Capacitor Discharging with CMOS Gate

[John Larkin]

John, I imagine your output circuit is something
like this, to source terminate a 50-ohm coax?

7sz125 39
--|>o---/\/\----(o) BNC

Nope, no resistor. The intent was to allow anyone to check the
oscillator frequency, or to allow one module to be the master that
others phase-lock to. I figured that if they needed fanout or cable
drivers, they could be provided somehow. If another module or two is
next door, just a little jumper will work.

Actually, its...

7sz125
--|>o-------(o)--------R----+------+-----comparator---> to PLL

SMB | |
| |
C L
| |
| |
| |
gnd gnd



where the LC is resonant at 10 MHz and the comparator works against
ground. This way, any input waveform, even mangled by cable
reflections, can drive the PLL. I never expected them to apply a
sinawave to it, but we still can't figure out why a 3 dBm sine would
kill the tristate driver.

The board itself is a 4-channel ARB with fairly complex control
memory-- conditional jumps, loops, flags, analog and digital outputs.
As usual, it's the simple parts we have the most trouble with... the
complex stuff just works.

Next rev, I guess I'll do a series RC to the connector (just in case
anybody applies DC) and schottky clamp the gate. But I suspect this
customer is going to insist that we still leave it off the board.

John

 

[John Larkin]

Are you saying these buffers are off a bidirectional port on your
module and the customer is applying a 3dBm clock to it? How is this
terminated for input mode drive?

Bandpass filter and comparator, as noted elsewhere.

Some generators are very particular
about termination and can do all kinds of oscillations and non-linear
things when driven into Hi-Z, to include putting DC on the line as well
as damaging peak voltages. This may be especially true of a video grade
signal generator with lots of programmability in amplitude and DC
offset. This would also explain why the failures are not occurring in
the field. Did the numbskulls even non-invasively scope out their input
waveform, or maybe your circuit blows so fast this is not possible.
Obviously you need to hold them by the hand and walk them through it.
Pathetic!

They are not idiots, they are not pathetic, and we've had lots of
intelligent conversations and emails trying to figure this one out.
I've applied 20 volts p-p from a 50 ohm generator and haven't managed
to blow the chips. They've made lots of measurements.

Should I recommend you as a consultant? Could you make room from your
busy schedule?

John

 

[Joerg]

Bandpass filter and comparator, as noted elsewhere.



They are not idiots, they are not pathetic, and we've had lots of
intelligent conversations and emails trying to figure this one out.
I've applied 20 volts p-p from a 50 ohm generator and haven't managed
to blow the chips. They've made lots of measurements.

Should I recommend you as a consultant? Could you make room from your
busy schedule?

Do you have any spare ADC inputs on this system? Might be a good idea to
quietly log the DC level at this input node, with time stamp. RC filter
it so that big excursions still show. I remember a case where there was
a totally unexplainable interruption every so many seconds. A SW routine
caught that time interval. Everyone thought real hard and while gazing
out a window there was a weak reflection way in the distance, every x
seconds. Same repetition rate as the glitch (this is where you can
really show off an aviator watch ...). Closed the metallic blinds,
problem gone. Aha! It was some kind of military radar station on the
next hilltop.

 

[Jim Thompson]

Do you have any spare ADC inputs on this system? Might be a good idea to
quietly log the DC level at this input node, with time stamp. RC filter
it so that big excursions still show. I remember a case where there was
a totally unexplainable interruption every so many seconds. A SW routine
caught that time interval. Everyone thought real hard and while gazing
out a window there was a weak reflection way in the distance, every x
seconds. Same repetition rate as the glitch (this is where you can
really show off an aviator watch ...). Closed the metallic blinds,
problem gone. Aha! It was some kind of military radar station on the
next hilltop.

Sounds like radar from Moffett Field near San Jose.

Use to drive me nuts when I was at Philco-Ford, Santa Clara.

...Jim Thompson

 

[Michael A. Terrell]

Sounds like radar from Moffett Field near San Jose.

Don't ignore those TV weather RADAR systems.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

[Fred Bloggs]

Bandpass filter and comparator, as noted elsewhere.

Oooooh yeah- almost guaranteed to blow the sig gen...

They are not idiots, ...

Should I recommend you as a consultant? Could you make room from your
busy schedule?

There won't be any friggin consulting to it, have walked in many a
situation to correct a multi-month 'mystery' in five minutes with people
like that. Where are these people, WrightPat or Hanscom or
someplace...too far if farther than five minutes away, especially for
any defense related *crap*.

 

[John Larkin]

Oooooh yeah- almost guaranteed to blow the sig gen...

Yeah, not many sig gens can handle a load like this...


in (o)------------470r-----+-------+---------comp+
| | |
gnd | | +---comp-
L C |
| | |
| | gnd
| |
gnd gnd


Terrifying, just terrifying.


John

 

[Fred Bloggs]

Yeah, not many sig gens can handle a load like this...


in (o)------------470r-----+-------+---------comp+ |
| | gnd | | +---comp- L C
| | | | | | gnd | | gnd gnd


Terrifying, just terrifying.


John

Is that 470R or 47R, either way, I know of at least one fairly popular
lab grade type that would not like that at all...you should have used a
constant impedance filter. See the tutorial in US6608536, know-it-all.

 

[Jim Thompson]

John Larkin wrote: [snip]

Yeah, not many sig gens can handle a load like this...


in (o)------------470r-----+-------+---------comp+ |
| | gnd | | +---comp- L C
| | | | | | gnd | | gnd gnd


Terrifying, just terrifying.


John

Is that 470R or 47R, either way, I know of at least one fairly popular
lab grade type that would not like that at all...you should have used a
constant impedance filter. See the tutorial in US6608536, know-it-all.

Cute. Patent what everyone already knows :-(

...Jim Thompson

 

[Joerg]

Is that 470R or 47R, either way, I know of at least one fairly popular
lab grade type that would not like that at all...you should have used a
constant impedance filter. See the tutorial in US6608536, know-it-all.

Then ditch it and get a real signal generator

 

[Fred Bloggs]

John Larkin wrote:
[snip]

Yeah, not many sig gens can handle a load like this...


in (o)------------470r-----+-------+---------comp+ |
| | gnd | | +---comp- L C
| | | | | | gnd | | gnd gnd


Terrifying, just terrifying.


John

Is that 470R or 47R, either way, I know of at least one fairly popular
lab grade type that would not like that at all...you should have used a
constant impedance filter. See the tutorial in US6608536, know-it-all.

Cute. Patent what everyone already knows :-(

...Jim Thompson

I know, only been common knowledge 100 years, didn't bother reading past
that he had the time constant equalities right...

 

[Fred Bloggs]

Then ditch it and get a real signal generator

The best damned RF generators anywhere are Rohde & Schwartz, it's the
programmable video grade jobs that act up...

 

[Fred Bloggs]

The best damned RF generators anywhere are Rohde & Schwartz, it's the
programmable video grade jobs that act up...

I take that back, any RF gen with reasonably precise ALC will not like
JL's circuit, usually they just passively shut-down. The solution is
very high tech, get that RL to -3dB or more usually does it, duh-huh, a
coaxial attenuator...

 

[Joerg]

Then get the rep who sold them to you out there and read him the riot act.

I take that back, any RF gen with reasonably precise ALC will not like
JL's circuit, usually they just passively shut-down. The solution is
very high tech, get that RL to -3dB or more usually does it, duh-huh, a
coaxial attenuator...

Hmm, I use lots of generators and my loads are typically rather weird.
Transformers with all kinds of resonances and the like. I expect them to
survive and not shut down. And they all complied (so far).

The only RS generator I have here is the old SMF RF generator. The most
spectrally quiet gen I ever had. But the steel tube in the oscillator
has become tired after roughly half a century and it's pretty much
unobtanium :-(

 

[Baron]

The only RS generator I have here is the old SMF RF generator. The
most spectrally quiet gen I ever had. But the steel tube in the
oscillator has become tired after roughly half a century and it's
pretty much unobtanium :-(

Which tube is that ?

 

[Barry Lennox]

Which tube is that ?

Don't know about that one, but I have an old Philco-Sierra generator
that uses an ML7211X. They are pretty rare, anybody know of a
source?

Barry

 

[Joerg]

Which tube is that ?

I'll have to open it up again, some German type (AF...?). Maybe I should
rework it so a newer Svetlana or Sovtek tube fits.

 

[Baron]

I'll have to open it up again, some German type (AF...?). Maybe I
should rework it so a newer Svetlana or Sovtek tube fits.

If you do, open it up to find out, let me know.

 

[John Larkin]

Is that 470R or 47R, either way,

Our convention is that 470r means four hundred and seventy ohms.

I know of at least one fairly popular
lab grade type that would not like that at all...you should have used a
constant impedance filter. See the tutorial in US6608536, know-it-all.

You know of a generator that blows up when it drives a 470 ohm load?
Let me know, so I'll be sure to not buy one.

Oh, the reason to keep the input impedance high is so that one 10 MHz
source can lock a lot of modules. The bandpass/comparator makes the
thing pretty much indifferent about waveforms.

John

 

[John Larkin]

I take that back, any RF gen with reasonably precise ALC will not like
JL's circuit, usually they just passively shut-down.

So they shut down when open-circuited, or when plugged into a high-Z
scope input? I've never seen any generator do that.

Once they shut down, how do they ever recover?

John