I've got parasites

[Tim Williams]

I know no one can tell me exactly why, but someone can at least give me some
thoughts here. Ok, my problem. First of all, when I set my Tek 475 to
2mV/div, 200MHz BW with an open circuit probe, I get about 20mV (it's a 10x
probe) of noise, primarily at approximately 90MHz. I've tried turning off
basically everything in the house and that hasn't stopped it. (Amazingly, I
didn't see any change at all when turning off my computer and monitor just
five feet away!)

So I've reassembled my medium power induction test circuit.
http://webpages.charter.net/dawill/Images/Induction701.jpg
(All neet and shiny with tight 12AWG hookups and bypass caps.
Same old sextet of STW11NB80's.)

When I turn on the drive circuit,
http://webpages.charter.net/dawill/tmoranwms/Elec_SG3524Driver.gif
I immediately get something like 0.5V of amplitude modulated noise across
the ground-return resistor (0.1 ohm, eh probably inductive). It's a perfect
keyed CW envelope, modulated by the 5-200kHz drive signal as it is.

I tried sniffing around with a balanced twisted pair with 1" dia. loop on
the end, but that didn't tell me much besides the circuit is somewhere
radiating noise...no shit!



Tim

 

[John Miles]

I know no one can tell me exactly why, but someone can at least give me some
thoughts here. Ok, my problem. First of all, when I set my Tek 475 to
2mV/div, 200MHz BW with an open circuit probe, I get about 20mV (it's a 10x
probe) of noise, primarily at approximately 90MHz. I've tried turning off
basically everything in the house and that hasn't stopped it. (Amazingly, I
didn't see any change at all when turning off my computer and monitor just
five feet away!)

FM radio station, most likely.

I tried sniffing around with a balanced twisted pair with 1" dia. loop on
the end, but that didn't tell me much besides the circuit is somewhere
radiating noise...no shit!

Sometimes the easiest way to track down low-level stability problems is
to "build" your circuit in LTSpice. LTSpice has the advantage of
perfect immunity to environmental effects.

-- jm

 

[Tim Williams]

FM radio station, most likely.

- I don't know of one any of siginificant power nearby. I'm not in a big
city. The college has one but I don't think it's that strong.

There is a local AM geezer (WGEZ, really!) radio station that gets into just
about everything, but being a non-relatavistic rate of voltage change, I can
squash that easily enough in most cases, including here.

For Pete's sake I can't even short the VHF signal at the scope connector,
how does that work!?

At any rate, as much as it has to be an external signal, how would you
explain the keying effect when the drive circuit is turned on? The local
circuit would have to be acting as a passive resonator or something!

Sometimes the easiest way to track down low-level stability problems is
to "build" your circuit in LTSpice. LTSpice has the advantage of
perfect immunity to environmental effects.

Well, that won't help much, because I already know the circuit works, on
paper and in reality. Besides which, simulators tend to go with perfect
connections. I don't know about LTSpice in particular, but I'm guessing it
wouldn't know the difference if I had one of the MOSFETs on the other side
of the room on some ten foot pigtail leads.

Tim

 

[Mark]

90 MHz would be about right for a college station..

how far away is it?

And what about if you disconnect the probe complely from the
scope...see it then?


Mark

 

[John - kd5yi]

- I don't know of one any of siginificant power nearby. I'm not in a big
city. The college has one but I don't think it's that strong.

There is a local AM geezer (WGEZ, really!) radio station that gets into
just
about everything, but being a non-relatavistic rate of voltage change, I
can
squash that easily enough in most cases, including here.

For Pete's sake I can't even short the VHF signal at the scope connector,
how does that work!?

You mean, you remove probes and all connections to the scope and short the
BNC connector at the scope with, say, a screwdriver and the problem is
still there? If yes, the problem is internal to the scope.

There is usually a switch near the input that selects DC/AC/GND. Put it to
GND. Still there? Yes = scope internal problem.

2 cents, please.

Cheers,
John

 

[Ian Stirling]

I know no one can tell me exactly why, but someone can at least give me some
thoughts here. Ok, my problem. First of all, when I set my Tek 475 to
2mV/div, 200MHz BW with an open circuit probe, I get about 20mV (it's a 10x
probe) of noise, primarily at approximately 90MHz. I've tried turning off

Set the trigger right, and see if you can resolve the frequency accurately.
Now, tune a FM radio to that freq, and see if it is or isn't.

 

[Tim Williams]

You mean, you remove probes and all connections to the scope

Sorry for the confusion, I was using a tee connector. Still, that's
balls-all inductance, I don't see how the signal can get past a short like
that, at least for another few hundred MHz.

There is usually a switch near the input that selects DC/AC/GND.
Put it to GND. Still there? Yes = scope internal problem.

With no antenna, or the input GND'd, it shows no signal. Even a 2" stub of
wire will register a signal.

87.5MHz is the exact value I get, something like 7 cycles in 8 div at
10ns/div. Ten percent error smears it over half the damned frequency band,
so it could be only what, 40 channels on the dial, or one of 40 more that
don't exist! (under 88MHz) I don't have a frequency counter high enough, or
any other way to lock it more precisely.

BTW: there is a college radio station a few blocks down the street(!), but I
don't know how strong it is (tell me, how many FM radios do you own that
came with AVC and an indicator? .

Does anyone have an explanation how the induction drive circuit appears to
be amplifying the signal in modulation to the drive output?

Tim

 

[Rich Grise]

87.5MHz is the exact value I get, something like 7 cycles in 8 div at

Turn your TV to channel 6, and listen as you watch the trace.

Or, turn your TV _off_. (or maybe your computer. and unplug the microwave
and VCR.) ;-)

Good Luck!
Rich

 

[Tim Williams]

Or, turn your TV _off_. (or maybe your computer. and unplug the
microwave and VCR.) ;-)

I tried just about every appliance in the house, including the network (the
CAT5 does happen to run quite near the experiment... which caused network
troubles when I was playing with small CW Tesla coils , but there was no
change.

Tim