Started by June 30, 2015
```Now sure how to describe the problem in few words, sorry.

I have two circuit boards (to be designed), A and B. A has the power supply=
. A and B will be connected by a longish, 2 conductor wire. The wire will b=
e used to supply 12V+ and ground from A to B. B has no other source of powe=
r.

B needs to signal A when something happens.

For various reasons, I can't replace the wire with a 3 conductor version, a=
nd wireless solutions aren't practical.

Normally, board B draws maybe 60mA at most, mostly for LEDs. But occasional=
ly board B will close a relay, and feed the 12V into a larger load: a DC-DC=
converter, to generate 5V @ maybe 1-3A. Presumably that will show up as a =
larger current draw on the 12V line, but I don't know how much. (I can slap=
a high wattage resistor in parallel with the load to make it draw more cur=
rent, if that helps.)

That's what A has to detect.

I don't know how to detect current changes and I don't entirely trust my es=
timates on the current change anyway.

Is there a clever and inexpensive way to overlay some sort of signal on the=
wire that is reliably detectable? Or is there an easily adjustable way to =
detect current changes on a power line? It is probably safe to say that B d=
raws considerably less than 500mA normally and something over 500mA during =
the event.=20

TIA.
```
```In article <af1675d4-9fde-47b8-a41d-27139759fee7@googlegroups.com>,
scott.a.mayo@gmail.com says...

[snip]

> I don't know how to detect current changes and I don't entirely trust
> my estimates on the current change anyway.
>
> Is there a clever and inexpensive way to overlay some sort of signal
> on the wire that is reliably detectable? Or is there an easily
> adjustable way to detect current changes on a power line? It is
> probably safe to say that B draws considerably less than 500mA
> normally and something over 500mA during the event.

HTH
```
```On Tuesday, June 30, 2015 at 9:07:54 PM UTC-4, Randy Day wrote:

>
> HTH

Interesting, but it doesn't look trivial to implement, and since I'm trying to send one bit of information ("it happened") it feels like overkill.

One idea I'm toying with is trying to put a very brief (100ms) very low resistance across the power supply, to try and get the (unregulated) 12v supply to sag. By comparing the sag against the previous 12v voltage (stored with a cap) with a comparator, can I catch the event? But I'm not sure I like the idea of using a very low value resistor to get an unregulated 12v 4A power supply to sag.

If there some way to inject a brief oscillation, say at 20kHz, on the line that the other side can detect?
```
```<scott.a.mayo@gmail.com> wrote in message
>Now sure how to describe the problem in few words, sorry.

>I have two circuit boards (to be designed), A and B. A has the power
>supply. A and B will be connected by a >longish, 2 conductor wire. The wire
>will be used to supply 12V+ and ground from A to B. B has no other >source
>of power.

.B needs to signal A when something happens.

>For various reasons, I can't replace the wire with a 3 conductor version,
>and wireless solutions aren't practical.

>Normally, board B draws maybe 60mA at most, mostly for LEDs. But
>occasionally board B will close a relay, and feed the 12V into a larger
>load: a DC-DC converter, to generate 5V @ maybe 1-3A. Presumably that will
>show up as a larger current draw on the 12V line, but I don't know how
>much. (I can slap a high wattage resistor >in parallel with the load to
>make it draw more current, if that helps.)

>That's what A has to detect.

>I don't know how to detect current changes and I don't entirely trust my
>estimates on the current change >anyway.

>Is there a clever and inexpensive way to overlay some sort of signal on the
>wire that is reliably detectable? Or is >there an easily adjustable way to
>detect current changes on a power line? It is probably safe to say that B
>draws >considerably less than 500mA normally and something over 500mA
>during the event.

What about putting a very low resistance ,say about .1 to .5 ohms in series
with the wires and then an IC compariator circuit across the resistor.  YOu
set the comparitor to trip when say over 100 to 150 ma is drawn.

```
```> What about putting a very low resistance ,say about .1 to .5 ohms in series
> with the wires and then an IC compariator circuit across the resistor.  YOu
> set the comparitor to trip when say over 100 to 150 ma is drawn.

This sounds interesting, but at 12v and 4A, and assuming the load wants most of that when the relay closes... wouldn't I be looking at a 50W resistor?

Assuming that's feasible, what sort of comparator do I use?
```
```scott....@gmail.com wrote:

> > What about putting a very low resistance ,say about .1 to .5 ohms in series
> > with the wires and then an IC compariator circuit across the resistor.  YOu
> > set the comparitor to trip when say over 100 to 150 ma is drawn.
>
> This sounds interesting, but at 12v and 4A, and assuming the load wants most of that when the relay closes... wouldn't I be looking at a 50W resistor?

** Ohms Law a bit rusty with you ?

4A and 0.1 ohms give 0.4V drop and the power is 1.6 watts.

> Assuming that's feasible, what sort of comparator do I use?

** I would use a reed switch with a few dozen turns or so of wire wound around it - eg:

http://www.arunet.co.uk/tkboyd/ec/ec1RehaReedP0331.jpg

....  Phil

```
```<scott.a.mayo@gmail.com> wrote in message
>
>> What about putting a very low resistance ,say about .1 to .5 ohms in
>> series
>> with the wires and then an IC compariator circuit across the resistor.
>> YOu
>> set the comparitor to trip when say over 100 to 150 ma is drawn.
>
> This sounds interesting, but at 12v and 4A, and assuming the load wants
> most of that when the relay closes... wouldn't I be looking at a 50W
> resistor?
>
> Assuming that's feasible, what sort of comparator do I use?

At 4 amps and a .1 ohm resistor,a two watt resistor would be ok.  The power
of the resistor would be 4 x 4 x .1 or 1.6 watts.  A 2 watt resistor  would
allow a safety margin.  A 10 watt resistor would be enough if a .5 ohm
resistor was used.  4x4x.5 for the wattage.

You can start by looking at this IC.

```
```In article <4bc07c9a-c944-478b-a744-49fb1239814b@googlegroups.com>,
scott.a.mayo@gmail.com says...
>
> On Tuesday, June 30, 2015 at 9:07:54 PM UTC-4, Randy Day wrote:
>
> > Google 'power line communication'.
> >
> > HTH

[snip]

> If there some way to inject a brief oscillation, say at 20kHz, on the
> line that the other side can detect?

A small-value capacitor feeds signal from
the transmitter to the power trace on B. A
small-value cap on the power trace goes to

DC in ---/\/\--+---/ /--+--/\/\--dc out
L1   |        |    L2
= C1     = C2
|        |
rcv      xmt

You'd want inductors L1/L2 to prevent
the signal vanishing into your filter
caps.
```
```Phil Allison wrote:

>
> ** I would use a reed switch with a few dozen turns or so
> of wire wound around it - eg:
>
> http://www.arunet.co.uk/tkboyd/ec/ec1RehaReedP0331.jpg

** Probably I should explain that a bit more.

The coil wound around the reed switch *replaces* the previously mentioned resistor in line with the 12V DC supply.

Adjust the number of turns ( which may be wound on a plastic tube for convenience ) so the reed switch reliably closes when the current reaches the higher level.

A 20mm long reed switch operates in less than a millisecond and needs about 20 turns at 1 amp DC.

....  Phil

```
```On Tue, 30 Jun 2015 17:14:50 -0700 (PDT), scott.a.mayo@gmail.com
wrote:

>Now sure how to describe the problem in few words, sorry.
>
>I have two circuit boards (to be designed), A and B. A has the power supply. A and B will be connected by a longish, 2 conductor wire. The wire will be used to supply 12V+ and ground from A to B. B has no other source of power.
>
>B needs to signal A when something happens.
>
>For various reasons, I can't replace the wire with a 3 conductor version, and wireless solutions aren't practical.
>
>Normally, board B draws maybe 60mA at most, mostly for LEDs. But occasionally board B will close a relay, and feed the 12V into a larger load: a DC-DC converter, to generate 5V @ maybe 1-3A. Presumably that will show up as a larger current draw on the 12V line, but I don't know how much.

---
If board B is drawing 60mA quiescently from the 12V supply and then
something happens to energize the relay and turn on the DC-DC
converter, the current from the 12 volt source must increase in
order to feed the relay coil and the 12 volt side of the DC-DC
converter.

Assuming one of those jellybean 400 milliwatt coil relays, that'll

.          P     0.4W
.     I = --- = ------ = 0.033A = 33mA
.          E     12V

required from the 12 volt supply for the relay.

Then, if the output of the DC-DC converter supplies:

.     P = IE = 5V * 1A = 5 watts

into a load, and the efficiency of the DC-DC converter is 85%, the
DC-DC converter will need:

.            Pout,W       5W
.     Pin = --------  = ------ = 5.88 watts
.             0.85       0.85

from the 12 volt supply, which equates to:

.          P     5.88W
.     I = --- = ------- = 0.490A = 490mA.
.          E      12V

The sum of the three currents, then, is:

.     It = Iq + Ik + Ic = 60mA + 33mA + 490mA = 583mA.
---

>(I can slap a high wattage resistor in parallel with the load to make it draw more current, if that helps.)

---
That shouldn't be necessary.
---

>That's what A has to detect.
>
>I don't know how to detect current changes and I don't entirely trust my estimates on the current change anyway.

---
The classical way is to use a low-valued resistor (curiously called
a "shunt") in series with the supply and to measure the current
through the resistor by measuring the voltage across it and applying
Ohm's law,

.          E
.     I = ---
.          R

---

>Is there a clever and inexpensive way to overlay some sort of signal on the wire that is reliably detectable? Or is there an easily adjustable way to detect current changes on a power line? It is probably safe to say that B draws considerably less than 500mA normally and something over 500mA during the event.

---
The suggestion to use a reed relay wrapped in a coil with a wire
diameter large enough to make its resistance negligible when placed
in series with the 12 volt supply, and enough turns to meet the reed
switch's ampere-turn MAKE requirement when the supply's target
current is met, or exceeded, is a good one.

However, there are at least a couple of caveats.

One is that reed switches exhibit large hystereses, and once the
switch contacts are MADE, they may not break when the current in the
coil reverts to quiescent.

Another potentially nastier one is that trimming the winding to just
what the reed switch needs to turn ON is tricky and, if you have to
do more than a few, less than an attractive way to go about it.

My suggestion would be to use a high-side shunt, a dual opamp, and a
reference, like this:

https://www.dropbox.com/s/i6hlcrv73ara2il/Current%20threshold%20detector.asc?dl=0

you have any questions about the circuit or you need a circuit
description, post a request and I'll do what I can to help you out.

Version 4
SHEET 1 2680 692
WIRE -1312 -1392 -1360 -1392
WIRE -1248 -1392 -1296 -1392
WIRE -1184 -1392 -1232 -1392
WIRE -1072 -1392 -1120 -1392
WIRE -1008 -1392 -1040 -1392
WIRE -944 -1392 -976 -1392
WIRE -880 -1392 -912 -1392
WIRE -816 -1392 -848 -1392
WIRE -752 -1392 -784 -1392
WIRE -688 -1392 -720 -1392
WIRE -624 -1392 -656 -1392
WIRE -560 -1392 -592 -1392
WIRE -496 -1392 -528 -1392
WIRE -432 -1392 -464 -1392
WIRE -368 -1392 -400 -1392
WIRE -304 -1392 -336 -1392
WIRE -240 -1392 -272 -1392
WIRE -160 -1392 -208 -1392
WIRE -1360 -1360 -1360 -1392
WIRE -1184 -1360 -1184 -1392
WIRE -1120 -1360 -1120 -1392
WIRE -160 -1360 -160 -1392
WIRE -1056 -1344 -1264 -1344
WIRE -992 -1344 -1056 -1344
WIRE -864 -1344 -912 -1344
WIRE -816 -1344 -864 -1344
WIRE -688 -1344 -736 -1344
WIRE -1360 -1296 -1360 -1328
WIRE -1184 -1296 -1184 -1328
WIRE -1120 -1296 -1120 -1328
WIRE -1056 -1296 -1056 -1344
WIRE -608 -1296 -608 -1328
WIRE -160 -1296 -160 -1328
WIRE -1360 -1232 -1360 -1264
WIRE -1184 -1232 -1184 -1264
WIRE -1120 -1232 -1120 -1264
WIRE -320 -1232 -320 -1248
WIRE -160 -1232 -160 -1264
WIRE -768 -1216 -768 -1232
WIRE -352 -1216 -432 -1216
WIRE -864 -1200 -864 -1344
WIRE -800 -1200 -864 -1200
WIRE -240 -1200 -288 -1200
WIRE -688 -1184 -688 -1344
WIRE -688 -1184 -736 -1184
WIRE -352 -1184 -688 -1184
WIRE -1360 -1168 -1360 -1200
WIRE -1184 -1168 -1184 -1200
WIRE -1120 -1168 -1120 -1200
WIRE -1056 -1168 -1056 -1216
WIRE -992 -1168 -1056 -1168
WIRE -864 -1168 -912 -1168
WIRE -800 -1168 -864 -1168
WIRE -160 -1168 -160 -1200
WIRE -608 -1136 -608 -1216
WIRE -560 -1136 -608 -1136
WIRE -432 -1136 -432 -1216
WIRE -432 -1136 -480 -1136
WIRE -1056 -1120 -1056 -1168
WIRE -1040 -1120 -1056 -1120
WIRE -1360 -1104 -1360 -1136
WIRE -1264 -1104 -1264 -1344
WIRE -1184 -1104 -1184 -1136
WIRE -1120 -1104 -1120 -1136
WIRE -1056 -1104 -1056 -1120
WIRE -864 -1104 -864 -1168
WIRE -160 -1104 -160 -1136
WIRE -432 -1088 -432 -1136
WIRE -608 -1072 -608 -1136
WIRE -1360 -1040 -1360 -1072
WIRE -1184 -1040 -1184 -1072
WIRE -1120 -1040 -1120 -1072
WIRE -160 -1040 -160 -1072
WIRE -1360 -976 -1360 -1008
WIRE -1184 -976 -1184 -1008
WIRE -1120 -976 -1120 -1008
WIRE -160 -976 -160 -1008
WIRE -1264 -960 -1264 -1024
WIRE -1056 -960 -1056 -1024
WIRE -1056 -960 -1264 -960
WIRE -1040 -960 -1056 -960
WIRE -864 -960 -864 -1024
WIRE -864 -960 -1040 -960
WIRE -768 -960 -768 -1152
WIRE -768 -960 -864 -960
WIRE -608 -960 -608 -1008
WIRE -608 -960 -768 -960
WIRE -432 -960 -432 -1008
WIRE -432 -960 -608 -960
WIRE -320 -960 -320 -1168
WIRE -320 -960 -432 -960
WIRE -1360 -912 -1360 -944
WIRE -1184 -912 -1184 -944
WIRE -1120 -912 -1120 -944
WIRE -1040 -912 -1040 -960
WIRE -160 -912 -160 -944
WIRE -1360 -848 -1360 -880
WIRE -1312 -848 -1360 -848
WIRE -1248 -848 -1296 -848
WIRE -1184 -848 -1184 -880
WIRE -1184 -848 -1232 -848
WIRE -1120 -848 -1120 -880
WIRE -1072 -848 -1120 -848
WIRE -1008 -848 -1040 -848
WIRE -944 -848 -976 -848
WIRE -880 -848 -912 -848
WIRE -816 -848 -848 -848
WIRE -752 -848 -784 -848
WIRE -688 -848 -720 -848
WIRE -624 -848 -656 -848
WIRE -560 -848 -592 -848
WIRE -496 -848 -528 -848
WIRE -432 -848 -464 -848
WIRE -368 -848 -400 -848
WIRE -304 -848 -336 -848
WIRE -240 -848 -272 -848
WIRE -160 -848 -160 -880
WIRE -160 -848 -208 -848
FLAG -768 -1232 +12
FLAG -1040 -912 0
FLAG -1040 -1120 +12
FLAG -320 -1248 +12
FLAG -608 -1328 +12
FLAG -240 -1200 KON
SYMBOL res -896 -1184 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R3
SYMATTR Value 10K
SYMBOL res -880 -1120 R0
SYMATTR InstName R4
SYMATTR Value 100K
SYMBOL res -720 -1360 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R2
SYMATTR Value 100K
SYMBOL res -896 -1360 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R1
SYMATTR Value 10K
SYMBOL res -1072 -1312 R0
SYMATTR InstName R5
SYMATTR Value .1
SYMBOL voltage -1056 -1120 R0
WINDOW 0 14 102 Left 2
WINDOW 3 24 96 Invisible 2
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
SYMATTR InstName V1
SYMATTR Value 12
SYMBOL Opamps\\LT1366 -768 -1248 R0
SYMATTR InstName U1A
SYMBOL current -1264 -1104 R0
WINDOW 3 24 80 Invisible 2
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
SYMATTR InstName I1
SYMATTR Value PULSE(0 3 0 1 1 1U 2)
SYMBOL Opamps\\LT1366 -320 -1264 R0
SYMATTR InstName U1B
SYMBOL References\\LT1634-5 -608 -1040 R0
WINDOW 3 -96 40 Left 2
WINDOW 0 -64 -1 Left 2
SYMATTR InstName U2
SYMBOL res -464 -1152 R90
WINDOW 0 63 58 VBottom 2
WINDOW 3 67 57 VTop 2
SYMATTR InstName R7
SYMATTR Value 9100
SYMBOL res -448 -1104 R0
WINDOW 3 41 70 Left 2
SYMATTR InstName R8
SYMATTR Value 1000
SYMBOL res -624 -1312 R0
SYMATTR InstName R6
SYMATTR Value 6800
TEXT -1020 -936 Left 2 !.tran 3
TEXT -1088 -792 Left 4 ;CURRENT THRESHOLD  DETECTOR
TEXT -416 -872 Left 2 ;J FIELDS  01 JULY 2015
TEXT -1344 -872 Left 3 ;BOARD B
TEXT -1104 -872 Left 3 ;BOARD A

John Fields
```