BLDC pump suddenly runs dry...what happens next?

Flyback said:

OK, but this is a Fixed speed pump.
Why wouldn't the following work for controlling this fixed speed pump?

Click to expand...

Some how this new control scheme does not make any sense!
I see some contradictions in the control loop.
Where did the speed sensor come from all the sudden!!!
As I said in the first few pages of this thread and was just mentioned on this page again, Your motor will not start like this!!!!!!!!!!!!
This setup is fundamentally flawed. You are trying to put bandaids on a system that can't work right.
Your best bet is to throw it all out and start over!
Sorry for shouting but you are just not getting it,
I never shouted before ever, but I thought you were more open to commonsense.
Even if you were able to deal with the pump suddenly running dry it still will not start and will have other problems.
You really need to consider running your brushless motor in a more conventional way.

This sim (albeit with crude modelling of the converter and ignoring motor commutation) shows the likely effects of a rapid reduction in motor load.
U2 provides current limiting at ~20A during the first part of motor startup (up to ~ 0.9 sec).
From 0.9-1.3 sec current ramps down as speed ramps up.
From 1.3 sec, speed limiting provided by U1 kicks in and acceleration is zero.
At 2 sec the load drops (pump begins to dry run), acceleration commences, but the speed control loop rapidly catches that and the motor supply current drops to a new level sufficient to maintain the fixed speed at reduced load. There is a corresponding drop in motor supply voltage.


Edit: One major problem which the ML4425 does not seem to cater for is a locked rotor (pump jammed/clogged). How will you cope with that?

I thought your company worked with LEDS?????? why are you talking about tupils and water pumps?? in another thread your making 5 million thingies for something else, so why do you grow tulips? Am I missing something? When did a PSU or LED light company start growing tulips?

p.s Dont forget my like for posting

4pyros said:

Some how this new control scheme does not make any sense!
I see some contradictions in the control loop.
Where did the speed sensor come from all the sudden!!!
As I said in the first few pages of this thread and was just mentioned on this page again, Your motor will not start like this!!!!!!!!!!!!
This how setup is fundamentally flawed. You are trying to put bandaids on a system that can't work right.
Your best bet is to throw it all out and start over!
Sorry for shouting but you are just not getting it,
I never shouted before ever, but I thought you were more open to commonsense.
Even if you were able to deal with the pump suddenly running dry it still will not start and will have other problems.
You really need to consider running your brushless motor in a more conventional way.

Click to expand...

Calm yourself 4pyros calm yourself, it distresses me to see you get all upset like this

LG said:

I thought your company worked with LEDS?????? why are you talking about tupils and water pumps?? in another thread your making 5 million thingies for something else, so why do you grow tulips? Am I missing something?[/QUOTE]

Yes, you are unaware of the way the real world works, sometimes.

Our esteemed OP may have changed companies.
It could be that he works for a "general engineering company" which does all sorts of things.
It may be that he is a freelance consultant, a bit like myself, or as I once described myself to an employee of a company where I had just arrived:
"I am not an employee here, but an independant consultant. You could describe me as an Electronic Prostitute"
At which he laughed and understood perfectly how I fitted in to the organisation!


LG also said:

Calm yourself 4pyros calm yourself, it distresses me to see you get all upset like this

Click to expand...

You are getting the idea LG.
Sometimes you just have to shake your head, walk away and take a chill pill.
And realise that nothing you can say or do will alter the situation.

Click to expand...

So he is a lady of the of the curb side that campaigns against certain light systems and grow's tulips? Is that about the size of it?

It was only recent with the 5 million thingies, and before that something to do with installing lights he was worried some fangled light system would run the world and make him unemployed. Fair do's stuff changes but TULIPS?

sorry to hear your a prostitute even of the electronic type, does that mean you work nights mostly?

Hey LG

How do you change Fonts so easily when you post ???

I battle with that

Regards,
tvtech

yes I am and proud of it!
Anyway, the fixed switching frequency schem can work, it just has to "soft start" up to it....and the method I illustrate is "the" standard way to drive 3 ph bldc's, ie by just switching the bldc coils one by one, and controlling the speed by controlling either the output voltage , or output current of the upstream SMPS.

Flyback said:

yes I am and proud of it!
Anyway, the fixed switching frequency schem can work, it just has to "soft start" up to it....and the method I illustrate is "the" standard way to drive 3 ph bldc's, ie by just switching the bldc coils one by one, and controlling the speed by controlling either the output voltage , or output current of the upstream SMPS.

Click to expand...

In what world is that a standard control scheme?
Can you show an real word examples?

The font thing is on the bar when you post 8th button to the left of the smiley face one next to the font height one

It really is obvious....I mean, in a high power bldc nobody is going to wire from the motor coils down to a current sense resistor on a PCB.....there would be too much wiring inductance.....you are going to put a hall sensor on the pcb upstream of the inverter/bldc, and regulate the current out of that, or regulate the voltage of the upstream smps...and limit the current with a hall sensor.....the hall will be on the pcb and you wont have to wire to it.

How will you address the locked rotor problem?
Have you determined from your lab work what the ML4425 does if it can't get the Back-emf loop to lock within the times set for rotor alignment and ramp-up? The datasheet is silent on this. I have a nasty feeling the commutation cycle could be left in an indeterminate state and an arbitrary coil could be left fully on at maximum current

alec_t said:

I have a nasty feeling the commutation cycle could be left in an indeterminate state and an arbitrary coil could be left fully on at maximum current

Click to expand...

That would solve one problem though
Then you could get a proper pump and controller instead of a experts interpretation of one
Now mr flyback why are you messing with tulips? Are they Led ones?

locked rotor's not hard to sense, for obvious reasons...then you just deal with it.

Flyback said:

locked rotor's not hard to sense, for obvious reasons...then you just deal with it.

Click to expand...

If its that easy then why start this thread.

locked rotor's not hard to sense, for obvious reasons..

Click to expand...

Quite. The column of smoke from the motor coil will be a dead giveaway

Flyback said:

then you just deal with it.

Click to expand...

You mean 'smack it with a hammer?'. How big are the water tanks? I do actually know a bit about fields and crops, I am a bit surprised by the irrigation system. It might be worth jumping on a farming forum and posting pics of the field, one thing I can tell you is modern ways of irrigation rarely involve much in the way of pumping, most are done with clever valves and properly prepared fields, then the rest is gates and channels .
Yours sounds very much like late 19th century farming,modern tractors pretty much all have the ability to have cheap electronic furrow levels fitted, very few fields these days are not dead flat unless you wanted them bumpy. We were not a crop farm, we did have some acres for feed though, we had a GEO pole in the corner of the farm, GEO poles send signal to the tractor and get info from satalights, basically the tractor knows where it is and how level the ground. Takes around 12 months to first prepare a field properly, after that its easy.
Modern flower growing is a long way from tanks and pumps the way you have them. Save the money and look into modern farm practice, you will up the %profit by a long way
LG