AC Voltage Conditioner for Appliances on Generator Power?

Hello -

I have company that makes small, low flow, high-pressure pumping
systems.

The systems consist of small digital controller (MSP430 based) with an
LCD display. The pump motor plugs into the controller and power is
supplied to the pump motor via a relay inside the controller, i.e.
there is only one power cord for the unit. The units are designed for
120V 60 Hz power. When the pump motor is running, the steady-state
current draw is 7A, when the motor is not running, the controller
draws 300 mA.

Recently a number of our systems have been installed on small
Caribbean islands that have powered supplied by diesel generators. We
have had an issue with the controllers failing - either via the fuse
blowing, or the micro getting fried. Since we have near perfect
reliability with the systems on domestic utility power, my guess is
that either power spikes or electrical noise from the generator power
is causing the failures.

The purchaser of the systems in the Caribbean has told us "everything
else here works fine so yours should too", while our contract
electrical engineers are giving us the "we can't be responsible for
anyting running on generator power". I haven't had much luck Googling
around on the topic.

So, my question is: what options would I have as far as some sort of
reliable voltage conditioning device that could strip out any electric
noise and/or provide surge suppresion when the units are running on
the generators? I need something that could be incorporated into the
unit for < $75.

Thanks

 

You probably need a better voltage regulation circuit. Can you post
your design?

 

Try this....

http://img57.imageshack.us/img57/4306/controllerioschematiczq2.jpg

....I believe this schematic constitutes the voltage regulations/
transformer circuit (apologies but I'm a ME not a EE!). The rest of
the board just contains some additional relays for other devices that
the controller triggers. The LCD and micro are located on a separate
boad that is piggy-backed onto the power board.

 

It can't be driving the micro with 11.6V. What's between the 11.6V
and micro?

 

This:

http://img74.imageshack.us/img74/4264/regsq7.jpg

and here's where it ties to the micro:

http://img74.imageshack.us/img74/1469/watchdoggr4.jpg

 

I would double check on your LM317 design. It doesn't look right.

 

The "we cannot be responsible etc" is _too_ true. The output of starting
and stopping generator is a mess of voltages and frequencies till it
stabilases. So any device connected during this time can fail.
One of simplest things would be a time delayed relay on line, activated
by line voltage and connecting the pumps and controllers after ~1 minute.
When connected beware of stopping generators.

HTH

Stanislaw

 

That's very true. Nevertheless, a properly designed regulator circuit
should be able to prevent damages. The OP's circuits are wrong and
very inefficient. It doesn't make sense to have 12V transformer/
switcher output and wasting 75% of it in the linear regulator. 300mA
for an LCD controller is ridiculous if not illegal. I am not sure if
we are still on track to ban such linear regulator by law.

 

You have big problems doing this cheaply because you tap off the
internal line feed to drive the pump. Might as well figure 1KVA for
whatever you do.

 

Very hard to say without seeing the whole unit. It is best to get a
local consultant to take a look at how it is wired, where the cables are
running and so on.

This will be more a shot in the dark but has helped a few times: Get a
linear regulated power supply of a good brand, such as Condor. Proper
amperage with lots of reserves, it can get hot in the Caribbean. Should
be one that has a uA723 chip for regulation (most do). Adjust that down
from 12V to 11.6V, place a good inline power filter up front. The Corcom
brand is good. Fuse it all properly.

A Corcom filter alone right in front of T1 might also work. Use a good
filter, not just a common mode choke. Right now it seems you have
nothing like that in front of T1, that lone transzorb isn't going to cut
it. The motor power should be tapped off before that filter.

Keep in mind that spikes can also come in via control lines into the
controller board and then you'd really need a consultant that knows EMI
stuff well.

With generators you need to be prepared for lots of brown-out
conditions. Slow and sputtering starts, running out of diesel, big loads
being switched where voltage and frequency can go all over the place etc.

 

Feed the input voltage to the MC34063A through a diode installed
before the R8/C6/pin 6 junction, and add a 100 uf from that
junction to ground. On the output side, change C15 to 270.

You can also add the optional output filter to the MC34063 -
see the datasheet. What is the purpose of R1 and R6?

Ed

 

Change the power supply to an offline SMPS, they do not care what
frequencies they receive, and if done correctly can have a wide input
voltage. Low cost off the shelf units can easily handle from 90V DC to over
250V AC with out a glitch. Sounds like your product is low volume, so I
would not try to roll your own due to regulatory testing. The good news is
some are available with several voltage outputs, so you may be able to
eliminate your other power supplies entirely, potentially saving money!
Efficiency would improve also.

 

There are also huge "magnetic stabilizer transformers" that were popular
in the former Eastern block countries because the power grid was often
so lousy that TV sets would repeatedly use sync or something would fry
in there. Those monsters must still be around somewhere.

 

Fried an oscilloscope once with one. The thing didn't get along with its
load (a few old non-PFCed SMPSes, only 50% of rated stabilizer power).
Stabilizer overheated, produced weird 100Hz (!) output harmonics, and
fortunately it was the cheap TDS2xxx scope that went first (got it repaired
on warranty).

I didn't really bother to investigate the cause of the weird behavior, I
just chucked the whole damn stabilizer into the bottom of some cabine where
I'm sure it still is today. It's a 1.5kVA unit made by Block. Maybe it has
some future use as a gravity lens.

robert

 

Some scopes in the TDS series are truly deplorable IMHO. That's why I
recently bought an Instek instead.

Watch out for that nasty conducted EMI at around 40/80kHz. It can really
fool you when looking at analog signals. Probably leakage from the
backlight inverter or maybe its switcher.

Well, you aren't supposed to connect switchers and PFC stuff to it. They
weren't meant for that. In this thread it's a simple air conditioner
where 99% of the load is motors.

 

I didn't have time to dive into all this.
Let me just say that if the generators are configured to provide
backup power, (automatic transfer), then you'll sometimes see systems
that do not bother to synchronize the Utility-provided AC waveform
with that provided by the backup generator. They just throw the
switch when normal utility returns.

So, when the automatic transfer switch goes back to Utility power,
it's possible to switch the load from the max voltage of one souce, to
the min voltage of the other. There are also usually slight frequency
variations between the two sources, but this is likely unrelated to
your experience.

If any of this is the case, your customers really need an in-phase
monitor on their automatic transfer switches. (If manual, it probably
takes too long to throw the switch?, and this problem would not
occur.)

I agree with other posters here that a robust power supply on your end
could be designed to handle utility transfer switches for generators
if you feel there is a reasonable design justification for same.
Good luck.

-mpm