Relay Failure

[Rosso]

I am using a PCB mounted DPDT relay rated at 1250VA at 250VAC, to switch a 1KW 250VAC immersion heater on and off, and which (the relay) has failed after a couple of weeks of use.
(The coil still actuates, but one of the contacts does not throw)

It is simply wired 250VAC to both common contacts, and the purely resistive load wired across the normally open contacts.

Should I have included something to protect the contacts, like a capacitor?

The relay is a Maplins **broken link removed**

 

[ericgibbs]

I am using a PCB mounted DPDT relay rated at 1250VA at 250VAC, to switch a 1KW 250VAC immersion heater on and off, and which (the relay) has failed after a couple of weeks of use.
(The coil still actuates, but one of the contacts does not throw)

It is simply wired 250VAC to both common contacts, and the purely resistive load wired across the normally open contacts.

Should I have included something to protect the contacts, like a capacitor?

The relay is a Maplins **broken link removed**

hi,
Are the contacts welded 'closed'.?

I guess as its a relay controller, the relay contacts can close/open at any point in the mains phase cycle.?
A resistive/capacitor across the contacts would help.

 

[Reloadron]

In addition to Eric's comment:

I guess as its a relay controller, the relay contacts can close/open at any point in the mains phase cycle.?

Another problem could be the paralleling of the relay contacts. Generally, though it looks good on paper, this is not a good practice. The contacts will not always make/break at the same instant in time. This can cause one contact to begin to fail before the other and with use the small problem compounds. I would add, as Eric suggest, a snubber across the load.

I would also consider using a good SSR (Solid State Relay) with zero cross over rather than a mechanical relay.

Just My Take
Ron

 

[Rosso]

Eric, no not welded 'closed', open circuit on one of the poles.
and yes they can open/close at any time in the cycle. I think that the situation is probably worsened because the relay driver is fed via a voltage comparator chip (LM339), which when the two inputs are very close to each other causes the relay to chatter before cleanly switching over. I tried to introduce some hysteresis to combat this, but couldn't get it to work properly, as it defeated the accuracy of the voltage comparator.
What would you recommend as a suitable resistive/capacitive circuit across the contacts?

Ron, I would prefer not to use a SSR if possible, but now the relay has failed, I could replace it with a single pole 10Amp relay (N29AW) instead of the DPDT 5Amp relay.
As per above, what would you recommend as a suitable resistive/capacitive circuit across the contacts?

 

[Reloadron]

As to an RC circuit I would look at maybe .25 uF and 470 Ohms. They are also commercially available as versions like these. I would go with the single contact relay and if at all possible get some hysteresis in that LM339 circuit. That On/Off chatter will wreak havoc on those relay contacts.

Ron

 

[ericgibbs]

hi,
As Ron says fit the snubber.
Chatter is a common problem with LM393 comparators without hysteresis.
Post your full circuit, we may be able to suggest a LM393 solution.

 

[Rosso]

hi,
As Ron says fit the snubber.
Chatter is a common problem with LM393 comparators without hysteresis.
Post your full circuit, we may be able to suggest a LM393 solution.

This is the full circuit, a LTSpice file and a screenshot.

 

[ericgibbs]

This is the full circuit, a LTSpice file and a screenshot.

hi,
I am still looking over you schematics and LTSim.
The relay is being operated at close to the minimum of 9V , actual 9.5V, the transistor base current is ~0.7mA and the relay coil current is ~36mA..

I cannot see a transistor type number.???

EDIT:
I suspect the reason you couldn't get the hysteresis to work is due to the lack of 10K input resistors to the final comparator, the 10uF caps would cause oscillation.

EDIT2:
What are the OPA types.?

 

[Rosso]

hi,
I am still looking over you schematics and LTSim.
The relay is being operated at close to the minimum of 9V , actual 9.5V, the transistor base current is ~0.7mA and the relay coil current is ~36mA..

I cannot see a transistor type number.???

EDIT:
I suspect the reason you couldn't get the hysteresis to work is due to the lack of 10K input resistors to the final comparator, the 10uF caps would cause oscillation.

EDIT2:
What are the OPA types.?

The transistor is a BFY53 (that I bought 30 years ago and never got around to using!!)
The OP amps are **broken link removed**

Would the inclusion of 10K resistors on both the inputs to the comparator, effectively isolate the op amp functions from the comparator function and enable me to then include an element of hysteresis?

Apart from the relay problem, the circuit works very well, and has performed well aver the past month or so.

 

[Rosso]

I have tried to introduce some hysteresis without success, as the input voltages into the comparator vary between 600mV and 3V , so the feedback voltage has little effect when the input is 3V but a overwhelming effect when 600mV.

I have had partial success, by linking a 10mf capacitor and 500k resistor in series between the output and input of the comparator, which when the comparator state changes, it provides a feedback pulse which provides hysteresis for a few seconds.
Not ideal....but..

 

[ericgibbs]

I have tried to introduce some hysteresis without success, as the input voltages into the comparator vary between 600mV and 3V , so the feedback voltage has little effect when the input is 3V but a overwhelming effect when 600mV.

I have had partial success, by linking a 10mf capacitor and 500k resistor in series between the output and input of the comparator, which when the comparator state changes, it provides a feedback pulse which provides hysteresis for a few seconds.
Not ideal....but..

hi Rosso,
Like you I am looking at the LTS circuit, I can improve the hysteresis if I remove that diode/1000uF and connect to 0V.
I am trying to introduce that 0.7V offset at the solar OPA output, I will post later.

 

[ericgibbs]

hi,
Look at this option.
Added buffers to isolate the 1meg/10uF from the FWrect outputs to the comparator., which now has hysteresis.
Removed the diode/1000uF to 0V from the mains opa's.
Added a diode/1meg between the two sections of the solar rectifier to give the 0.6V offset you were requiring.

 

[Rosso]

hi,
Look at this option.
Added buffers to isolate the 1meg/10uF from the FWrect outputs to the comparator., which now has hysteresis.
Removed the diode/1000uF to 0V from the mains opa's.
Added a diode/1meg between the two sections of the solar rectifier to give the 0.6V offset you were requiring.

Thanks Eric
I don't seem to be able to replicate the 0.6V offset, instead I get about 250mV?
I like the idea of using op amps to isolate both sections, and it makes the hysteresis work. Presumably by adjusting the feedback resistor value, I could fine tune how much hysteresis is applied (only need about 10 - 20mV).
This afternoon, I have bought a 10A SPDT relay and a cap/resistor to fit across the contacts, which will hopefully help this first prototype, but to make the changes above will need a complete rebuild (no space on the circuit board!).

 

[ericgibbs]

Thanks Eric
I don't seem to be able to replicate the 0.6V offset, instead I get about 250mV?
I like the idea of using op amps to isolate both sections, and it makes the hysteresis work. Presumably by adjusting the feedback resistor value, I could fine tune how much hysteresis is applied (only need about 10 - 20mV).
This afternoon, I have bought a 10A SPDT relay and a cap/resistor to fit across the contacts, which will hopefully help this first prototype, but to make the changes above will need a complete rebuild (no space on the circuit board!).

hi,
I have posted the LTS asc I have be working with, for reference, you may have to insert your model for the LM339.??

I have added a voltage/controlled switch on the solar current meter, in order to try the solar levels.

Lets know how it goes.

 

[KeepItSimpleStupid]

Use a contactor, which is of a special design.

 

[Rosso]

Use a contactor, which is of a special design.

A contactor?? What's that?

 

[KeepItSimpleStupid]

Example: Your Search Results | Newark.com

Used "30A contactor" and then sort for DPST

 

[Rosso]

Thanks Eric, yes, that works fine and moving the diode to create the 0.6V bias is a much neater solution than what I have.
The hysteresis also works well, just providing a little bias to ensure that the relay energises cleanly, without undermining the work of the comparators.
As I said earlier, these mods cannot be physically accomodated in my first prototype, as it is built and working (well urm.. it will be when I put the new relay & snub in!) but, if the relay chatter continues I can expect the new relay to only have a limited life, so I will build version 2 in the coming week or so, and swop it over.
Many thanks for your continued help & guidance.