Inductive AC load is killing my stamp...

billmi

The Application:
Controlling an 110 VAC diaphragm pump (basically a coil with a magnet connected to a diaphragm, so the magnet moves back and forth at 60hz causing the diaphragm to pump water) to deliver water to a saltwater aquarium.

The Circuit: I'm using a 9VDC wall transformer to power a BS1-IC. The stamp has a couple of control buttons, a variable resistive bend sensor activated by a water level float, an output signaling LED, and a MOSFET to control a small relay. The NO contacts of the relay can close to connect the hot leg of a controlled AC outlet, the neutral leg connected full time.

The Software: The app I've written in PBASIC lets me prime the pump (run continuously while button is held to purge air from the pump lines after changing sources,) refill the tank with salt-water after draining for a partial water change (continuous pump until float indicates full,) and slowly add calcium hydroxide laced refill water at 1 drop per second (a 50ms pulse of the pump delivers one drop) to compensate for evaporation, but doing it slow enough to not change Ph too fast.

The Problem: By itself, without the pump plugged into the controlled outlet, the circuit works properly, and the stamp has no trouble controlling it. When the pump is plugged in, the stamp has no problem turning on the pump. About 30% of the time when the stamp turns the pump off, it resets. But it's not just when the stamp cuts power to the pump... I can plug the pump into a wall outlet across the room on a different circuit, and about 30% of the time, when I unplug the pump from the wall, it kills the stamp.

The stamp is not the only thing affected either, when power is cut to the pump, when near the aquarium, the magnetic ballasted florescent aquarium lights blink dark for a moment.

So... I need to either block whatever RF/EMF is being caused by the pump entirely, or properly protect the stamp from it. I'm probably going to add a CDS photocell to the stamp, so that it only drips in the limewater at night anyway, so interference with the lights won't be a big deal.

Any advice would be appreciated.

Thanks.

skylight

The pump could be causing a voltage sag in the supply due to high starting current in the motor of the pump as indicated by the aquarium lights dipping,

you could·feed the pump on a circuit downstream to the power supply for the control circuit ie stamp or have a reservoir capacitor supplying the power rails for the stamp and control circuitry to allow for the momentary dip or alternatively run the stamp from a maintained battery supply or ups.

sam_sam_sam

billmi

If i where having this problem i would use something like this·and put to the basic stamp board with a LM317 reg in the current to charge the battery's

I would use two of these

http://www.allelectronics.com/cgi-bin/item/NCB-322/160300/3.6V_380_MAH_NICKEL-CAD_BATTERY_PACK_.html

or two of these

http://www.goldmine-elec-products.com/prodinfo.asp?number=G14949

or four of these

http://www.allelectronics.com/cgi-bin/item/NMH-16/160/7.2V_40MAH_NIMH_COIN_CELL_PACK_.html


▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
··Thanks for any··that you may have and all of your time finding them

·
·
·
·
Sam

Post Edited (sam_sam_sam) : 2/4/2008 1:40:52 AM GMT

Beau Schwabe

billmi,

Do you have a reverse biased diode across the Relay /and the Mosfet transistor?

It may not be the Inductive AC load itself, rather than the back EMF generated from the relay coil when the unit turns off.

A steady "isolated" supply to the BS1 might be necessary also...

http://forums.parallax.com/attachment.php?attachmentid=39280

Reference Thread:
http://forums.parallax.com/showthread.php?p=553973

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Post Edited (Beau Schwabe (Parallax)) : 2/4/2008 5:55:35 AM GMT

billmi

Skylight - I don't suspect that motor starting current is the issue, since the pump has no motor, and the problem is when the pump stops, not when it starts.

sam_sam_sam - That's been one direction I have been thinking, basically a little battery back-up. Giving this more thought, I suppose I should run the circuit on a 9v battery for testing, to confirm that a power dip is the real problem.

Bau - Yep, I did add a diode across the relay just in case that was the problem (though I've not had it with similar sized relays that I use to flip-flop current direction on my model railroad.) But I don't think it's backwash from the relay coil, since the problem doesn't happen when the stamp is opening and closing the relay without the pump plugged into it, and the problem happens if I plug and unplug the pump in a wall outlet across the room (i.e. the stamp and its relay are not controlling the pump at all, just unplugging the pump from any wall outlet in the room knocks out the stamp.) I'll try a 9v battery to supply the stamp, and if that solves the problem, that isolated supply circuit should do the trick, I hope.

skylight

Instead of a dip it could be a spike then? This would suggest a heavy load that when switched off causes a voltage spike thats killing your stamp, as suggested use a different supply.

billmi

Well, it looks like it's neither a power dip or a spike - it appears to be some sort of RF issue, or invisible gnomes.

I did what I should have done earlier on - I isolated the stamp, powering it by a 9v battery, and connected nothing to the attached circuitry (i.e. there are no wires from the stamp and its circuit connected to anything else, not to the pump, not to AC power, nothing.)

Running a test program on the LED that pauses for 5 seconds, then rapidly flashes the LED (the pause then becomes the indicator that the stamp has reset) I left the stamp running, and turned the pump on and off by plugging it into the wall outlet and unplugging it. Same results. About 1/3 of the time, when I unplug the pump from the wall, the stamp resets. The effect continues for a range of about 20 feet between the stamp and pump, and then begins diminishing after that.

I've shielded the stamp in a cage of grounded heavy gauge aluminum foil, and done the same with the pump - neither solved the problem.

The pump that vexes me: http://www.aquatichouse.com/Pumps_files/aqualifter.asp

Barring some amazing epiphany, I'm simply going to accept that it resets the stamp. I've re-written the code so that the first thing it does is a level check, then performs delivers a pulse of makeup water - if the stamp resets, it just goes though the loop it would do anyway if no buttons were pressed. The only difference is I don't have as large of a time window for wave elimination (multiple checks of the level sensor to ensure that a ripple in the water doesn't result in a false low-level reading.)

skylight

Its incredible after doing everything you have to avoid the problem it can still affect the stamp from a distance of 20ft! you say you have totally isolated the stamp and pump even the ground connections? Is there an earth path say from a metal worktop or something that is still connecting them somehow?

If as you say its an RF problem then the grounded shielding you made up should have stopped any interference, its quite baffling??

is the reset pin on the stamp connected?

Is the battery powering the stamp fresh? If the power supply drops below 4.2v the stamp will reset.

Bruce Bates

skylight -

None of the following may be the cause of your problems, but let's get them out of the way none the less.

Tell me more about the receptacle where the pump is getting its power. Is it ground fault protected? Is it an older circuit, or a fairly new one (relatively speaking)? Is the receptacle a 3-wire (grounded) or 2-wire (ungrounded) one?

It seems reasonable to presume that the fluorescent lamps on the aquarium are on the same circut as the pump. Can you confirm this? Is there anything else on this circuit (aquarium heater, etc)?

Regards,

Bruce Bates

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"Genius is one percent inspiration and ninety-nine percent perspiration."

Thomas Alva Edison

skylight

Bruce, Its Billmi that posted the problem, i'm just replying.

Beau Schwabe

billmi,

Even though your BS2 is isolated, do you have a programming cable connected or anything such as long leads that could be acting like an antenna?
Try placing a 1K resistor from the ATN (Pin #3) to Vss (Pin #4). This will help attenuate any noise that might be persuading the BS2 to reset.
I've had similar experiences with a BS2 controlling a large industrial vacuum system I put together several years ago... Once I isolated the cause, the 1K resistor solved the problem.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

billmi

Bruce,

When the pump is affecting the lighting, they are on the same circuit - haven't tried different circuits to see if it does.
The problem exists during testing, on my workbench, but once installed, the stamp, pump, circulation pump, filter pumps, and aquarium lights are all drawing power from the same GFCI outlet (test bench is GFCI protected as well) and all the house wiring is 8 years old.

Beau,
Yes, there are long leads. Even when the programming cable is disconnected, the leads to the sensor head (water level, and aquarium lights - just changed that this morning to IR break-beam, since the bend sensor didn't last in the salty air environment) are 4 to 5 feet, and there's a couple of feet of lead to the relay that turns the pump on and off. It's a BS1-IC, not a BS2, but I assume that the same thing could happen, enough current inducted to trigger a rest, yes? For the BS1-IC, it's a low reset, not a high reset, and the manual says not to drive it high. Is there a similar approach that I could take though to make sure it doesn't get pulled low by EMF?

For the meantime, since it's better to drip calcium supplemented water at night, I've added a CDS cell to detect if the lights are on (day) and the circuit can just run as night, with it's main loop such that the last thing to happen in the regular cycle is the pump shutting off, so if it does reset the stamp, it just starts the loop again, the same as it would anyway.

skylight

I can understand you getting a problem if all is connected as above and I would suggest as previously stated that it is either a dip or spike or noise·in the power circuit hence the flickering aquarium lights·or maybe a loose connection·but you said it also happens when the stamp is totally isolated from everything else?



Have you thought about putting a mains filter across the pump?

Post Edited (skylight) : 2/17/2008 4:17:45 PM GMT

David H.

It sounds like a EMF issue caused by inductive kick when the pump motor is shut off. The best way I know of isolating this would be to use a solid state relay. Good luck.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
David


There are 10 types of people in this world,...
Those that understand binary numbers, and those that don't!!!

billmi

Skylight - yes, it happens when totally isolated.

When the stamp is powered by a battery, I can put it in a cardboard box, no part of the stamp or its circuits, or connected wires, coming out of the box.

If I plug the pump into a wall outlet in the vicinity of the box, about 1/3 of the time when I unplug it, it will reset the stamp.