Help with Water Pump

[ronv]

Current Limit

Alec,

Maybe just a simple current limit like this one would work. You7 could or all 4 together and then AND it with say a 2 count and reset the "ON Clock"? Then an over current would kind of hic-up along and there would be time for the motor to start? I haven't read thru all the circuit so I'm not sure it would work or is easy.

 

[alec_t]

Ah, we're thinking along the same lines. I've done a rough draft which uses just such a source resistor and BJT, but am still working on timing and latching bits and bobs, with an option either to 'hic-up along' (i.e. auto-reset) on over-current or else latch off after, say, 2 secs and wait for a manual reset.
Rather than OR all 4 together I think individual over-current protection would be preferable, so that one pump out of action won't affect the others, this being a 24/7/365 system?

 

[ronv]

Here is another idea. Since we know max draw is 1.3 amps (Joe's spec. on page 1) and locked rotor current is around 10 -12 amps it could probably be protected with a resetable fuse like the 1.5A one here:
https://www.electro-tech-online.com/custompdfs/2012/06/Littelfuse_2920L.pdf

 

[salty joe]

@Joe
What is the adjustment range of your power supply?

19.2-23.8V

@Anybody
Does it make sense to try to modify the OEM controller to run at 24V or be adjustable?
That way joe can take a pump and run with it. Drawback, protection not as robust as intended.
Just looking at the idea if Joe could take a 24 3a supply or an adjustable V supply at 3A and run a pump 24/7 with the regulator being a modified OEM controller?

If it could be made to simulate a tide and run as a wave generator. What the controller that came with the pumps does is of no use to me.

Joe: Unlikely, but does your power supply have the luxury of a settable current limit?

The only adjustment is for V output.

 

[KeepItSimpleStupid]

@sj
There is a LT-blue and brown wire going to the pump. Do you know which one is positive and ground? Can you find out?

Is Lt-Blue negative?

@sj
I was just trying to get to a test mode to run a pump 24/7 with a current limited supply NOT use it for the design. It's just an opportunity to see if a single pump would survive.

I thought this was a 24 V supply, so why 23.8V?

 

[salty joe]

I just cut into the pump wire by the plug and found an orange and white wire. The orange is +. Different colored wires on different pumps. That's just goofy.

I re-checked the power supply and got a read of 24V. I must have had a poor connection at the probe or something.

 

[KeepItSimpleStupid]

@sj

So, your 24V is with the voltage adj at one extreme?

So, you don't have the wire from this pump? It's kinda important that I know what the polarity is. If you have the wire from the pump I have, all you have to do is plug it into an OEM controller and measure the voltage.

@anybody
I can't make heads or tails with the reverse engineering of the pump. The hall sensor has be baffled. All three lead sensors have a +V, -V and output. That output could be "programmable", open drain or open collector that I can find.

If the hall sensor had inverted outputs with a single power source, then maybe. e.g. V+, OutA, OutB or GNG, OutA or OUTB like a ULN2003 chip has, then maybe. Something where an external pull-up powers the chip. Otherwise I'm stumped.

My meter needs batteries. If the solder blob FET is good, it might shed some light.

4 lead sensor can offer a lot more options.

 

[()blivion]

I can't make heads or tails with the reverse engineering of the pump. The hall sensor has be baffled.

Just adding my two cents....

It should be one lead is V+, one lead is gnd, and the other is output, normally open drain/collector. The rest of the circuit deals with the switching and order. Note a few things (you prolly already have)...

1: One FET might be P-channel.
2: One FET might be depletion mode, not enhancement.
3: THERE *MIGHT* BE A BACK EMF COMPONENT HELPING SENSE AND SWITCH THINGS, ON TOP OF THE HALL SENSOR.
(yes, would be very very odd)

Some PC box fan circuits use a single hall sensor just to tell rotor orientation. Then back EMF for the actual switching and sensing. No matter what, the back EMF exists. They either had to account for it, or make use of it.

 

[ronv]

The only thing I can think of is that 1 FET is N type and the other P type. So when 1 is on the other is off and vise versa. Whatca think?

 

[()blivion]

The only thing I can think of is that 1 FET is N type and the other P type. So when 1 is on the other is off and vise versa. Whatca think?

... Minds = read?

We posted with in seconds of each other.

 

[KeepItSimpleStupid]

Sorry about the super horrible schematic

Print the jpg.
Draw over lines with a pen.
Put a junction dot under LT_Blue

I'll try to figure out what happened.

 

[()blivion]

Oh wow! I see. That is odd indeed.... thinking on it...... Maybe a raw Hall element rather than an IC?

Edit: Something like this will oscillate if the raw sensor has enough SNR....

View attachment 64845

 

[KeepItSimpleStupid]

That's the kind of concept I came up with, but I can't find anything.

 

[()blivion]

Yeah. Just sifted through a bunch of Hall sensor datasheets and I can't find anything resembling such a setup either. Everything in 3 pin packages are arranged with the normal pinout (Vdd, gnd, output). Also, the Hall element MUST be amplified in some way which makes the above circuit less likely. This is because the actual Hall effect is a few mVolts at best. Though strictly theoretically speaking, with a clever arrangement of transistors on a bare Hall element the above circuit is still plausible.

Looks like your going to need to take the sensor out of the circuit and play with it. No matter what the Gnd and V+ in should be correct based off the FET arrangement. So at least we should know the correct polarity.

Maybe we could ask the manufacturer what the sensor is and how it works? (HA HA, The idea that the manufacturer would help with this is absurd... yes)

 

[KeepItSimpleStupid]

Goto patents.google.com and type 6005320. Possibilities?

 

[KeepItSimpleStupid]

The sensor has had it. Partly my fault. Partly not.

 

[KeepItSimpleStupid]

Have an idea. Push pull outputs. https://www.digikey.com/us/en/techz...r-Hall-Effect-Sensor-ICs-for-Mobile-Apps.html and swap two leads on one of the FETs.

 

[()blivion]

The sensor has had it. Partly my fault. Partly not.

:sad days: D:

Goto patents.google.com and type 6005320. Possibilities?

Looks close! That circuit uses BJT's and a standard Hall sensor though. But it's almost the same arrangement as you have laid out above.


Salty Joe, When you get around to reading this could you possibly get some iron filings (hacksaw on a nail for a bit) or some black dirt or something and rub it on the spare rotor/pump blade? Also take a picture or two and post it if you would. I am trying to figure out how many poles the magnet has if we can, and this test might tell us. If you don't know what I'm asking I can elaborate farther or we can just forget it. But I'm basically just saying stick magnetic dirt all over the rotor part till it start's making shapes like this....

View attachment 64846

I'm guessing it probably has only two or four poles.

 

[salty joe]

I found the plug that got melted to the jack, but the wires got pulled completely out while trying to remove the plug from the jack. Really sorry.

 

[salty joe]

I taped the magnet to the bottom of cardboard. Hoping it's a good enough pic. Thanks all.