EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: SgtRock on February 15, 2012, 06:01:36 am
-
Greetings EEVBees:
--I cannot afford 40,000 Dollars for a Hybrid Automobile, so little projects like this are my effort to save the planet. Just kidding, I am only trying to help a friend save 20 dollars. The small (120V, 1500W) Everstar space heater seen in the pictures below belongs to a friend of mine. During the Winter here in central Florida, often, one can get away with using a heater like this in the bedroom over night. The lady that this belongs to has probably used it on fewer than 50 nights, so it has seen very little use.
--She brought it to me because it had ceased to heat. I should explain that the user sets the temperature by controlling a lever which closes or opens a set of points. One leaf of the points is a bimetallic strip which opens the points back up when it is hot enough and closes them again, when is cools down. When I opened the heater up I could find no problem. The closed points had practically no resistance. I did notice they looked a little scorched, so using an old shade tree mechanic trick, I used the striker from a matchbook cover to clean them up a little, bobs your uncle. Only problem is after about 10 nights the problem reoccurs. After turning out all the lights I noticed bright blue arcing, just before the points close. Apparently metallic oxides are forming that stop major current from flowing, but will still allow a continuity tester to work. There is a plating layer that lasts for a while but, if the points are constantly opening and closing, eventually the plasma vaporizes the alloy coating away and problems begin. When the unit is used in a very small space, it cycles frequently, increasing the rate of wear on the points. I have seen this kind or arcing also, with well pump, and exterior heat pump points sets.
--My question is what kind and size of capacitor can I connect across the points, to suppress the arcing? It never gets very hot inside the plastic case, as it is never more that warm to the touch. Your help and advice is humbly solicited.
"I wear suspenders and a belt. I am a security man all the way"
Justin Wilson (The Cajun Cook) 1914 - 2001
Best Regards
Clear Ether
-
http://www.allaboutcircuits.com/vol_4/chpt_4/2.html (http://www.allaboutcircuits.com/vol_4/chpt_4/2.html)
which would fall back to the F=1/(2*pi*R*C) for your values,
-
Dear Rerouter:
--Thanks for pointing me in the right direction. I think I have a suitable capacitor and resistor combination now.
"He was born ignorant, and has been losing ground ever since."
Fred Allen 1894 1956
Best Regards
Clear Ether
-
100nF and around 10 ohms right? available off the shelf as a contact suppressor. I wanted some, but it was cheaper to go out and buy the 3 resistors and 3 100nF 2500V polycarbonate capacitors ( wanted a long life out of them, was tired of replacing that particular plug in relay) to do the same. Running 5 years now, normally it would eat the relay in 2 years. Over 2 million cycles of operation there.
-
Dear SeanB:
--I was considering going with a 0.33 uF XG-FS 250 Volt Safety Capacitor, which was formerly strapped across the mains switch from a 120 V computer monitor, combined with a 2 Watt 100 Ohm Metal Oxide Film Resistor. I do not know much about this sort of thing, but I have seen similar values used, according to my research on the net. It has not been cold enough lately, here in Florida, to want to heat up the place, testing a heater, not that I am complaining. Thanks for your interest.
"Measure twice. Cut once"
Norm Abram 1950 -
Best Regards
Clear Ether
-
You only need to put snubber circuit (capacitor and resistor) in case when you have an _inductive_ load.
In your particular case - you have mostly _resistive_ load - the heating element has very little inductance compared to motors, pumps, etc. where you should definitely put snubber across relay contacts. You don't need a snubber there - it's pointless.
So, if contacts of that thing is made out of sh*t - the snubber won't help you either way.
-
Dear Phil_ip1:
--Thanks for your concern. Just to be clear. Are you saying I cannot stop the arcing with a snubber? Or are you saying that it will do no good to stop the arcing? This particular heater does not have a relay per se, merely points that open or close, actuated mechanically by a bimetallic strip, according to temperature. Thanks again.
"He didn't know where he was going. But he knew where he was wasn't it."
Lord Buckley 1906 1960
Best Regards
Clear Ether
-
Dear Phil_ip1:
--Thanks for your concern. Just to be clear. Are you saying I cannot stop the arcing with a snubber? Or are you saying that it will do no good to stop the arcing? This particular heater does not have a relay per se, merely points that open or close, actuated mechanically by a bimetallic strip, according to temperature. Thanks again.
"He didn't know where he was going. But he knew where he was wasn't it."
Lord Buckley 1906 1960
Best Regards
Clear Ether
No, you would not completely stop the arcing with that. It could help a little bit, because that heater has a fan inside, which is inductive load. But it's pretty small inductive load, so I'm not sure how much it would help. Not too much, I guess.
And about relay: it has one! It's just not electromagnetic relay, but thermal bimetallic one. The contacts are the same, so the same principles apply in this case.
On how to solve this problem, I can propose a couple of things:
1. Replace bimetallic temperature sensor/relay/switch (that arching thing, for short) with a better quality one.
2. Buy external thermostat, and connect the heater through that.
3. Install beefy external relay, so that internal bimetallic thermostat will switch very small load - relay, instead of the whole heater with the fan.
4. Keep properly cleaning the contacts.
5. To throw the heater out and buy a higher quality one.
-
Dear Phil_ip1:
--Thanks for the clear answer. I shall advise the lady how to clean the points with a match striker. The unit only cost 10 or 15 dollars and, here in Florida is used very infrequently. Other that those stupid points the heater looks brand new inside.
"We are all born ignorant, but one must work hard to remain stupid." Benjamin Franklin
Best Regards
Clear Ether
-
You only need to put snubber circuit (capacitor and resistor) in case when you have an _inductive_ load.
Why will the snubber not work with a resistive load? The same principle applies when a circuit is interrupted regardless of the load.
Consider: the contacts start to open, and as soon as a tiny air gap appears between the contacts the full supply voltage appears across that gap (depending on timing of AC cycles, and so on...). A large voltage across a tiny gap produces a spark, which erodes the contact surfaces. Eventually the contacts get so far apart (or the AC cycle crosses the zero point) and the spark is extinguished. But it has already done the damage.
Now place a capacitor across the gap. While the contacts are closed the capacitor is shorted out and therefore remains discharged at zero volts. When the contacts start to open the full supply voltage tries to appear across the gap and the capacitor in parallel. Only this time the capacitor limits the rise in voltage due to the dV/dt rate limit across the capacitor. If the capacitor is big enough no voltage can appear across the gap as the capacitor absorbs it. Eventually the gap gets large enough that sparking is no longer possible and the capacitor has done its job.
So I say try the capacitor. If it works for inductive loads, it will work even better for less stressful resistive loads.
-
You only need to put snubber circuit (capacitor and resistor) in case when you have an _inductive_ load.
Why will the snubber not work with a resistive load? The same principle applies when a circuit is interrupted regardless of the load.
Consider: the contacts start to open, and as soon as a tiny air gap appears between the contacts the full supply voltage appears across that gap (depending on timing of AC cycles, and so on...). A large voltage across a tiny gap produces a spark, which erodes the contact surfaces. Eventually the contacts get so far apart (or the AC cycle crosses the zero point) and the spark is extinguished. But it has already done the damage.
Now place a capacitor across the gap. While the contacts are closed the capacitor is shorted out and therefore remains discharged at zero volts. When the contacts start to open the full supply voltage tries to appear across the gap and the capacitor in parallel. Only this time the capacitor limits the rise in voltage due to the dV/dt rate limit across the capacitor. If the capacitor is big enough no voltage can appear across the gap as the capacitor absorbs it. Eventually the gap gets large enough that sparking is no longer possible and the capacitor has done its job.
So I say try the capacitor. If it works for inductive loads, it will work even better for less stressful resistive loads.
I see your point, but..
Did you give it a thought why no one in the industry uses snubbers for a resistive loads?
And why there is a resistor used in parallel with a capacitor in snubber circuits?
In the scenario you just described - capacitor will do exactly that. It will prevent arc from happening.
But what about second scenario, where contacts are closing?
In this case capacitor will be fully charged (of course if voltage amplitude at it's fullest) and when contacts will start to come together and touch at some point,
there will be point when resistance between contacts would be pretty large (in comparison to closed contacts). But it would be enough to dump the whole charge stored in capacitor + current that flows through the heater and convert all that energy to heat at the contact point.
So, the current in that case would be many times higher, than normal contacts closing current and thus damage would be much bigger.
When opening contacts - you gain some, when closing contacts - you loose some..
But in case of inductive loads - snubbers really help when load is disconnected. They protect contacts from inductive "kick".
In this case a really cool solution would be to switch on and off the heater at zero crossings.
Just a simple MCU (attiny13 would do just fine), transformer-less PSU (couple of passive components), transistor and a relay. Since the relay switch on and switch off times are pretty much a constant across similar relay models - it'll be really easy to crank this thing up, and no additional tuning is needed. It'll work like a charm!
But in this particular case, with a cheap-o-heater - I guess, it'll be an overkill. But, anyways, it's not for me to decide what to do..
-
Use the switch to operate a triac.
-
Was scrapping a HP laser last week, and looked at the heater circuit. A lot of input noise and surge protection ( a lot of assorted chokes, capacitors and VDR's after the HRC fuse) and, on the line to the fuser heater ( thin film resistor) there are both relays ( gross control ) and triacs. The triacs have a snubber across them, a integrated unit of 100n and 33R. As close to a pure resistor as you can get, but has a RC snubber. This on a high dvdt triac as well, which theoretically does not need one with a resistive load.
-
Use the switch to operate a triac.
Triacs give away a lot of heat. Might as well need a big heatsink (and some room to mount it inside).
Was scrapping a HP laser last week, and looked at the heater circuit. A lot of input noise and surge protection ( a lot of assorted chokes, capacitors and VDR's after the HRC fuse) and, on the line to the fuser heater ( thin film resistor) there are both relays ( gross control ) and triacs. The triacs have a snubber across them, a integrated unit of 100n and 33R. As close to a pure resistor as you can get, but has a RC snubber. This on a high dvdt triac as well, which theoretically does not need one with a resistive load.
Most triacs need a snubber circuit for protection from mains power surges.
However there are snubberless triacs exist. They (theoretically) don't require snubber even with inductive load (as far as I remember).
Also snubbers sometimes used in relays even with resistive load to reduce EMI.
-
Triacs give away a lot of heat. Might as well need a big heatsink (and some room to mount it inside).
Not much of a problem in a device designed to make heat. Another solution might be to put a power PTC (such as an A/C solid state starter) across the contacts, such that it gets cooled off by the fan. When the contacts open, the PTC would shunt the current until it heats up.
Or if less heat output is OK, put in a series power diode so the probability of the switch opening when it is conducting is cut in half. That would also reduce the cycling that causes the problem in the first place.
The best solution would be to implement a PID controller using a microcontroller and triac, but quite overkill.
-
Time for a sing-a-long :)
Too Much Time On My Hands (https://www.youtube.com/watch?v=_JxiY6zGtGc#)