off-power Answer Mach: responses
jo...@jach.hawaii.edu
along with the original posting. The long-distance ground loop is
something I had not experienced before, nor considered...
Anyway, I think I've now got a handle on what to try next (when I get
off this hill), and I thank all of you for the rapid replies.
Aloha!
====================================================================
Newsgroups: sci.electronics
From: jo...@jach.hawaii.edu
Subject: off-power Answering Machine
Organization: UK/Canada/Netherlands Joint Astronomy Centre, Hilo, HI
Date: Sat, 4 Jan 1992 08:32:33 GMT
I've got a cute puzzle for some wiz out there... (whiz out there?)
I live off-power, my home system runs on solarvoltaics, batteries, and
an inverter. My power capacity is limited, and I spend about two weeks
every month away from home (on the summit of a 14000' mountain). The
phantom loads and minimum inverter power consumption are such that, if I
leave the inverter on while gone, the batteries are slowly but
inexorably being depleted... while at home, I usually need to use a
generator to boost my batteries every coupla daze.
So- to the problem- I decided to turn off the inverter whenever I am
away from home, but want to keep my answering machine powered. Simple,
since the answer mach is powered by a "power cube" labelled "120vac in,
12v 500ma out". I ran a 12v line from my batteries to the machine.
But- when I connect the 12v to the machine, I get a powerful and
annoying hum on the phone lines (it can be heard from any phone in the
house, and by anyone calling the machine). I checked polarity &c, and
for voltage drop due to line resistence, all OK. I'm getting a good 12v
to the machine. Thought I had a ground loop or something- but the hum
persists when I disconnect ALL other telephones, modems, etc that share
the phone line. Tried disconnecting the ground lead at the phone juntion
box outside the house- no effect. Checked the 12v line with an o'scope-
it looks perfectly clean, duzzn't seem to be picking anything up (tho
how could it, with no AC or LoRF within a half-mile?). Checked the
"power cube"- it puts out about 18v DC with the answer mach load
(notwithstanding its 12v rating). Figured this was just poor regulation,
that there is regulation in the answer mach. Looked in the machine-
there appears to be lotsa CMOS/TTL plus all the cassette drive stuff, no
regulators visible except for some diodes. Tracing is a nightmare, as
there are sundry voltages everywhere, and you can't take the board out
without unplugging some in-line connectors- I don't really want to power
it up with stuff unplugged.
One more thing- the phone part of the answering machine works fine with
NO power at all to the machine, ie powered only by the phone lines- it
is only when you plug the 12v batteries to the machine that the hum is
evident. This seems to belie the possibility that it needs more than 12v
to operate properly. As far as I can tell, I've tested it with
absolutely no connection (ie loops &c) between the batteries and the
machine, except for the power input jack...
This is driving me even nutsier- any suggestions? something else to
check?
joel.
===========================================================================
From: MX%"br...@UCSD.EDU" 4-JAN-1992 06:06:03.95
From: br...@UCSD.EDU (Brian Kantor)
Organization: The Avant-Garde of the Now, Ltd.
I think you're getting a ground loop, with the hum induced by all those
other people on your island who are on the grid leaking current back
through earth returns. Of course your batteries and the phone company
battery are separately grounded, many km apart.
When you're using the power cube, you're isolated from ground. You will
have to solve the problem by finding
- an answering machine that doesn't connect one side of the
phone line to its dc ground internally
- a phone line isolator that uses a transformer to pass audio
and relays or opto-isolators to do ringing and hookswitch
- ditto fiber-optics
- some sort of DC-to-DC converter to isolate the ground
Or maybe you could just float the answering machine on a separate
battery just large enough to handle taking a message, and use a relay
to disconnect the small battery from your large battery whenever the
answering machine has the phone off-hook.
- Brian
=========================================================================
>Brian-
>
>hmmm... I hadn't considered a ground loop that large (small-scale
>thinking, I guess), but that does make some sense... actually, the
>next thing I was going to try was a separate battery. It's not clear
>to me that my batteries are "currently" (ha ha) grounded in any way,
>although they are connected to the inverter, which I have grounded. I
>would have thought that the inverter case ground was isolated from the
>battery -V, but I could be wrong. 'Course I've probably got minor leakage
>from the battery cases, as well.
>
>DC-DC converter is another possibility I was considering- just
>throwing a regulator in, but the cheap regulators would require a
>2 volt drop, and I'm not sure the machine would like that. I may
>give that a try anyway.
>
>So thanks- you're the "first-responder" to my dilemma, and with
>intelligent suggestions too! I like it.
>
> Mahalo, aloha, & the rest- joel.
========================================================================
From: MX%"sor...@triton.unm.edu" 4-JAN-1992 11:03:33.46
Hi, Joel,
I have a similar problem, to the extent that I must power
my telephone answering machine from a special isolation
transformer, from the local 120v mains. I even tried a
small audio transformer on the phone line going to the
machine. Although it eliminated the hum, the machine
wouldn't work correctly when dc isolated from the phone
line.
Grounding, even through a fairly high resistance up to
1k or so, either side of my phone line makes mucho hum.
That's with nothing else connected to it but an AT&T
handset.
That's why I wonder if maybe your batteries are sitting
on the ground, since there would be fair capacitance to
earth, even if there were absolutely no leaks or cracks
or other paths to ground.
If the 12 volt batteries are connected to earth ground, this
could produce the hum you are having. Even grounding one
side of the battery supply might not solve anything.
The test to determine this would be to power the answer
machine with a truly isolated 12 volt source, e.g. a
car battery sitting on a plastic sheet on a wooden table.
Or a string of flashlight batts, also isolated.
Just some thoughts. Good luck.
Stan i.e. Stan Orrell sor...@triton.unm.edu
Support research on artificial intelligence, it's our only hope.
===========================================================================
From: MX%"wh...@milton.u.washington.edu" 4-JAN-1992 11:53:15.51
From: John Whitmore <wh...@milton.u.washington.edu>
Organization: University of Washington, Seattle
This sounds like typical ground loop problems; the phone ground
and your local power ground are connected together. Solutions: put
a surge-supressing transformer core around the 12VDC input (so that
the power current and return current are both coupled to the same
hunk of iron); Radio Shack #273-104 or #273-105 are easily available
components for doing this. Second solution: disconnect the
GND line (i.e. run only the +12V line from your battery); this
will remove the ground loop but will run the extra current
through the phone wiring (so it isn't optimal). Third solution:
put a resistor in series with the power GND wire (10 to 100 Ohms).
Fourth solution: use a DC-DC converter to isolate the
input power from the output (expensive, but very effective).
The DC-DC converter can be faked up with a '555 oscillator
and a 1:1 transformer; it is basically a small, very efficient
inverter which is good for only low power.
The loop is probably through the telco ground connections
(and even if you can lift one of those, there's another one
at the next house, and on every Nth telephone pole, and...).
John Whitmore
======================================================================
From: MX%"gops...@helix.bms.com" 4-JAN-1992 12:08:17.69
Organization: Bristol-Myers Squibb PRI
Well, why not take a small 6v battery (or four 1.5v batteries) in series
with the 12v battery to make 18v. Then test your low voltage theory...
Rich
--
Rich Gopstein
gops...@bms.com
rutgers!bms.com!gopstein
==========================================================================