Maker Pro
Maker Pro

Simple DIY power transient monitor?

D

DaveC

I've Googled but don't see any circuits for a simple mains power transient
monitor.

Nothing fancy. I'd be happy to have an array of LEDs, calibrated to certain
levels of dropout, spike, etc.

Does such a DIY circuit exist?

Alternately, for devices powered by a DC "brick" supply (ie, network routers,
etc.), would monitoring of the DC output be easier/better than monitoring the
mains AC?

Ideas?

Thanks,
--
Please, no "Go Google this" replies. I wouldn't
ask a question here if I hadn't done that already.

DaveC
[email protected]
This is an invalid return address
Please reply in the news group
 
A

Al

DaveC said:
I've Googled but don't see any circuits for a simple mains power transient
monitor.

Nothing fancy. I'd be happy to have an array of LEDs, calibrated to certain
levels of dropout, spike, etc.

Does such a DIY circuit exist?

Alternately, for devices powered by a DC "brick" supply (ie, network routers,
etc.), would monitoring of the DC output be easier/better than monitoring the
mains AC?

Ideas?

Thanks,

Try this. Get a digital multimeter with an RS-232 output. Write a BASIC
program to monitor the multimeter output. Have it record all deviations
from a norm you either establish or from a running average computed by
your software. The response of your multimeter will determine the
bandwidth you can monitor.

Al
 
D

Don Taylor

DaveC said:
I've Googled but don't see any circuits for a simple mains power transient
monitor.
Nothing fancy. I'd be happy to have an array of LEDs, calibrated to certain
levels of dropout, spike, etc.
Does such a DIY circuit exist?

I remember seeing a VERY simple circuit in something like Radio
Electronics way back in the early 1970's. Once or twice I went
back and searched to try to find that again, without success. It
seems like it was only a few components but would latch to record
spikes or power failures. I can't remember which, too many years,
too many dead brain cells.

However, having talked with a few of the guys who spent a bunch of
their time on the subject, they said that the more accurate you
want to be about this the harder it gets, and it ramps up quickly.
They were looking at lots of the cheap "surge protectors" and were
not impressed. I kept mumbling I should find get a big heavy Sola
transformer. Any single cycle, or less, spikes or dropouts should
meet some stiff opposition with one of those, even if they are hot
and weigh fifty kilos.
 
I

Ian Stirling

In sci.electronics.design DaveC said:
I've Googled but don't see any circuits for a simple mains power transient
monitor.

Nothing fancy. I'd be happy to have an array of LEDs, calibrated to certain
levels of dropout, spike, etc.

Does such a DIY circuit exist?

Sound card, resistors, job done.
 
W

w_tom

Long before asking how to measure it, first ask WHAT is to
be measured. Transverse mode transient or longitudinal?
Impulse? How fast? Milliseconds single digit voltage
increases or a microsecond pulse of thousands of volts? A
transient of no current - blackout or brownout? A voltage
source transient or a current source transient?
 
M

Mac

I've Googled but don't see any circuits for a simple mains power transient
monitor.

Nothing fancy. I'd be happy to have an array of LEDs, calibrated to certain
levels of dropout, spike, etc.

Does such a DIY circuit exist?

Alternately, for devices powered by a DC "brick" supply (ie, network routers,
etc.), would monitoring of the DC output be easier/better than monitoring the
mains AC?

What you are referring to as a "brick" is more correctly called a
"wall-wart." AIUI, "bricks" are line-powered, switching DC power supplies.
They are not transformer-based, and the outputs are usually fully
regulated. Wall warts usually output uh, got to be careful here, uh let me
just say that what they put out is not fully-regulated DC. ;-)

I could be wrong.
Ideas?

Thanks,

--Mac
 
S

Sam Goldwasser

Al said:
Try this. Get a digital multimeter with an RS-232 output. Write a BASIC
program to monitor the multimeter output. Have it record all deviations
from a norm you either establish or from a running average computed by
your software. The response of your multimeter will determine the
bandwidth you can monitor.

Transients may be of short durationg. Typical cheap RS232 compatible
multimeters have a 1 or 2 second integration time. Might miss it.

A cheap data acquisition board would be better.

--- sam | Sci.Electronics.Repair FAQ Mirror: http://repairfaq.ece.drexel.edu/
Repair | Main Table of Contents: http://repairfaq.ece.drexel.edu/REPAIR/
+Lasers | Sam's Laser FAQ: http://repairfaq.ece.drexel.edu/sam/lasersam.htm
| Mirror Sites: http://repairfaq.ece.drexel.edu/REPAIR/F_mirror.html

Note: These links are hopefully temporary until we can sort out the excessive
traffic on Repairfaq.org.

Important: Anything sent to the email address in the message header above is
ignored unless my full name is included in the subject line. Or, you can
contact me via the Feedback Form in the FAQs.
 
J

John - KD5YI

Long before asking how to measure it, first ask WHAT is to
be measured. Transverse mode transient or longitudinal?
Impulse? How fast? Milliseconds single digit voltage
increases or a microsecond pulse of thousands of volts? A
transient of no current - blackout or brownout? A voltage
source transient or a current source transient?

DaveC wrote:

Damn, Dave. You sure know how to rain on a parade. ;-)

John
 
I

Ian Stirling

In said:
YOU FORGOT THE ISOLATION TRANSFORMER!!! Other than that, it could work
pretty well (if you lived).

True.
Personally, I'm happy with only the resistors.
Three in series in each of the legs, two different makes, each one
of a rating able to take 2* the input voltage, and rated for the input
voltage, and three in parallel for the shunt resistors). As well as output
isolation resistors after the shunt.
So a dangerous fault would be at least 4 failures away.
 
D

Dimitrios Tzortzakakis

Your utility does this all the time.Why reinvent the wheel?
 
A

Al

Sam Goldwasser said:
Transients may be of short durationg. Typical cheap RS232 compatible
multimeters have a 1 or 2 second integration time. Might miss it.

A cheap data acquisition board would be better.

--- sam | Sci.Electronics.Repair FAQ Mirror: http://repairfaq.ece.drexel.edu/
Repair | Main Table of Contents: http://repairfaq.ece.drexel.edu/REPAIR/
+Lasers | Sam's Laser FAQ: http://repairfaq.ece.drexel.edu/sam/lasersam.htm
| Mirror Sites: http://repairfaq.ece.drexel.edu/REPAIR/F_mirror.html

Note: These links are hopefully temporary until we can sort out the excessive
traffic on Repairfaq.org.

Important: Anything sent to the email address in the message header above is
ignored unless my full name is included in the subject line. Or, you can
contact me via the Feedback Form in the FAQs.

You used to be able to get a free data acquisition kit at Dataq. It was
crippled, of course. But you might be able to write your own program for
it. It was a promo to get you to buy better modules and/or software. Try
this web site: http://www.dataq.com/

Al
 
B

Bruce Esquibel

: I remember seeing a VERY simple circuit in something like Radio
: Electronics way back in the early 1970's. Once or twice I went
: back and searched to try to find that again, without success. It
: seems like it was only a few components but would latch to record
: spikes or power failures. I can't remember which, too many years,
: too many dead brain cells.

In the same boat here with memory, so take this with a grain of salt...

I think the project was in Popular Electronics, being Elementary Electronics
was more "kids" and Radio Electronics was more "adult".

From what I remember, I'm pretty sure I built one, but it wasn't very good.
It was a simple over or under voltage monitor (one led for each and a reset
button).

I think it used a "surplus junk transformer" out of your junk drawer that
had either a 6.3v or 12.6v secondary. Whatever chip it used (could of been
like a LM317) had a simple circuit around it to monitor the secondary
voltage of the transformer (probably after passing through a 1N4004, no
filters). I think the point of the article was if you used a 12.6v
transformer, you were dealing with 10:1 ratios or something.

Anyway, whatever adjustment was inside set the high or over voltage led to
come on and stay on. Think that was set to (in theory) to 130v.

It didn't really detect low voltage if I remember correctly, was it did was
if voltage was lost completely, when restored the other led would light.
Same bit when you first plugged it in, the low led automatically lit, you
pressed reset to clear.

So I guess if you came home and the low led was on, you knew you lost power.

I remember this thing because it was around the same time Radio Shack
started selling that chip, which for the life of me I can't remember now, it
was like designed for building an audio sound VU meter. There was a follow
up at some point on building "an improved version" of the monitoring gizmo
using it. So instead of a simple two led thing, you could read the voltage
directly using 10 or so led's.

Remember this was all prior to cheap LCD meters being around (probably pixie
tubes were the closest digital meters) so it was sort of neat having a
visual multi-led ac meter, even if it only worked from 108-130 or whatever
the range was.

Point is, even the fancy version didn't detect spikes.

Pretty sure Sam G. had articles in all those mags at that time, if there is
an index to be had, he should know.

-bruce
[email protected]
 
N

NSM

You used to be able to get a free data acquisition kit at Dataq. It was
crippled, of course. But you might be able to write your own program for
it. It was a promo to get you to buy better modules and/or software. Try
this web site: http://www.dataq.com/

Offer now is "Chart Recorder Starter Kit for $24.95"

N
 
D

DaveC

Thus spake Ian Stirling:
Sound card, resistors, job done.

I'm sure there's a *bit* more to it than that. Care to elaborate?
--
Please, no "Go Google this" replies. I wouldn't
ask a question here if I hadn't done that already.

DaveC
[email protected]
This is an invalid return address
Please reply in the news group
 
| In sci.electronics.design [email protected] wrote:
|> Ian Stirling wrote:
|>> > I've Googled but don't see any circuits for a simple mains power transient
|>> > monitor.
|>> >
|>> > Nothing fancy. I'd be happy to have an array of LEDs, calibrated to certain
|>> > levels of dropout, spike, etc.
|>> >
|>> > Does such a DIY circuit exist?
|>>
|>> Sound card, resistors, job done.
|>
|> YOU FORGOT THE ISOLATION TRANSFORMER!!! Other than that, it could work
|> pretty well (if you lived).
|
| True.
| Personally, I'm happy with only the resistors.
| Three in series in each of the legs, two different makes, each one
| of a rating able to take 2* the input voltage, and rated for the input
| voltage, and three in parallel for the shunt resistors). As well as output
| isolation resistors after the shunt.
| So a dangerous fault would be at least 4 failures away.

Add a fuse. Put it on a branch circuit.
 
| Thus spake Ian Stirling:
|
|> Sound card, resistors, job done.
|
| I'm sure there's a *bit* more to it than that. Care to elaborate?

I've been putting some thought into a power monitor. But I want to not
only monitor voltage, but also current. And I do want some reasonable
resolution out of it but I am not sure how much I'll really need. What
I have in mind right now is to run the power from the main breaker to an
enclosure having 2 sets of terminal blocks. The first set would just be
a connection point. The 2nd set would distribute power to the branch
panels. In between there would be three current transformers. Tapped
into the terminal block will be fused voltage sensors. All this will
drive a circuit board (to be designed) that will convert levels to 16-bit
digital at some data rate like 96 kHz, serialize all channels, and drive
an LED viewed through a small covered window. A pickup with then feed
all that data to a computer to de-serialize and process as desired. This
should be sure to keep high voltages away from the computer (especially
the CT voltages). The only metal in and out of that box would be the
power conductors themselves, and ground (also metered).
 
I

Ian Stirling

In said:
| Thus spake Ian Stirling:
|
|> Sound card, resistors, job done.
|
| I'm sure there's a *bit* more to it than that. Care to elaborate?

I've been putting some thought into a power monitor. But I want to not
only monitor voltage, but also current. And I do want some reasonable
resolution out of it but I am not sure how much I'll really need. What
I have in mind right now is to run the power from the main breaker to an

Current transformers are of course good.
However, I needed something quickly to monitor power usage.
I ended up with a simple box, with a .1R resistor in the live, and
three similar resistor networks from live/neutral/sense (3 of
3 500V rated resistors in series, to 3 resistors in parallel to form the
shunt, and another 3 resistors on the output to isolate the output. (using
neutral as 'ground', and bringing that out through another 3 resistors
in parallel.)

I actually used a scope - however, there is in principle no reason why
this couldn't be done with a sound card.
The basic idea is to keep impedances high, and all currents low, so
that even in the event of multiple failures, nothing bad happens.
This does of course greatly reduce the speed, but in many cases that's
not a problem - I would have been quite happy with a bandwidth of 1Khz,
which was comfortably exceeded.

This is not an approach to take lightly, you need to understand the reasons
why this sort of circuit is not allowed in production, and carefully
engineer in the safety, rather than taking the 'easy' approach with
current/voltage transformers, when it can be more or less obvious from
construction that nothing very bad can happen.
Otherwise, you can win a Darwin award.
 
D

DaveC

Thus spake Ian Stirling:
Otherwise, you can win a Darwin award.

Posthumously (by definition).
--
Please, no "Go Google this" replies. I wouldn't
ask a question here if I hadn't done that already.

DaveC
[email protected]
This is an invalid return address
Please reply in the news group
 
Top