Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Dave4720]

I'm going nuts using this XTBM VoLpmeter!  Not because of the meter, but because of what I am finding.

I use the VoLpmeter during the day to find and solve issues.  I get noise identified and beat well down.

My work extends into darkness.  Now I get .20 to .30 noise all around the house!

I finally find that the X10 XPDI3 and WS12A modules are putting out noise when at full brightness.  Now, we know that at full brightness we get max current and minimum TRIAC off time - but yes, the TRIAC is not 100.0% on even at full brightness.  Seems like a noise source.

Dimming back down JUST EVER SO SLIGHTLY causes the noise to fall back to N00. Three devices total in three different locations on two different circuits.

So my workaround now is to set these lights slightly dimmed, and hope the “ramp back up to where it was last set” feature keeps them from going back to 100% bright.  Most are ON/OFF lights where I don't want to dim anyway.  But I'm sure something will get hosed and the slight dim setting will be lost, I'll get noise, and I'll have to re-fiddle.

Any thoughts from the helpful audience here?  Can others verify my observations?

[HA Dave]

....... I finally find that the X10 XPDI3 and WS12A modules are putting out noise when at full brightness.  ........  Seems like a noise source.

Could these/those be halogen bulbs.... or just drawing too much (or at the max of) wattage? Just curious.

[JeffVolp]

The X10 transmission takes place within the first millisecond after each zero crossing of the powerline.  That is when a dimmer set to almost 100% brightness will turn on.  (Actual 100% would turn on at the zero crossing itself.)

The XTBM looks for several kinds of noise.  It is most sensitive to noise that can mimic X10 transmissions, but it will also register particularly large transients that are shorter than a normal X10 transmission when they occur inside the X10 reception window.  As I recall the ringing had to last at least half as long as that required for a X10 module to register a logic "1".  The rational is that if there is something there now, it might get worse.

Most dimmers include an inductor to prevent RF radiation by damping the very fast rise in current when the dimmer turns on.  Depending on how well damped the powerline is, there can be some ringing as that inductor dumps the energy it absorbed back onto the powerline.  So it is entirely reasonable that a dimmer switching on during the X10 transmission window could produce a transient that is picked up by the XTBM.  While that can register as a strong noise level, understanding the source, I doubt that it could increase in length enough to cause a problem.

During the design stage, the XTBM actually displayed several types of noise - long-term average, short-term average, peak, and transient - where decoded commands are now displayed.  That presented a bunch of confusing numbers, and I thought a single number indicating the worst-case noise would be the easiest to use.

FYI:  I just checked three heavily loaded dimmer switches here (Leviton 16383) all feeding incandescent loads.  (Those lights are in areas only used when we have guests, so they are still incandescents.)    On one circuit powering 360W of incandescent load, the XTBM on that circuit toggles back and forth between N.00 and N.01 when at full brightness.  The XTBM registers N.08 for the transient when the brightness is set one PalmPad dimmer step down from full-on, The transient moves out of the X10 transmission window with another step down, and the noise indication goes away.  Neither of the other dimmers results in any noise indication on the XTBM, regardless of the brightness.  Those are powering similar high wattage loads.

Jeff

[Brian H]

Not sure if this will make any difference, but Jeff may know.

The older Lamp Modules with out Soft Start have the coil in the Neutral lead. From the Neutral input pin to the Neutral output socket. The Triac was between Line input pin and the controlled Line to the socket.

The new Soft Start have a brown wire from the Neutral input pin and the Neutral Output socket. The triac is fed from the Line and its output is connected to the coil with the other end on the controlled line output. Also the power supply is now around 5 volts and drops to about 4.5 when the triac is being driven. Yes no Zener regulation diode any more. Just smaller dropping cap and larger series resistor to limit the voltage from the current being drawn.

[JeffVolp]

Since it is a two-leaded device, it shouldn't matter which lead the inductor is in from a functional standpoint.  However, I would prefer the new design with the neutral passed directly through.

I'm not sure how you can get away without some sort of shunt regulator in a transformerless power supply to prevent excessive voltage when lightly loaded.  If there is no zener clamp, perhaps it is a function of the controller chip itself.

Jeff

[dave w]

Also the power supply is now around 5 volts and drops to about 4.5 when the triac is being driven. Yes no Zener regulation diode any more. Just smaller dropping cap and larger series resistor to limit the voltage from the current being drawn.

Well I wonder how much X10 are saving on that excised zener? Yeah, like Jeff said the regulation could be in the chip now, but I am betting once again that X10 will have shot themselves in the foot.

[Brian H]

The new Microcontroller IC looks like it can run on a much wider voltage range. I believe 2.4 volts to 5.5 volts. The voltage varried in my soft start one on an average 5.2 triac off to 4.6 volts when the triac was being driven.

I don't think they save much on the diode but changing the series cap from .68uf to .22uf; the series resistor from 100 Ohm 1/2W to 330 Ohm 1/2W and the power supply filter cap from 220uf to 100uf may have shaved a few cents off.
The new appliance module has basically the same parts except an added Zener diode on the DC side of the supply and it measured 4.2 volts in my tests.