Jim,

On 05/22/2016 03:58 AM, Jim Palfreyman wrote:

Good that you have found the offender, but have you been able to remedy
it by other means than turning the AC off?
I think others H-maser owners would love to know, and potentially the
vendor you have.

Cheers,
Magnus

Hi

That’s a pretty good example for the “why you don’t do your timescale based on
a single brand of gear / single setup” file.

Thanks for sharing!!

Bob

Hi Jim,

Thanks much for the update. I can see how this was a pain to track down.

For those that don't remember the issue, the archive starts here:
https://www.febo.com/pipermail/time-nuts/2015-December/094904.html
And note that 1 / 0.704 ns = 1420 MHz, the frequency of a H-maser.

So it's either electrical or magnetic or seismic, yes? Does it happen every time the HVAC turns on/off? Can you run a high-resolution phase comparator during the event to find the time constant of the phase jump / cycle slip? 1 kHz sample rate should be more than enough. That may help narrow down which circuit is at fault.

/tvb

----- Original Message -----
From: "Jim Palfreyman" jim77742@gmail.com
To: "Discussion of precise time and frequency measurement" time-nuts@febo.com
Sent: Saturday, May 21, 2016 6:58 PM
Subject: [time-nuts] Maser 0.7 nsec jumps solved

Interesting math: Hydrogen maser frequency standards use the 1420 MHz line.

Period of 1420MHz is 0.7 ns.

It's not so clear to me that the maser itself is being disrupted, it seems
more likely the external noise is inducing an extra count or causing a
count to be slipped.

A different Australian observatory 1400 MHz RFI problem:
http://arxiv.org/abs/1504.02165 "Subsequent tests revealed that a peryton
can be generated at 1.4 GHz when a microwave oven door is opened
prematurely and the telescope is at an appropriate relative angle. Radio
emission escaping from microwave ovens during the magnetron shut-down phase
neatly explain all of the observed properties of the peryton signals."

Tim N3QE

On Sun, May 22, 2016 at 7:43 AM, Magnus Danielson <
magnus@rubidium.dyndns.org> wrote:

As far as a remedy goes we are going to try a solid state relay that only
switches on at 0V in the AC waveform. This should slow the inrush current,
and hopefully the magnetic impulse.

If this doesn't work, then a better model of air conditioner might have to
be installed. These ones do come on with a big "thump".

Jim Palfreyman

On 22 May 2016 at 21:43, Magnus Danielson magnus@rubidium.dyndns.org
wrote:

In message CALH-g5b0C+aBz53MUM7bkJ00441AvNxwFLc2RgjgPj=k1FPOKA@mail.gmail.com, Jim Palfr
eyman writes:

If that is not enough, consider a small VFD drive, and ramp up voltage+frequency
over 10 seconds.  It's slightly more intrusive, as you will need to remove the
starting capacitor from the motor to install the VFD.

--
Poul-Henning Kamp      | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG        | TCP/IP since RFC 956
FreeBSD committer      | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.

On Sun, May 22, 2016 at 11:15 PM, Jim Palfreyman jim77742@gmail.com wrote:

As far as a remedy goes we are going to try a solid state relay that only

If the load being switched on is inductive, it would be better to switch
the AC waveform at the voltage peaks, not at 0V.  This might seem
counter-intuitive, but it's real.  Switching on at the 0V crossing may
maximize the current pulse through the magnetics.

OTOH, if the origin of the impulse is mechanical in nature, neither remedy
may help.

Regards,
Andy

Am 23.05.2016 um 05:15 schrieb Jim Palfreyman:

In the context of transformers and motors, switching on at 0V is
actually the worst point in time.

< https://de.wikipedia.org/wiki/Einschalten_des_Transformators > (German)

<
http://electrical-engineering-portal.com/practical-considerations-of-transformer-inrush-current

regards, Gerhard

In message 57441ACA.8070609@arcor.de, Gerhard Hoffmann writes:

Well...

The motor is almost certainly an "Squirrel-cage" induction motor and
that means it is three phase, although one of the phases is probably
created with a "starting capacitor".

So which of the three phases is going to be the lucky one that
switches at maximum voltage, or are you going to switch the phases
on sequentially ?

A 4kW Variable Frequency Drive costs less than $1k and allows you
to control both the voltage/time and the frequency/time both
during startup and during rundown.

I wouldn't bother fuzzing around with hacks - I'd just go for the
known-to-work solution.

--
Poul-Henning Kamp      | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG        | TCP/IP since RFC 956
FreeBSD committer      | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.

LTspice shows  switching  at 0V is the best point in  time.  With no
flux in the magnetics, the inrush current is limited by  the circuit
resistance. The magnitude is given by Ohm's law: I = E / R.

Switching at 0V,

I = E / R
= 0 / R
= 0A

However, switching  at  the peak of the voltage can  give  very high
inrush currents. You can verify this with LTspice. The  schematic is
at

http://www.pst.netii.net/misc/48b961a6.gif

There are two circuits. They are identical except the second has the
applied voltage  shifted  by 90 degrees. The  inductors  have  1 Ohm
series resistance (not shown.)

The waveforms are at

http://www.pst.netii.net/misc/48b96217.gif

The currents are IL1 and IL2.

Switching at  0V, IL1 starts at zero. It rises to a  peak  of 900mA,
then falls  back  to  zero. The DC offset takes 3  or  4  seconds to
decay, then the current is stable at zero +/- 450mA.

Switching at the peak, IL2 is

I = E / R
= 169.7 / 1
= 169.7A

It takes over twice as long for the starting surge to decay. I could
still detect it past 8 seconds.

This analysis shows switching at 0V is the best option.

If you wish to do further analysis, the LTspice .ASC and  .PLT files
are at

http://www.pst.netii.net/misc/48b96262.zip

The wikipedia  article  states "To avoid magnetic  inrush,  only for
transformers with  an air gap in the core, the inductive  load needs
to be synchronously connected near a supply voltage peak."

https://en.wikipedia.org/wiki/Inrush_current

Clearly, from  the  above LTspice waveforms, switching  at  the peak
gives the  highest inrush surge that is possible to obtain.  It also
takes the longest time to decay.

MRM

In message b222ff$74b01p@out.teksavvy.com, "Mike Monett" writes:

You ought to have stopped and wondered at the "with an air gap in the core"
and pondered if you SPICE models use of an ideal inductor was appropriate
for something as electrically complex as an induction motor...

--
Poul-Henning Kamp      | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG        | TCP/IP since RFC 956
FreeBSD committer      | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.

On Wed, May 25, 2016 at 9:59 AM, Mike Monett timenuts@binsamp.e4ward.com
wrote:

Er, no.  It's an RL circuit for which:

V(R) = E(1-e^(-(R/L)t)

so I(R) = I(L) = (E/R)(1-e^(-(R/L)t)

for t close to zero where E is essentially constant, I is going to rise
with the time constant of R/L which for a 1H inductor and 1 ohm resistance
is 1 second.

I suggest running the simulation with a pulse voltage source, Tdelay 1ms,
period 1s, Ton 0.5s (Period and Ton are fairly arbitrary, the Tdelay is
important).

You only need to run the simulation for a few ms.  You will get a very
different result.

Now as Poul-Henning Kamp suggested, an induction motor is anything but an
ideal inductor...

Am 25.05.2016 um 18:59 schrieb Mike Monett:

No, it doesn't. :-)

First, the single inductor does not represent a transformer; the second
inductor
and the coupling declaration ( style: K1 L1 L2 0.99 or so) and the load are
missing.

The most important thing is that the Inductor is nonlinear which is not
represented in the model. If there has remained some magnetism in
the core from previous operation, the transformer won't be able to further
increase the magnetism as needed to induce an opposing voltage
in the primary winding. When the core saturates, the inductance
collapses and leaves only the copper resistance to limit the current.

The catastrophe builds up in the first 90 degrees of the source wave,
not right at the start.

In the simulation, the current at t=0 is at its maximum already when at
t=0 the input voltage is just switched on. You'd expect that from a
capacitor, never from an inductor.

The reason is that the simulation does not really start at t=0, but
much earlier. The simulator computes the conductance matrix,
applies the initial sources and waits until everything has calmed down.
That may require repeated recalculation of the matrix to respect
nonlinearities.

Your circuit must contain hidden resistors btw, otherwise the computation
of the initial condition at "t<=0" would result in numeric overflow
as required by an assumed initial DC voltage across an inductor, before
the transient simulation.

One can enforce initial conditions with statements like  .IC v(my_node) = 0V

regards, Gerhard

BTW I've got the first 20 pcs. of the VCXO carrier / voltage regulator /
lock to reference / squarer / iso amp or frequency doubler / 1pps board.
That won't be soldering for beginners or jittery hands.  :-)

Hi All,

Thanks so much for your input and thoughts. It has really proved helpful
here at the observatory.

As it turned out we easily obtained a zero-crossing solid state relay so we
thought we'd try it.

And, drumroll......

It made things so much terribly worse than ever before. (As predicted by
many of you above.)

We are going to try a SSR that switches at the peak - but we need to order
one. So stay tuned on those results.

There is of course the "move the bloody thing far away from the maser"
solution which could end up being a serious option. These air conditioning
units are small and cheap (window-type), so we are trying to find the
cheapest solution - and if that ends up being some ducting - so be it!

Jim Palfreyman

On 26 May 2016 at 13:13, Andy AI.egrps+tn@gmail.com wrote:

Jim,
My head is precisely in the get it away from the unit approach.
Did not mention it for the following reason.
Its well understood and for time-nut boring. Its more fun to figure out
peak currents and such.
But I tend to fall into the get it done camp and move on.
That doesn't mean its a simple answer. Flex duct is bad. So only use it in
the last few feet. You want low resistance hard duct.
Then the fun of the return feed. Often overlooked and poorly considered.
Regards
Paul
WB8TSL

On Fri, Jun 3, 2016 at 1:00 AM, Jim Palfreyman jim77742@gmail.com wrote:

One more thought.
Is the aircon on the same power phase as the maser?
Are you split phase in the facility at least.

On Fri, Jun 3, 2016 at 9:24 AM, paul swed paulswedb@gmail.com wrote:

On Fri, 03 Jun 2016 12:37:26 -0400
"Mike Monett" timenuts@binsamp.e4ward.com wrote:

Actually, spice (the engine behind LTspice) does a DC analysis before
almost all modes of operation. This DC analysis has the intention to
start the circuit from a steady-state point and thus to reduce simulation
time. In order for this to work properly, you have to specify the DC voltage
and currents for all sources correctly. Spice messes this up at times
making the first part of a transient simulation worthless (it has even
worse problems when you do an AC analysis). Additionally LTspice hides
too much of these small complications for the problems to be visible to
the untrained eye and also at times makes it harder to provide the correct
values. Thus, caution is advised.

The general rule of "Never trust a simulation you haven't
forged yourself" applies.

I only had a quick glance at your webpage, but it seems that you used
the standard LTspice transformer model. Unfortunately, this is not a
good model to study this kind of behaviour. For one, the only loss considered
in the model is the winding coupling, it doesn't even directly consider
resistive losses in the windings. In this case, the two most important effects
that you need to include are saturation and core losses, which are both
frequency dependent. The cores of electric machine transformers are very
poor when it comes to their "high" frequency behaviour. Where high frequency
starts somewhere closely above mains frequency. Ie 1kHz is already so far off
that somewhere around 90% of the energy would be dissipated in the core.
The sharp rise in voltage and the leading inrush current have frequency
components that are way higher than mains frequency. Hence the linear model
you used will give inaccurate results, to put it mildly.

Unfortunately, building an accurate transformer model in spice is not
easy and depends on higher order functions that might or might not be
available in the flavour you use. Not to mention that you will need
to have good (measured) numbers on the non-ideal behaviour of a transformer,
which are also not easy to get by.

			Attila Kinali

--
It is upon moral qualities that a society is ultimately founded. All
the prosperity and technological sophistication in the world is of no
use without that foundation.
-- Miss Matheson, The Diamond Age, Neil Stephenson