you suggesting that a low efficiency would it influence the stress on the relays etc?

I don’t know which the “track side” of the relay is, so I took photos of all sides which are attached here.

Hi,
contacts1.png

You need to monitor the Peltier temperatures because the element is rated for a particular temperature differential and MAXIMUM hot side temperature.
Do you have a heat sink on the hot side and a fan to keep it cool.
Have you measured the Peltier current?

MOSFET is best solution, BUT you need to take care of the Peltier element.

Tom.. :slight_smile:

TomGeorge:
Hi,
contacts1.png

You need to monitor the Peltier temperatures because the element is rated for a particular temperature differential and MAXIMUM hot side temperature.
Do you have a heat sink on the hot side and a fan to keep it cool.
Have you measured the Peltier current?

MOSFET is best solution, BUT you need to take care of the Peltier element.

Tom.. :slight_smile:

How would it even be possible to monitor the temperature of both sides as it is running? (as there are heatsinks blocking access to the surface of the element). The element has a small heatsink on the cold side and a much larger one on the warm side with two fans on it.

But like I mentioned, the peltier and its heatsink is inside a box where the ambient temperature is always around 50 °C. And to make matters worse, the warm-side heatsink is located a few cm from a heater element. So even with all the heatsinks and fans in the world its not possible to cool the heatsink below this. I’ve attached a photo of what the box looks like where the element is situated. Are you saying the peltier could fail if it gets to warm? Or are you suggesting that something potentially dangerous could happen?

petters:
How would it even be possible to monitor the temperature of both sides as it is running? (as there are heatsinks blocking access to the surface of the element).

The simplest method seems to measure the temperature of the heatsinks,
which will be quite near the element temperature.

alesam:
And don't trust Chinese rating. If it marked 10A do not let more than 5 amps flow through it.

"Chinese Amps" - is that sort of like "Internet Inches" - with a 2:1 conversion factor? :roll_eyes:

Hi,
Looking at jpg2 on the rhs near the terminals the contacts show burning and as the marks on the case show it has been overheated.
Since you have ordered a correctly rated relay the problem should be solved but as said by others the info on some imported components is seldom accurate and usually on the high side of actual characteristics.
The track side is usually the bottom side of the PCB where the devices are soldered onto the copper tracks but in the case of a double sided board you would need to specify which side was the "Track" side.
Cheers ..... Mike B