If you own a dual channel scope, and have an X10 system, you owe it to yourself to hook your scope to the powerline and watch those X10 signals fly by. Unfortunately, it'll take more than flipping the switch from X1 to X10 on the scope probe...
Apart from seeing X10 at work, a scope can literally display problems with your setup that are otherwise hard to find, such as noise or weak or degraded signals. The following description is the schematic for the ACT ScopeTest2 device, which Phil kindly posted. You can buy it for $40+S&H from ACT (Dec 2003: no longer available).
Alternatively, you can build it for about $4 worth of components as described here.
Little bit of theory
First, you must remember that a standard oscilloscope is
incapable of showing X-10 signals by itself. That's because the X10 signal
amplitude is just a few volts (sometimes just a few tenths of a volt), compared with the
110v or 220v that's powering the house grid. Showing a 220v line is also more than what
most scopes can handle, amplitude wise. Thus the X10 scope adapter does two things:
- It divides the mains voltage by a factor of 20 so that a 110v
line would be fed to one channel of the scope (say ch1) at a reduced amplitude of about
5v. This is done via a simple resistor voltage divider.
- It filters the X10 signal from the powerline and 'flatens' it a bit and feeds it to the other channel of the scope. This is done with a simple capacitor acting as a high pass filter (the X10 signal is 120Khz, compared with the 60 or 50Hz line frequency). The original ACT ScopeTest2 uses a fancier filter, that you can make yourself as well. The resistor loads the capacitor and flatens the curve to better show noise. The diference between the fancy filter and the single capacitor filter is that the cap filter will tend to pass to the scope any noise, even noise that doesn't affect x10 signal frequency much. The fancy filter has a bandpass filter around the x10 frequency and rejects noise not too relevant to X10. For practical purposes, though, it doesn't make much of a difference, and I'm happy with the simple cap.
Displaying both channels (on a dual or better trace scope) shows vividly the x10 signal, its amplitude, its timing in relation to the powerline zero crossing, any noise around its frequency (80-150Khz) and, of course, in real time.
Making the adapter
Refer to the figure above, you'll need:
- Two short BNC cables with BNC terminals - I simply cut in half a cable with two BNC connectors at its ends.
- Three resistors:
R1 - 2Mohm/0.25W
R2 - 100Kohm/0.25W. If you have a 220v system, you might want R2 to be 50Kohm/0.25W instead.
R3 - 100Ohm/0.25W - Capacitor:
C1 - 0.1mF/400v - A two-prong polarized plug. Get one with some space inside and you'll be able to fit the whole circuit inside the plug.
The diagram shows the line being split into two paths. One path goes through a
voltage divider (about 20:1) so that it can be used on an o-scope that has its highest
volts/div as 5v. That goes into input ch1 on the 'scope (since that is usually the
triggered input). I was able to fit the whole circuit inside my 220v standard plug. Word of caution: absolutely, positively make sure that the
outer ground in the BNCs are connected to neutral and not to line. Otherwise,
you'll put your scope case at high voltage and yourself and/or others in risk of electric
shock from touching it!
After you complete the adapter construction, hook it up to power, but not to the scope yet and do so without touching the BNC cables. Wrap them in electrician tape if necessary - but leave a spot so you can touch them with a tester probe. Take an electrician tester and touch the BNC outer ground with the tester. If you hooked it up correctly - the tester stays unlit. If you hooked it up the wrong way - the tester lights, indicating that the BNC ground is at high voltage - a NO NO! Reverse connections and retest. Do not use your adapter until you're absolutely sure you got this right. This is one test you'd really want to pass.
Using the adapter
Now you can adjust your sweep speed to 2ms/div (or if you
want only one sine wave, set it to 1ms/div and then adjust the fine sweep knob).
This circuit, when used with a dual trace oscilloscope, will display the X-10 signal, its
relationship with the 60Hz sine wave plus any noise within the range that most effects
X-10 systems (80kHz-150kHz). Use the following setting (assuming you've connected channels
as in the figure above:
- set channels to gnd coupling and position them overlapping on the 0v graticule. Switch both channels to dc coupling and set:
- ch1: 2V/div
- ch2: 10mV/div
- vertical mode: ADD
- Horizontal Display: A
- Time/Div: 2mS
- Trigger Mode: Auto
- Trigger Coupling: AC
- Trigger Source: Auto
Finally, make sure that your scope is not grounded through the power cord. Use one
of those little plug adapters that we used when old houses didn't have grounded
outlets. That way, if there is any voltage on the neutral, you will not get current
flow through the scope case. Another consideration is that all X-10 transmitters and
receivers use neutral as the reference, not ground, so when you use your scope, it too,
will use neutral as the reference. All the scope shots used in the tech article were
using the ScopeTest2.