What is the waveform of the current that supplies this pump?

[treez]

Hello,
We have to design a universal input offline SMPS to supply the Whale Gulper 220 Pump.
Our SMPS will have a settable output voltage, either 12V or 24V.
The datasheet for the pump just says "24V, 2A" or "12V, 4A".

Though we do not know if this has high current peaks or not.
Do you know what the current waveform will be like?....will it have high peaks, and what will the peaks be? What will be the frequency of the drawn current?

The customer gave us a sample waveform, which shows the current into a 24V pump. -It actually shows a scope shot of
the voltage across a 100milliohm sense resistor in series with this pump....the waveform was a 2Hz (sort of) sine wave, (with a DC offset, so not negative going) which had a 5 Amp peak
bit which lasts for about 100 milliseconds.
We have no idea if the scope used was calibrated or if it was accurate.
I am of the impression that scope voltages are generally not accurate...so we have no idea if the peak current on this waveform is accurate or not.

Do you know what the current waveform into this pump will be like?


Whale Gulper 220 pump datasheet:
https://www.defender.com/pdf/503200_Whale_Gulper220.pdf
Whale gulper 220 pump details
https://www.defender.com/product.jsp?path=-1|51|2234226|2234234&id=1777934

 

[Dan Mills]

They will all be a little different anyway, what with unit to unit variations, never mind variations due to starting up into different amounts of standing head and the fact that sooner or later someone will switch the pump on with the seacock closed (Because that always happens eventually).

Design for 5A cont, 10A surge for a second @12V, half that at 24V and include short circuit protection, assume a brushed DC machine so work on the loads of RFI and spikes from the load assumption.

Regards, Dan.

 

[treez]

Thanks.

assume a brushed DC machine

How do you know its not a Brushless DC motor?

As far as I know, the current draw into a brushed DC motor is pure, flat DC?....it doesn't have pulsing like a Brushless DC motor?
I suspect the current waveform we have been sent might have been with an unstable SMPS or something, because I don't see how a 2Hz waveform can be going into any kind of DC motor pump?

Sorry but I also don't understand what is the "seacock", and also I don't understand what happens when its closed?. I assume the worst case condition is locked rotor, because that's just a short circuit across the 24v supply.

 

[--BawA--]

As the datasheet witnessed that
It's a DC motor ,so the current waveform would be like


Also current is property of passive source , so you have to design a SMPS, which can fullfill the requirements of the motor,
keep in mind that , during startup , the motor draws a huge current , so at that time your SMPS should not take a false decision of OVERLOAD and shut down , . For this ,you could add a delay circuit ,before taking an overload decision.

 

[Dan Mills]

Treez, the thing is a running a diaphragm pump, so I would assume the 2Hz is the diaphragm cycle time ~(The torque demand will follow a roughly sinusoidal shape depending on the design of the cam).

A seacock is a valve placed just inside an opening in a boats hull and is usually closed when the boat is to be left unattended (In case there is a problem with the internal pipe connections), they are always annoying to get to and folk often forget to open them before doing something that requires water to be pumped overboard.

Regards, Dan.

 

[treez]

I see, I also originally thought it was a DC pump, but it is of course, a Diaphragm pump as you say..
**broken link removed**

presumably there is no motor in it, but just a solenoid which pulls the diaphragm up and down?
So presumably there is none of the usual , initial high inrush current such as is assocated with DC motor pumps.?

Actually, I take that back, because the following shows that there is in fact a motor inside it, turning a worm gear which ends up moving the diaphragm..
33 seconds onwards shows motor action...
http://www.youtube.com/watch?v=zb-cE_oq_do


The motor is obviously moving slowly, which begs the question as to why the oscillogram we have shows a sinusoidal current with a DC offset, and at approx 2 Hz....?
The motor in the above video is rotating smoothly, and will no doubt have step down gears, and so there should be no current pulses at 2 Hz.?

So now if I may please return to my original question about what is the current waveform looking like?, and is it a BDC or a BLDC motor?..and any ideas to the magnitude and duration of the startup current?

 

[Dan Mills]

Different style of diaphragm pump, I suspect this one has a motor working a cam to drive the thing (Most semi trash pumps work this way).

Regards, Dan.

 

[FvM]

The motor is obviously moving slowly, which begs the question as to why the oscillogram we have shows a sinusoidal current with a DC offset, and at approx 2 Hz....?
The motor in the above video is rotating smoothly, and will no doubt have step down gears, and so there should be no current pulses at 2 Hz.?

Apparently you don't understand that a geared motor driving a diapraghm pump will result in a pulsating current. But just this happens. Believe it or not.

You should take the measurement results as fact, as long you have no own more accurate measurements.

The amount of inrush current can't be determined by pure guessing. These pumps are regularly supplied by lead-acid batteries, so short inrush currents don't particularly hurt. Knowing if it's a brushed or brushless DC also doesn't tell you about actual inrush current. Needless to say that the question can't be answered by guessing either.

 

[treez]

Thankyou, the fact that the motor will literally be like a short circuit on startup is a problem for us, as we are using the L6566B PWM controller, and this has internal short circuit and overload protection, which will unfortunately be tripped by the startup "short circuit" of the motor.
I believe that we should use a simple controller like UCC38C43B which has no short circuit or overload protection, and we should add these things in ourselves with external circuitry. Of course, as you say, we need to get a motor and actually see what the startup short circuit really is like in terms of magnitude and duration, and also , measure the motor winding resistance.

L6566B PWM controller
https://www.st.com/st-web-ui/static/active/en/resource/technical/document/datasheet/CD00167474.pdf

 

[FvM]

Seriously, I don't know the motor armature resistance and expectable inrush current.

If it's a electronically commutated (BLDC) motor, the driver might implement inrush current reduction. I saw "electronical reverse battery protection" mentioned in a Whale Gulper document, which would only make sense for a BLDC.

But as already stated, inrush current can be still as high as for brushed DC. In lack of other vendor information, you can only rely on the input fuse specification.

 

[treez]

Yes I see what you mean, and the fuse spec just says "automotive fuse"......these are slow blow, and they are so slow that the fuse value is not useful to us, because it doesn't tell us the magnitude of transiently high currents which could be say 30 Amps for 300ms.......this is short enough to avoid blowing the automotive fuse, but long enough to totally trip out the short circuit and overload protection within the L6566B control chip.
I believe that what we need is to implement eg a "dumb" pwm controller such as ucc38c43b, and then add in the overload and short circuit protection externally. A high secondary start up current is going to saturate the transformer which cannot be tolerated even for short periods of a few 100 milliseconds.

What we actually need here, is an SMPS with an output voltage regulating, plus current limited output. If the level of current needed to start up the motor is going to put more than 350mT in the transformer, then we simply have to upsize the transformer....this will not go down well, as the current power supply for this motor is a mains transformer, and it is quite small , since as you know, mains transformers can easily handle massive overcurrents and overloads......I am seeing now why switch mode power supplies are not in this way as good as mains transformers for supplying motors, as the smps cannot deal with the start up situation without it being vastly overspecced for normal running operation (when the motor has spun up to nominal speed).

They didn't tell us this at University.....still we should have worked it out for ourselves...we were just told how superior smps's were to mains transformers in every way.....I can now see that this is not so....mains transformers are very very good at what they do....ie supplying the type of extreme overloads and overcurrents required by motors at start up.

 

[FvM]

It must not necessarily be very difficult to the design the power supply, but there are some parameters that can be only acquired by running the real motor in a test setup. There will be e.g. a minimal current to overcome the motor and pumps break-free torque. If it's low, you may get away with a simple power supply soft-start. But you won't find it by "exegeting" the datasheet.

 

[Dan Mills]

A switcher could be fine, just use a current mode controller rather then voltage mode and design to run in constant current mode if the current demand exceeds 10A or so.

You could put an ohm or so of series resistance in line and then bypass it with a mosfet after a few hundred ms (series resistance starters are an old technology).

I was under the impression that all else being equal transformer core flux dropped with increasing load current? Of course the switch mode controller has to respect any inherent duty cycle limits that your transformer design imposes.

Regards, Dan.

 

[treez]

We are using a flyback smps, so our core flux increases with load current....I should have made this clear that we are using a flyback, my mistake.
I will do that, though if we find the motor needs a current that puts more than 350mT in our transformer then we will need to upsize the ferrite transformer.......this will almost certainly mean the smps PSU being bigger in size than the mains transformer PSU.

 

[Dan Mills]

Or you could go to some sort of push pull forward converter that does not rely on storing the energy in the core flux.

Flybacks usually end up needing rather oversized transformers (or huge gaps) once you get above 50W or so, better then to move to something forwardish, maybe cuk of some kind?

Regards, Dan.

 

[treez]

yes, I see your point, though we do need the isolation for 90-265vac connection.
Our average power, when the motor is running, is actually 60w, hence they thought they could get away with flyback.....though yes, the forward or bridge topologies are better at the overload situation....the one transistor forward springs to mind....though there is an issue with the coupling between the reset winding and the main primary...meaning a heavy primary snubber ends up usually being needed.

Of course, the other way is to have a flyback with say a 50v output, and then supply the motor from a 24v output buck smps which runs off this 50V. And having a big enough 50v capacitor bank to be able to supply the startup surge into the motor, via the buck smps.

 

[chuckey]

Why not use your SMPS to constant current charge a very large capacitor and switch the DC after it, so the capacitor supplies the motor's inrush current.
Frank

 

[Dan Mills]

LLC with the cap split and a couple of fast diodes so it has inherent current limiting?
Efficient, zvs over a reasonable operating range and if your magnetics guy is good you can design the resonating L as the leakage inductance of the main transformer.

Flyback is utterly wrong for this use case.

Regards, Dan.

 

[treez]

Now we are seeking to setup a lab test jig which allows us to see the maximum power draw of this pump in its usage.....do you know if the following lab setup sounds suitable?....

As before, we are designing an SMPS to power our whale gulper 220 pump. We obviously wish to know what is the maximum electrical load that the pump presents. (ie the maximum pump current and voltage) We have a setup whereby the pump sits on the floor, and pumps water from a bucket up a 1.5m vertical length of tubing rising up above it, and then this water drops down another length of tubing, back into the bucket which sits on the floor, alongside the whale gulper pump.
Is this likely to present the maximum electrical loading of the pump?
(the tubing is approx 1 inch bore)

If this is not the maximum loading, then do you know how we can set up a lab setup whereby we have the maximum loading.?

The pump datasheet says its 24V and 2A, but that is the average, and the pump is a diaphragm pump which has peaks and troughs in the power flow. We have a scope shot of unknown origin which shows 100ms long intervals where the pump draw 5A from its 24V rail. We do not know if this represents the maximum load.
Presumably if the pump is pumping heavy sewage then it will draw more current?

Whale gulper 220 pump.
https://www.whalepumps.com/marine/p...=10021&Product_ID=10037&FriendlyID=Gulper-220

 

[Dan Mills]

Looking at the data sheet you want a total of 4 meters of height, maybe 2m of suction line below the pump, and 2m above, and it is important that the flow from the pipe above the pump does not end up forming a siphon with whatever hose you use to take that water back to the bucket.

I would also suggest that 3/4 inch tube be used as that is the smallest the datasheet specifies as suiting the fittings (3/4 inch is a MUCH smaller cross sectional area then 1 inch).

You also need to ensure nothing bad happens with a blocked outlet pipe (With positive displacement pumps like this that can result in a locked rotor), obviously it does not need to work, but nothing should fail catastrophically, also a partially blocked outlet pipe should not cause overheating, motor current will rise as the partial blockage gets more severe, you need to trip if sustained current becomes excessive and retry once things have cooled.

Come on treez, this is basic datasheet stuff look at the specs and figure out the worst case scenarios.

Regards, Dan.