[time-nuts] AC line distortion [Was: HP 105B Battery, the saga continues]
kb8tq at n1k.org
Tue Oct 4 07:36:30 EDT 2016
A lot of the distortion on the AC line is locally produced. Consider a very normal
bridge rectifier running into a capacitor. It draws “all” the current in narrow spikes
near the peaks of the sine wave. A half wave rectifier would be even worse (only
one spike per cycle). That highly non-linear load will work against the impedance
of the distribution system to distort the AC voltage.
It’s hardly a “electronics only” sort of problem. If you attach a motor to the line, it
likely is driving *something*. If that something is imbalanced (think of a fan) or has
a cycle (think of a compressor), that non-linear power demand reflects back on the
Needless to say none of this makes the power company very happy. They would
much prefer to drive nice zero phase angle linear loads ….. Their generators are
essentially giant sine wave generators. Load mismatch wears them out quicker
than a “proper” load.
> On Oct 3, 2016, at 8:52 PM, Jeremy Nichols <jn6wfo at gmail.com> wrote:
> As an experiment, I bought an AIMS sine-wave inverter for the 105B Quartz Oscillator. The inverter has a built-in transfer switch that is supposed to allow the load to operate from the AC line and automatically switch to battery/inverter should the AC power line fail.
> In fact the thing seems to work—the output is a nice 118 VAC sine wave measuring 60.189 ± 0.003 Hz and the transfer switch is fast enough that the 105B doesn't seem to notice the change. The "AC Interruption" light doesn't light and I don't see a flicker of the 5 MHz output on my scope.
> Just for the fun of it, I connected a filament transformer and ran the low voltage into my distortion analyzer. The result was about 5% distortion for the inverter and 1.5% for the AC line. This got me to wondering, we've discussed the AC power line frequency at length but not other "qualities" of that "signal.' I was surprised that the AC line had so much distortion but it's a subject I've never considered. Has anyone in this group looked at this? [Yes, this is perilously close to not being appropriate Time-Nuts discussion matter—sorry!]
> On 9/18/2016 4:33 AM, Scott McGrath wrote:
>> That NiCad pack is part of the power supply and as Jeremy points out is part of the filter system. And so one needs to restore it as part of the instrument as even the 28V external power supply floats these cells and trips power interruption indicator if lost
>> Power supply is not terribly hard to fix and the small signal transistors can be replaced with 2N 2222,3904 and 3906'es depending on rating. You don't even need a extender a Huntron tracker or similar current limited lissajous bridge will identify failed or leaky caps and semiconductors
>> Remember HP did nothing without a good engineering reason and that plate is there for RF shielding to prevent stray sources coupling with the outputs
>> If a proper rebuild is too expensive I'd suggest selling it on the well known auction site rather than hacking it up as 105's have been selling in the hundreds regardless of condition
>>> On Sep 17, 2016, at 10:16 PM, Jeremy Nichols <jn6wfo at gmail.com> wrote:
>>> How did you come up with the 33,000 uF number, Perry, and is it one big capacitor or lots of little ones tied together? The big cap will also filter out some of the remaining ripple in the power supply that may have been managed by the ni-cad battery.
>>>> On 9/17/2016 3:50 PM, Perry Sandeen via time-nuts wrote:
>>>> Where the nicad pack was located one can put in 33,000 uF of Nichicon 105C caps for $20 for a buffer hold over. <snip>
>>> time-nuts mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>> and follow the instructions there.
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts