# 4-20mA scaling ???????

Discussion in 'Electronic Basics' started by Pablo, Aug 17, 2004.

1. ### PabloGuest

I am working on a project where I have 1 DC drive which outputs a 4-20mA
signal to a VFD, I want the VFD to follow the DC drive which it does,
however I want to be able to scale the VFD. That is if the DC drive is at
say 50% then the VFD would be at 50%, but, in some cases I want the VFD to
be say 45%. I have gone through the programming and found that the DC drive
has no means of scaling the 4-20mA signal and the VFD would be too
complicated for my production people to program each time they wanted to
change scaling. Is it possible to adjust the 4-20mA signal in some way, say
with a potentiometer or some device. I would like to do it this way so as to
make it simple for my operators. This setup is basically a machine which is
driven by a DC drive and a colour feeder which sends a pigment into the
machine, when we push the output up or down on the machine the colour must
follow, but as you may have guessed the colours are fed at different
percentages. Thanks to all whom reply. Paul

2. ### CFoley1064Guest

Subject: 4-20mA scaling ???????
Hi, Paul. I kind of wish you'd given more information, specifically the
manufacturer and model number of the VFD (variable frequency drive, used to
vary the speed of AC motors by varying their frequency). Nonetheless, there's

Quite a few modern VFDs have a load resistor at the input, typically 500 ohms
or so, and use an analog-to-digital converter and software to interpret the
signal for 0-10V, 0-5V, 0-20mA, or 4-20 mA. If you're set up for 4-20 mA, you
typically also have the ability to program minimum current (the current reading
below which you'll get 0 Hz output) and maximum current (above which you'll get
100% of programmed speed). If you use a potentiometer set up as a shunt
resistor, you'll cut the slope of your input current, but you'll also bring
your minimum current well below 4 mA. That may cause a fault lockout on your
VFD (which you also might be able to program around).

If you're going to do the shunt pot route, you need to figure out the minimum
current necessary to keep a fault lockout from occurring, and also how much
attenuation you can deal with when the pot is turned all the way up. Here's an
example based on the assumption of a 500 ohm load resistance at the VFD input
(view in fixed font or M\$ Notepad):

VFD Current Divider
.----------o------.
| | |
| V D |
| - | .-------------.
o+ | | +| VFD Input |
4-20mA V D '--------o--. |
signal - | | |
o- | | | |
| V D | | |
=== - | .-.500 ohm |
GND | | | |internal |
.-. | '-'resistor |
10K| |<---. | | |
| | | | | |
'-' | | | |
| | | | |
o-----' .--o--' |
| | -| |
.-. | '-------------'
220 ohm | | |
| | |
'-' |
| |
=== ===
GND GND
created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

Use a 1N4001 or Si diode of your choice. This circuit has a couple of

* Because of the diodes, it won't start attenuating until there's about 1.8V
across the 500 ohm resistor (which will reduce your problem with error lockout)

* When the 10K pot is set for maximum resistance, you'll have minimal
attenuation of your signal. Of the 20mA, 19mA will go into the VFD. Actually,
you should be able to program around this with your maximum current setting.

* When the pot is at minimum, less than 1/3 of the current will go into the
VFD input. That would mean that 20mA output from your signal source will give
less than 33% full speed.

The only real negative of this setup is that almost all the control will be in
the first 1K of resistance. If you want to invest in a 10-turn pot with a
calibrated dial, that'll help a lot. You can also tell the machine operators
where to set the dial (that's assuming anyone on your factory floor can set a
verneir dial).

On the other hand, this is a simple and relatively bulletproof setup that
doesn't require external power, and that a maintenance person will have a hard
time messing up.

If you need/want a more complicated solution, you'll need an external power
supply and a couple of op amps. Have fun, and

Good luck
Chris

3. ### Louis BybeeGuest

This may not be the best group to pose the above question. I suspect it

alt.engineering.electrical or sci.engr.control

You would be much more likely to get appropriate responses to questions if
you were to provide more complete information/parameters surrounding your
situation.

Is this project for a manufactured system, or one up testing?
Building one, or multiple quantities?
Will you leave this scaling capability in the finished product, or is this
just for testing?
What kind of a budget do you have.

The answer to the above questions could drastically alter the answers to you

A few methods to accomplish what you are after (might not be the most
appropriate depending on the above answers) are:

There are some interesting electronic controllers available that have
programmable capability, and scaling the input in relation to the output is
easily accomplished. You could even compress the full swing of the input
into a section of what the normal output would be.

Another option would be the use of a small PLC. This could be quite
effective if your application is for testing on an assembly line. You could
write a table in the PLC to reflect your scaling needs. You could even write
a number of switch selected scaling tables that could be selected remotely
by unskilled assembly workers according to the needs of the module they were
testing.

Louis

4. ### PabloGuest

Got a good answer, basically the slave drive is as basic as I have ever seen
a VFD, almost no adjustable parameters, which is ok as it it not necessary
in most applications. I can either go with a signal isolator, like a KB
Electronics type where I can scale the output via a potentiometer or I can
get an operator interface for the master drive which can scale the output
internally. We are going with the operator interface as we are looking at
about 8 machines and the interface is capable of this and is capable of
recipe inputs. Software for the operator interface and the interface is
button pot. So I guess we will go high tech sort of speek. Thanks for the
help guys.