is there relay chatter on top or bottom limits?
what is the pump rating
you could drive a 3 phase line contactor of your relay ans share the load across all contacts
The issue we have is with an aircon condensate pump.
sauermann si 1730
Pump is basicly a solenoid piston pump.
Uses a float sensor for activation with a magnetic float and I suspect reed switches for level sensing.
The main control board consists of a small 240V to 12 transformer with basic rectification and filtering. The reed switch sensor effectively, through possibly an optoisolator, switches two relays, one for an alarm and the other for the solenoid pump.
The pump short cycles and is actually probably overrated for the aircon unit.
We did have a smaller pump originally but found on the real humid days at start up it would not keep up with condensate removal so the aircon man fitted this bigger pump.
The 12v/240V relay that switches the pump seems to burn out its contacts and fails every year. I have been replacing it with a higher rated set each time and am now up to 10amp 240v rating.
The solenoid pump does not operate direct on 240v but has a single diode in line for half wave rectification effectively giving it 50 pulses a second on the pump.
I have tried a solid state relay with zero cycle point switching but it was not successful I suspect due to the in line diode.
Would be interested in hearing your possible solutions.
Every time it ends up happening at night or on a weekend or when I can't get anywhere to get parts so have to run a hose into a bucket....
is there relay chatter on top or bottom limits?
what is the pump rating
you could drive a 3 phase line contactor of your relay ans share the load across all contacts
With both AC and DC, contact arcing can be minimized with the addition of a "snubber" circuit (a capacitor and resistor wired in series) in parallel with the contact, like this:
A sudden rise in voltage across the switch contact caused by the contact opening will be tempered by the capacitor's charging action (the capacitor opposing the increase in voltage by drawing current). The resistor limits the amount of current that the capacitor will discharge through the contact when it closes again. If the resistor were not there, the capacitor might actually make the arcing during contact closure worse than the arcing during contact opening without a capacitor! While this addition to the circuit helps mitigate contact arcing, it is not without disadvantage: a prime consideration is the possibility of a failed (shorted) capacitor/resistor combination providing a path for electrons to flow through the circuit at all times, even when the contact is open and current is not desired. The risk of this failure, and the severity of the resulting consequences must be considered against the increased contact wear (and inevitable contact failure) without the snubber circuit.
It can be calculated from the time constant of the load.
Rating of components very much higher due to the back EMF.
Same principal as Distributor Capacitor.
Kindest Regards, " The Druid ".
Thanks for your input gents.
The power rating of the pump is only 35w which gives a less than 1/2amp current draw, however being an inductive load I would expect that when the contacts break when the solenoid is drawing maximum current the back emf generated by the collapsing field would generate a reasonable arc. The relay is a circuit board mounted small device.
I was wondering if there is a SCR device that may be available on the market similar to the solid state relay I purchased for this sort of purpose. I could then do away with the in line diode and use the SCR to achieve triggering and half wave rectifying. Alternatively I may have to renew the relay and use it to trigger an SCR to supply the pump solenoid.
Last edited by beer4life; 10-01-10 at 12:49 PM.
Same as you I thought the problem was the Voltage/Current that the contacts see when the magnetic field collapses in the solenoid.
In physically small relays etc I've overcome this problem with Flywheel diodes in parallel with the solenoid coil.
As I said, I'm mechanical so what do I know.
additional variable, The voltage/current that the contacts see will also be influenced by the position of the armature inside the solenoid at the time the field collapses.
Last edited by BlackDuck; 10-01-10 at 01:17 PM. Reason: Additional thought.
I have my freezer controlled by an kit sold by Jaycar in Australia
Essentially it is an amplifier chip set up as a comparator.
When it gets up near to switching it chatters and this manages to eventually fuse the 240V relay contacts. Ive been messing around trying to get the thing to switch more cleanly but its not yet successful. A good hammer blow seems necessary to unstick the relay contacts
When its close to switching even turning on and off fluros in the vicinity set it chattering
Most Unsatisfactory-from an Electronics Australia magazine article.
I would not recommend them
My Family and my Property are not Government business. Governments should do what they were designed to do . Govern the issues best done by a central body.
To defeat corruption in the public service , give them three times the penalty.
In my last job I spent lots of time experimenting in ways to relieve relays and switch contacts from burning. It can be hard to stop and sometimes damn near impossible.
If you can lengthen the cycle time perhaps by running the condensate into a larger container and pump from there.
Also, You need a relay with a very fast mechanical response time and a larger contact gap to break away from the flame quicker.
Besides that, the following reading will give you advice on extra options :
Perhaps the most popular method of quenching an arc between
separating contacts is with an R-C network placed directly across the
contacts (same as Beerforlife has mentioned). As the contacts just begin to separate and an arc ignites, load
current feeding the arc will be shunted into the capacitor through the
series resistance, depriving the arc of some of its energy. As a result,
arc duration will be shortened and material loss will be minimized.
Contact Protection Diagram
Theoretically, the ideal arc suppression method would simply be a
capacitor placed directly across the contacts. However, with no resistor
in the circuit, when the contacts make, there is nothing to limit
capacitordischarge current. This nearly instantaneous discharge current
can generate a brief, but severe arc that may cause welded contacts,
depending on contact material and characteristics. Thus, the resistor is
necessary to limit capacitor discharge current. However, there is one
drawback. That is, the resistor tends to isolate the capacitor from the
very contacts the capacitor is supposed to protect. Because of this, the
amount of resistance should be kept as small as possible.
Many relay users are unfamiliar with the selection of a capacitor for arc
quenching service. To begin with, AC differs from DC in that AC crosses
zero 120 times per second for 60 Hertz service while DC, of course, is
continuous current. In AC service, the capacitor need not be as large as
in DC service because the AC arc will extinguish at a zero crossover
point. In DC service, the capacitor must continue to shunt load current
sway from the contacts until the contacts separate far enough apart for
the arc to extinguish.
Assume a DC application of 28 volts, 5 amperes. Further assume an
R-C network is needed that will result in contact voltage of perhaps 15
volts 1 msec. after the contacts have separated. Since the value of
resistance should be as small as possible, a 2 ohm resistor might be
chosen. At 2 ohms, peak capacitor discharge current will be 14 amperes
at time zero. Depending on contact material and size, this 14 amperes
may be quite acceptable for such a short period of time.
Conact voltage—that is, arc voltage—at any given instant of time is simply
the sum of the voltage drop of the resistor and the capacitor voltage.
Select a capacitor voltage of, say, 10 volts. The remaining 18 volts must
appear across the 5.6 ohm load and the 2 ohm resistor. Thus,
instantaneous capacitor current is:
I = E = 18V = 2.4 ampere,
and the voltage drop of the 2 ohm resistor is 4.8 volts. Arc voltage,
therefore, one microsecond after contact separation is 4.8V + 10V =
14.8V, or about 53% of supply voltage.
To determine the size of capacitance needed, the basic equation for
capacitor voltage may be used:
ec = E(1 - ε -t/RC)
Rearranging the equation to solve for capacitance gives 1.1 mfd.
equation to solve for capacitance gives 1.1 mfd.
c = ____-t____ = 1.1 x 10-6 farad
Where: t = 1 msec.
ec = 10 volts = capacitor voltage at time t.
E = 28 volts (for AC, use peak value).
R = 2.0 ohms.
The next question concerns capacitor construction. Can the capacitor
withstand discharge surge currents? When the contacts close, the
capacitor will discharge through the resistor. For a 1 mfd. capacitor and
a 2 ohm resistance, the time constant is: R x C = 2 x 1 mfd. = 2.0
To determine discharge di/dt:
i = C dv = 1 x 10-6 28 x .63 = 8.8Aavg/ msec.
.......dt ...............2.0 x 10-6
where; .63 is the capacitor voltage loss during one time constant
of 2.0 msec.
This di/dt isn’t very severe and a wide variety of capacitors should be
able to withstand it. However, the di/dt of a 5 ampere 240 volt AC
application would be 107A/ msec. at peak of the AC line—that is, 340
volts; and capacitor selection* should be made accordingly.
Of course, di/dt may be lowered by a larger value of resistance to limit
capacitor discharge current even more. But, the greater the value of
resistance, the less effect the capacitor has on the arc.
Other Arc Suppression Methods
For quenching DC arcs in certain applications, relays are available that
have a permanent magnet located in close proximity to the contacts.
The magnet repels the DC arc, thereby stretching the arc and causing it
to extinguish quickly.
Some relay users connect a diode across the inductive load to prevent
countervoltage from reaching the contacts. When the relay contacts
open, the storedenergy of the inductance recirculates through the diode,
not through the arc. While this is an acceptable method of protecting the
contacts, it does result in lengthened hold-up time of the inductive load.
For those applications that cannot tolerate lengthened hold-up time, a
resistor may be placed in series with the diode. The resistor does,
however, lessen the effectiveness of the diode and, usually, a compromise
must be reached by trial and error.
By using a zener diode in place of the resistor, hold-up time is greatly
reduced. This is because the diodes cannot turn on until the voltage
across them equals the sum of their voltage drops.
In some circuits, space is at a premium and there may not be sufficient
room for a zener and a regular diode. In such circuits, some designers
use a metal oxide varistor. The MOV performs in a manner similar to
back-to-back zener diodes. And, since the MOV is a bidirectional device,
it can be used in both AC and DC circuits.
An added benefit of arc suppression is the minimization of EMI. An
unsuppressed arc between contacts is an excellent noise generator. Such
noise can be troublesome to sensitive components in a circuit, or within
the RFI field. In worst-case conditions, EMI can cause unwanted turnon
of IC logic gates, SCRs, and triacs, and can cause damage to other
*Suggested capacitor atypes are metalized foil and film foil. Check capacitor specifications for
dv/dt and di/dt ratings.
• Measure the current inrush of the circuit to be switched before
specifying the relay.
• Never parallel relay contacts to double the contact rating.
Unless the relays are specially adjusted, they will not pick up
and drop out simultaneously. Even if they were to be specially
adjusted, they would not hold this adjustment over life.
• Paralleling Form C contacts may result in an unwanted makebefore-
• Contacts rated low level to 2 amps may be used to switch a 2
amp load. Once having done so, however, they cannot be
used to switch reliably a low level load.
• In a circuit comprising a series of open relay contacts (AND
logic), all but the last set of contacts to close will be dry.
Likewise, in a circuit comprising a series of closed contacts,
all but the first to open will be dry.
• The use of many relay contacts in series may be limited by
total circuit contact resistance.
• A “low level” circuit that pulls a capacitive inrush current or
develops an inductive counter emf is not low level. Worst
case circuit conditions dictate contact rating, not steady state
Hope this helps
Last edited by loopyloo; 20-01-10 at 02:08 PM.
I relented and bought a new smaller pump. I'll be limping for a month after the hit on the hip pocket.
However I am going to manufacture a small solid state relay replacement utilising an opto triac and an SCR. I will use the existing relay dc coil supply to switch the opto on with an appropriate resistor to limit current to the led. The triac will then trigger the SCR. I can do away with the series diode with the pump coil as the SCR will effectively do this and it will naturally switch off at zero crossing. Looking at the opto's and suitable SCR, resistors and a chunk of veroboard at Tricky Dicks should be able to make it up for around $15 -$20. The difficult part will be that the extra board won't fit in the existing housing so will have to look for another box to mount it all in.
I looked at the bigger reservoir option but there is little room in the head unit for this. There are addons that hang under the head unit but they look really out of place.
I obtained another relay for the moment to keep it going and placed a reverse diode accross it to see if this one lasts. 16a rating on the contacts. I also added a small capacitor accross the coil to help stabilise its holding voltage as I think it may be close to its limit being supplied by the existing circuit.
Thankyou one and all for your suggestions. I suppose my desired resolution differs from my thread title in that I will be eliminating the relay altogether.
Interesting reading loopy's thoughts on the subject.
Recently I was handed some circuit boards to fix. The description of the fault pointed straight at the board's relay and there's no prizes for guessing what I found when I ripped the relay apart.
I suggested that I could fix the problem so that it would not happen again (or so I thought) but I was told, Oh shit no... don't solve the problem ! We make too much money replacing the modules, and you would kill our best money spinner.
So I didn't bother too much about solving the problem, and I now think of myself as part of the problem instead of part of the solution.
I had pondered over the circuit which has a snubber in it. The values of the snubber circuit were not what I would have used, but never the less that's what was there.
Being 240VAC, it is rather limited to choices of snubbing.
I had considered some neon lamps in series with a striking voltage of >350V or some 400V geiger voltage regulators.
If the load was a particularly big one, I think I would have considered staging a series of relays with progressively larger snubbing resistances across the contacts.
I haven't put a lot of thought into it... as mentioned, I'm now part of the problem
There is a new type of arc suppressor available that solved my arcing problem.
I had a repeated contact burn out problem due to arcing as described in this thread.
I used Loopy's great post to do resistor and capacitor sizing calculations and built a RC snubber. (Thanks Loopyloo for the useful info in your post!)
Unfortunately the snubber did not completely solve the arcing problem as the relay failed again after 6 months of normal operation.
It lasted about 1 month longer with a snubber compared to the previous periods with no snubber.
Replacing the relay is not difficult - however been called out in the middle of the night every few months to fix this problem was annoying.
After some internet sleuthing (I Googled "arc suppression") I found a new type of suppressor called "NOsparc" available from an Aus company.
This NOsparc suppressor is new technology - it is not an RC snubber so it does not require any sizing calculations. You only need to wire it in parallel to the arcing contact.
The NOsparc suppressor seems to have fixed the problem - the problem relay has been operating without failure now for 9 months. I visually inspected the relay contacts today - and they seem to be in perfect condition after 9 months - whereas previously after 3-4 months you could easily see contact damage.
Last edited by beer4life; 29-11-13 at 06:21 PM.
The cost of the NOsparc arc suppressor was not a significant factor for the company I work for. The suppressor has already paid for itself by helping us avoid the usual shutdown that occured with relay burnout.
Actually - by avoiding even 1 hour of a shutdown - the device paid for itself 50 times over.
The shutdowns also always seemed to occur in the middle of the night (or at least it felt that way) - so the other benefit is that I avoid the 2am callouts to fix the problem.
Replacing the relay regularly would also solve the problem - but relay costs also add up - not to mention having to keep track and remember to replace the relay. Using a larger relay (higher contact current rating) was also tried - but this only delayed the failure as contact damage from repeated sparking slowly built up.
Anyway - to sum up - management did not gripe at all about the cost. Based on my limited experience in solving my own contact arcing problem - I suspect there are probably heaps of other companies that may benefit from these NOsparc arc suppressors.