Help with Water Pump

[KeepItSimpleStupid]

Remember that the OEM controller has the LC ripple filter composed of the 68uH and 470 uF and this https://www.premmagnetics.com/Inductors company sells inductors to mortals.

 

[KeepItSimpleStupid]

Kinda interesting: http://www.marinedepot.com/Hydor_Ko...or_Aquariums-Hydor_USA-HD00588-FIWMEW-vi.html

The manual is in the above link too.

and a review:

**broken link removed**

It is a 12 V powerhead and looks like the controller does what it's supposed to do or what Joe wants. It apparently has problems with insulation breakdown and doesn't have enough Omph for some. It does appear to be a more reasonable price. Nice connectors, but they put them in the wrong place. The front? If you want high power, then 24 VDC is the way to go.

It's just an interesting read.

Here is a really interesting paper: **broken link removed**

 

[KeepItSimpleStupid]

ronv was in this thread about a motor. Here is the OP's comment at the end of his escapeds: https://www.electro-tech-online.com/threads/mysterious-self-destructing-lm317s.128302/#post1068216

and I did suggest PWM which was the right way. He does suggest a fast diode across the motor as well.

 

[alec_t]

A lot of headaches trying to build a homemade buck converter.

Yes. And a lot to modify an OEM controller, or build any other driver .

Spice doesn't model inductor saturation so it is easy for the current to be out of spec (for the inductor) during start up

Does that matter? We're concerned especially with a locked rotor (i.e. a permanent start-up state), which is essentially a DC situation where current is determined mainly by the coil resistance.

How about just latching the over current in the one you have so that it must be reset by hand on a failure?

That was a feature in one of my earlier designs (I've lost track), but Joe asked for an auto-reset solution.

 

[KeepItSimpleStupid]

Does that matter? We're concerned especially with a locked rotor (i.e. a permanent start-up state), which is essentially a DC situation where current is determined mainly by the coil resistance.

Locked Rotor conditions are detected by the IC
when there is no motion for one second and will shut off the motor drive for five seconds. Then, the IC will turn on the drive current for one second. This sequence continues indefinitely until the locked rotor condition is fixed. This feature prevents overheating

So, if this is the IC that's used (the no Vdd is patented), then is Locked rotor a DC condition or is it 1 second on, 5 sec off? I guess you could find out at the expense of a possible ruined motor. Use the OEM controller set at say 10.25 V and hold the rotor. See if a 1 sec on, 5 sec off is felt and HOPE that the motor survives the test.

 

[alec_t]

is Locked rotor a DC condition or is it 1 second on, 5 sec off?

1 second is an eternity for our electronics and I consider it a DC situation, compared to the 50kHz switching period, for design purposes.
BTW, does the OEM controller always start up at 10.xV, or is it random which of the 3 speeds it starts? If it's always 10.x then that answers ronv's question as to whether the pump will start reliably at that voltage. If not, then a test by Joe with a 12V car battery (pump in water!) would be a useful indication, especially if we also knew the start-up speed (roughly) at 12V.
In one of those links above I gather that 'feed mode' = pump OFF for 5 mins. So speed control not needed?

 

[KeepItSimpleStupid]

This explains various modes: https://www.advancedaquarist.com/2010/12/aafeature#section-6 That particular controller slows the speed for up to 10 minutes.

Some pumps have a night mode too, but I think that one is for the mortals to cut down the noise at night.

Fish can't regulate body temperature and they are very sensitive to temperature. Turning off the pumps as well as slowing them down may cause the fish to feel a sudden temperature difference so a slower speed is better than off.

 

[ronv]

A lot of headaches trying to build a homemade buck converter. One thing to worry about is the start up current. Spice doesn't model inductor saturation so it is easy for the current to be out of spec (for the inductor) during start up. It might be better to use a chip like the LT one you used before that takes care of all that "stuff'.

Below is what I was talking about. Note the inductor current as it charges the cap.

How about just latching the over current in the one you have so that it must be reset by hand on a failure?

Yes, I remember it as a nice feature. If I have time today I'll go through the posts and see if I can generate a dream spec. With a reset to a blank sheet of paper it might not be that hard to meet.

 

[alec_t]

That sort of current is experienced by the inductor in the OEM controller, so presumably the inductor is rated accordingly or there'd be lots of irate customers

 

[ronv]

That sort of current is experienced by the inductor in the OEM controller, so presumably the inductor is rated accordingly or there'd be lots of irate customers

The IC limits the current a plain pwm has no such safety.

Here comes the spec.

 

[ronv]

Spec summary

Drive a pump with a running current of 1.3 amps at 24 volts.
Run and not run time is adjustable
Pumps are on in pairs. Total of 4.
Bulk supply is 24 volts at 15 amps with small adjustment range. (20.8 to 24????) Post 115
Soft start in the order of a few ms. Post 167 This is probably the start time of the controller switcher.
On for 30 seconds, off for ~ 5 - 10 seconds. #178
Reset switch on failure instead of fuse #178
#229 6 hours run time 6 hours off time. Flick or short pulse on to keep fish away from the pumps that are off.# 270 ability to reduce flow.
# 692 auto reset on failure.
Locked rotor protection.

Methods and drawbacks:
Basic switch. No locked rotor protection

Linear regulator - Initial Current limit set only by IC spec. Does not completely shut down on failure (1.25 volts) Repeated retries would probably result in thermal shut down.
Voltage adjustable

OEM controller modification:
Advertized as unreliable.
Initial current limit set by IC spec.
Starting Current limit not adjustable.
Voltage adjustable

Current limited FET switch.
Voltage not adjustable
Thermal problems on auto retry.

LT1074 Switcher
More robust than OEM IC.
Adustable current limit for initial start up
Adjustable voltage
More dificult to build on a proto board - but meets all requirements (I think)

 

[KeepItSimpleStupid]

OEM controller broke when turned on/off within a short time

That little test you asked me to do, trashed the OEM controller.

The fuse didn't blow because my supply is current limited. The input is effectively shorted, so I SUSPECT the regulator IC. I did see things like the power LED was dim and then none of the Wave LEDS were on with a dim power LED.

The only real troubleshooting I did was to check for shorted capacitors and I checked the fuse in circuit.

 

[KeepItSimpleStupid]

I think eventually there will be 6 pumps, 4 for the wavemaker and 2 for the tide maker.

I believe Joe has something like an 8 and 15A supply at 24V. One can adjust V a little lower than the other.

On for 30 seconds, off for ~ 5 - 10 seconds. #178

4 pumps, 2 at a time.

#229 6 hours run time 6 hours off time. Flick or short pulse on to keep fish at bay.

2 pumps, 1 at a time, other pump flicks at some rate.

Is "Feeding" appropriate for both the wave and tide controller? Wave - definitely. Tide - not so sure.

I saw a post somewhere in this thread that flick was used in the wave controller too. I just don't know want the flick cycle timing is supposed to be.

Do you want 3 pumps to "flick" at once?

OEM controller modification:
Advertized as unreliable.

Now confirmed unreliable.

Basic switch. No locked rotor protection

Still not sure if true or not. Overvoltage may have been the culprit.

 

[ronv]

That little test you asked me to do, trashed the OEM controller.

The fuse didn't blow because my supply is current limited. The input is effectively shorted, so I SUSPECT the regulator IC. I did see things like the power LED was dim and then none of the Wave LEDS were on with a dim power LED.

The only real troubleshooting I did was to check for shorted capacitors and I checked the fuse in circuit.

What test was that Kiss?

 

[ronv]

I think eventually there will be 6 pumps, 4 for the wavemaker and 2 for the tide maker.

Joe can you fill in what the 5th and 6th pumps do?

Do you want 3 pumps to "flick" at once?

Still not sure if true or not. Overvoltage may have been the culprit.

Either way....... 10 amps thru the pump will fry it.

 

[KeepItSimpleStupid]

The test

Was to see if the OEM controller came up at the same voltage each time. Now it does. Basically off.

The 5th and 6th pump make a tide controller. Place say one at each end of the tank and call them left and right. The left one is on for 6 hours and then the right one is on for 6 hours. During the time the other pump is "OFF", it "flicks" or turn on briefly so the fish don't go to sleep next to a pump that could blow them away, literally. REPEAT So the tide pumps push the water to one side of the tank for 6 hours and then push the water to the other side for 6 hours. UNTIL

Waves are a rapid motion, tuned to the resonance of the tank. I'm sure that would annoy the fish when they are feeding, They would go "gulp" and the food would move away.

Some of the other controller stagger the start-up of the pumps by a second or so even when they are supposedly working together. "Flicking" out of sync is good for the electronics and the controller may have to.

That was one of the reason Joe bought two supplies. One for the tide controller and one for the wave controller, so they re basically sized appropriately.

I remember the wierdest things, but the "flick time" is a specific number. I don't remember that easily.

 

[ronv]

Wow, so it just went belly up.

So I guess the other 2 pumps don't take any more logic they just need drivers.

 

[salty joe]

What I had in mind was two seperate controllers. One for tidal simulation that would control two opposing pairs of pumps (4 pumps). One pair would run 6 hr and be able to be adjusted for about 30 sec on, 5 sec off. The idle pair of pumps would "flick" every minute or so to prevent fish from sleeping by the pumps. This feature is critical. The ability to reduce flow would be nice, but not critical. BTW, very nice spec summary Ronv-thank you.

The other controller for a wave would control two opposing pumps where only one pump ran at a time and having an adjustable run time of about 1-2 sec to about 15 sec.

Hey KISS, do you want another factory controller?

I'll hook a pump to my car 12V car battery and let you guys know what it does. I was just going to take a bucket of water to my truck and toss the pump in and not bother disconnecting the battery. That's OK, isn't it?

Thanks guys.

 

[()blivion]

I was just going to take a bucket of water to my truck and toss the pump in and not bother disconnecting the battery. That's OK, isn't it?

Should be. Just don't run or start the vehicle while your doing it. You may want to disconnect one cable too while your at it. But if your careful it shouldn't matter.

What I had in mind was......[What salty had in mind]

If you could give a quick drawing, a rough layout, or maybe some pictures of your unfinished tank and what you want to do, that might be nice for us to have. It's not required though and I don't actually *KNOW* how it can help. I just want to look like I'm doing something useful I guess.

Current limited FET switch.
No current limit on initial start up.
Voltage not adjustable
Thermal problems on auto retry.

I don't think initial start up current is going to be too much of a problem with this design myself. Unless of course the rotor is locked and the coil fully saturates. Even a pump with shorted FET's is going to naturally oppose high currents for a time, do to inductance of the coils. But a simple over current shutdown would fix all of this easy. I think one of alec_t's current limited FET switch designs already did this well enough. something like 6 Amp instant shut down wasn't it?

I do however want to emphasize my opinion that what ever design is eventually used, it *MUST* have all of the current limiting ever thought up...... ever. And basic transient HV protection should also be added, as we are dealing with FET's and CMOS logic. It would also be nice to add a water contamination shut down of some kind. As long as it's not to hard to do. A way to directly trip the PSU's protections would be nice, but that would require modding a mains voltage PSU, which in it's self is not 100% safe. Also, if salty goes nuts with the conformal when he is done, then this will not be as much of a problem. A plastic coating tends to repel water pretty good.

You can never have to much safety.

 

[ronv]

Thanks Joe,
The reason I ask about the other pumps is because it may be easier to use some of the existing circuitry for the second controller. Is that okay?

Trying the pump on the battery probably won't tell us much as the battery voltage is already as high or higher than the current limit in the factory controller. The "Kiss modification" to the factory controller would let us dial it in using the old reliable ohms law. In other words we would like to use as low a current as is reliable to start the motor. Since we know the resistance of the motor at rest to be 2.15 ohms we would know that if it started at 6.5 volts we could limit the current to 3 amps, 8.6 volts 4 amps and so on.