From: John De Armond
Subject: Green Plug 1, Refrigerator 0
Date: Sat, 12 Mar 94 07:46:34 GMT
Keywords: smoke, flames
Back in November I posted some preliminary results of my testing under
actual conditions, a Green Plug with a 1977 yearmodel Hotpoint Model
CTF18C refrigerator and some squirrel-cage fans. I promised that I
would post the long term tests on the refrigerator when they were
completed. The long term energy consumption tests were complete in
February but I decided to wait until now to post so as to let a process
I saw developing run to completion. The process was the destruction of
the compressor in this refrigerator. The refrigerator fired back by
destroying the green plug.
The process started back in January when I started seeing on the
attached recording ammeter an increased duration of starting current.
This increase appeared to be caused by the "soft start" feature (sic) of
the green plug that is supposed to limit inrush current. At the time I
got the green plug, I thought it foolhardy, because limiting starting
current drags out the starting process which actually subjects the motor
winding to MORE heat. Shortly after I noted this increased starting
duration, the refrigerator tripped the GFI on the circuit it was plugged
into. Prime evidence that the winding insulation in the compressor was
I decided to let this play itself out so I plugged the refrig and green
plug into a non-GFI outlet. The refrigerator continued maintaining the
temperature setpoint as indicated by the digital thermocouple meter
attached to the refrig. I noticed, however, that the compressor
overload would occasionally trip on excess temperature. And the total
current draw continued to creep up as the leakage current continued to
This afternoon the process ended. I heard the thermostat on the refrig
turn on and a moment later, the green plug erupted in a pleasing ball of
smoke and flame. I checked resistance between the refrigerator cord
prongs and ground. Dead short. I cracked the service fitting on the
compressor and was greeted by what is probably the worst compressor
burnout I've ever seen. This was a long term roast and not a quick
After I replace the compressor, I plan to open the shell of the old one
to inspect the damage and make sure it wasn't a tight bearing or
something else mechanical. But based on the electrical history I have,
I'm pretty positive that the green plug took out this refrigerator. I
plan on sending this device along with a letter asking them to pay for
the compressor to the factory to see what kind of reaction I get.
So the final score is this. The green plug was "saving" electricity
at the rate of about a dollar a month. Meanwhile, the replacement
compressor will cost me about $80 plus the better part of a day
to replace it and flush all the contamination out of the freon system.
I think that extra buck a month for power without the Plug was a
From: John De Armond
Subject: Re: Green Plug 1, Refrigerator 0 and more info
Date: Fri, 15 Apr 94 03:58:50 GMT
firstname.lastname@example.org (Willie Smith) writes:
>Anecdotal data _is_ nice to hear, and I actually read John's post with
>interest. However, John's conclusion, that Green Plugs ruin fridges,
>is, given the data, pretty far-fetched. I was just trying to warn the
>readers of this thread that a single inconclusive datapoint should not
>be taken as a scientific experiment damning the Green Plug. And, yes,
>I do failure isolation and analysis all day long using scientific
>procedures and carefully designed experiments (and I've been wrong in
>my 'conclusions' a number of times),
If you don't conclude any better than you quote (or even paraphrase),
I can understand your mistakes. To correct the record, what
I said is this. My green plug killed my refrigerator. I *KNOW*
this much. From the operational characteristics I observed and my
couple of decades of refrigeration experience, it is likely this thing
will kill other refrigerators. That is what I said THEN. NOW I have
>and I know for a fact that John's
>conclusion is at best a guess.
Funny thing is, my "guesses" are almost always right.
I've had a spirited exchange of messages via a third party with
Chris Riggio, chief engineer at Green Technologies, Inc. Following
is the relevant part of our communications:
To: John De Armond, Performance Engineering Magazine
From: Chris Riggio
Subject: GreenPlug Experience
In response to John De Armond's GreenPlug experience. The destruction
process you described in your March 15, 1994 text is indicative of a
failure mode called half cycling. While this misfortune is rare it can
and does occur. A more rigorous analysis (for those so inclined) would
have included voltage and current waveforms on several GreenPlugs as a
means of detecting the cause of failure. A properly functioning GreenPlug
has a soft start ramp that elapses in five line cycle or 82 milliseconds
during a motor start. The advantages of this design philosophy are self
evident to those skilled in the motor design art and are well understood
in industry. When voltage is applied symmetrically to a motor load in a
ramped profile, shaft rotation begins before peak current is reached.
This allows the overall start current to be reduced as well as reducing
winding and laminate stresses. The exception to this, as evidenced by
your ammeter, and eventually the compressor, is the half wave rectified
output induced by a faulty plug. This condition will certainly increase
motor current and suspend or prevent motor starting. Our failure rate for
GreenPlugs as a percentage of total production to date is approximately
0.2%, not bad for Power Control Electronics.
We certainly understand and accept the responsibility that comes with
interposing a voltage controller between the line and it's load. I think
you will find our customer service policies to be commensurate. Usually,
people recognize a malfunction of the type you described by virtue of the
low frequency hum, flickering cabinet light, or tripping of the thermal
breaker which are common to half cycling malfunction. In your case, return
for replacement would have yielded additional insight as to the real
malfunction and maybe saved you and us a compessor repair, barring, of
course, a bad compressor going in.
I never tell everything I know the first time around and this is no exception.
Riggio's enthusiastic claim that the green plug failed in the half cycle
mode (where only half the 60 cycle waveform is passed to the
refrigerator, resulting in a high DC component and heating.) tells me
that this is a common failure mode. It also tells me that my conjecture
that this "soft start" mode is really to protect an under-rated power
semi-conductor was right on the money. What I didn't say in my first post
is that I had a digital storage scope connected to the output of the
green plug at all times and at no time was half cycling apparent.
I have the hard copy screen dumps to prove it too.
Another thing I didn't mention (mainly because it is mostly irrelevant
to this dicussion) is that I ran my observations past a friend of mine
who does small motor failure analysis under contract for several
insurance companies and he agreed with my conclusions.
More important, since half-cycling is a predictable failure mode and
since this thing has to monitor the current and voltage waveforms in
order to function, it is reckless for the green plug to not detect this
failure mode and shut down. Food would be ruined but at least the
refrigerator would not be damaged. It would really be nice too, if that
nice little green light on the front was designed to go out when the
thing failed. Mine still shines brightly. Connecting it to the output
terminals would do the trick.
A couple of other important things are evident from his message.
The most glaring thing is the failure rate, 0.2% - 2 failures for
every thousand units. By any measure, this is a hideously high
failure rate, particularly for a device which controls the fate of
appliances which may cost thousands of dollars. Finally, his last
paragraph tells me that they're fixing a lot of refrigerators.
It should be fair warning to those who continue using their green
As I've said before, I am greatly disappointed by this product. It is
a genuinly good idea but with a terrible implementation. Depending
on the local power rate, a properly designed green plug would save
enough energy to pay for itself in a couple of years and from there
on out, it would be free money, allbeit a small amount. I wouldn't even
have much of a fault with the unit if it had a high failure rate but
detected said failures and either shut down or even better, went
into pass-thru mode accompanied by an alarm (such as using a bi-color
LED and having it change from green to red or even simply to go out.),
assuming they really do have a no-questions-asked warranty. (and I have
no reason to believe they don't.) This would add only a trivial amount of
cost to the product. As it is, this thing has a high rate failure mode
and the mode identified by Riggio WILL toast whatever is plugged
into the device. This is unacceptable.
The only criticism from Riggio that is a bit legitimate is that I should
test several plugs and refrigerators. Were I writing a refereed paper
for the IEEE Spectrum or ASHRAE, I would. As it is, I have absolutely
no intention of spending hundreds or thousands of dollars on multiple
unit testing just so I can write a free Usenet article. If anyone
doesn't like that, well..... Tough shit, pardon my french. If someone
wants to fund such a test program, I'm available.
The bottom line is, if you think putting your up-to-several thousand
dollar refrigerator at risk to save a buck a month or less in power is
a good idea, go right ahead and get one of these gadgets. If not,
you can probably save as much energy by turning off the butter warmer.
So Willie, seems my "guess" wasn't so far off, after all, huh?
That tends to happen after a few years of age and experience.
Oh, and BTW. You're a rank amateur in the rebuttal game.
Work on it a few years. You'll get there.