Applicability of VFD to a small (1/4 or 1/2 hp) pump

BillB3857

Joined Feb 28, 2009
2,570
If you look at the spec and setup sheets for the "off the shelf" units, you will be surprised at the programability of these units. Everything from Volts/Hz, Max volts, Max current, etc, etc.
They can be programmed to match almost any 3 phase motor within their HP/voltage range.
 

Thread Starter

kingdano

Joined Apr 14, 2010
377
i will certainly do all my research before making a choice.

doing it with a motor control uC would benefit my growth in electrical engineering skills infinitely more than buying a unit off the shelf.

i am interested in bettering my skills first, company comes second.
 

Jaguarjoe

Joined Apr 7, 2010
767
Start with something simple. Learn how VFD's work with something that already works, like this:

http://tinyurl.com/29flshu

It has about 350 pages of parameters and set up instructions you can program into it. Far more than you'll ever need to spin the pump. The only thing lacking is the PID controller but you can get a Honeywell UDC to handle that.
 

Thread Starter

kingdano

Joined Apr 14, 2010
377
yes, something like that

my concern would be the tuning process, ie autotune that would not be available to me.

an off the shelf controller with auto tune PID would be the safest bet - paired with an off the shelf pressure sensor.

we will see how things go, but thanks for all the feedback - its hard to balance bettering myself with company BS and deadlines - its helpful to get a variety of opinions
 

Thread Starter

kingdano

Joined Apr 14, 2010
377

Jaguarjoe

Joined Apr 7, 2010
767
yes, something like that

my concern would be the tuning process, ie autotune that would not be available to me.

an off the shelf controller with auto tune PID would be the safest bet - paired with an off the shelf pressure sensor.

we will see how things go, but thanks for all the feedback - its hard to balance bettering myself with company BS and deadlines - its helpful to get a variety of opinions
Whatever controller you end up with-whether you buy one or build-make sure it doesn't suffer from "reset windup" and has bumpless transfer when switching between manual output and auto output.

Auto tune is very nice but you don't learn much from it.
 

Thread Starter

kingdano

Joined Apr 14, 2010
377
Whatever controller you end up with-whether you buy one or build-make sure it doesn't suffer from "reset windup" and has bumpless transfer when switching between manual output and auto output.

Auto tune is very nice but you don't learn much from it.

agree

but if i mess up the PID constants enough times we may think its not possible to solve this problem!

:D
 

Jaguarjoe

Joined Apr 7, 2010
767
Only three variables but they can really mess you up. First thing, put the loop in manual and see if you can get the pump to run at any speed and remain stable. If it does that it will work under PID control somehow or another.
30 years ago, I used to tune surge controllers on 39,000hp air compressors. 3ft diameter driveshaft that never, ever stopped turning. When the main motor stopped a hydraulic jacking system kept the shaft turning. Far out stuff.
 

Thread Starter

kingdano

Joined Apr 14, 2010
377
*sigh*

ive got a chemist/project manager telling me the best approach to this problem is to use a DC motor - despite all of the research i have shown everyone.

ill tell him how to make chemical reactions next, for act 2.


what a joke
 

Jaguarjoe

Joined Apr 7, 2010
767
what do you mean by "put it in manual" - is there a way to bypass the PID controls on most off-the-shelf units?
Yes, that's a standard feature. The PID controller's output is disconnected and a manually adjusted output signal is substituted. It goes from 0-100%. This is where you get in trouble with reset windup because the PID controller thinks it still has control of the process but in reality it doesn't. The PID output will wind up to 100% output or 0% output. When the PID does regain control it has a hard time getting back to normal because of the hugh offset between SP and PV. This is because of the "I" part of PID- integral or reset. Bumpless transfer allows you to switch from auto to manual and back again without a bump in the output which would upset the process.
 

Jaguarjoe

Joined Apr 7, 2010
767
*sigh*

ive got a chemist/project manager telling me the best approach to this problem is to use a DC motor - despite all of the research i have shown everyone.

ill tell him how to make chemical reactions next, for act 2.


what a joke
What kind of meth lab is he running?

I haven't seen every motor controller out there but from what I did see, VFD's were hands down more flexible.
 

Thread Starter

kingdano

Joined Apr 14, 2010
377
What kind of meth lab is he running?

I haven't seen every motor controller out there but from what I did see, VFD's were hands down more flexible.

i have done my research and presented a strong case for the use of 3-phase AC motor/pump bring controlled via pressure sensor feedback at the point of interest through PID and a VFD drive.

if the senior persons here are afraid of "new" technology - that is their problem. they can buy and wire up a dc motor controller for a pump on their own.

i wont half-ass the job and try to pound a square peg into a round hole when there is a clear precedent technically for the use of 3-ph AC VFD in applications such as this.

being experienced makes people stubborn and stuck in technology ruts - copy/paste engineering.

thats NOT what i do.

:D
 

Jaguarjoe

Joined Apr 7, 2010
767
You might have a hard time finding a pressure transducer. You'll probably want something like this but I'd bet it costs a fortune:

http://www.sensonetics.com/moldcav.pdf

Besides handling the heat, it should have open passages like this diaphragm unit has so wax doesn't have a chance to harden up and clog the path to the sensor.
 
Last edited:

Thread Starter

kingdano

Joined Apr 14, 2010
377
we are likely going to avoid measuring the pressure through the liquid flow, and measure the pressure in the airspace above the fluid in the reservoir.

we will have to (and by we, i mean the mechanical engineer) calculate the pressure drops to ensure correct regulation further down the line.

the air sensors are considerably cheaper and more common, i am told.

thanks for digging that up for me though.

i did manage to be enough of a stubborn jacka** to get my way.

proof of concept with off the shelf, and once/if it works do a proof of concept PCB design since this would be new to the company (3-phase VFD control)
 
Top