<miek>
is there any way to skip the connect dialog when launching glscopeclient?
<sam210723_>
passing a bad scope string works
<sam210723_>
eg. "0:0:0"
electronic_eel has quit [Ping timeout: 276 seconds]
electronic_eel_ has joined #scopehal
electronic_eel_ has quit [Client Quit]
electronic_eel has joined #scopehal
<azonenberg>
sam210723_, miek: long term the whole startup/welcome dialog is going to get totally redone
<azonenberg>
Katharina was supposed to be working on that, then disappeared
<azonenberg>
She was out for several months with covid so not sure if complications from that, or just busy at work / lost interest
<azonenberg>
i want to be able to connect to scopes, load recent scopesessions, reconnect to recently used instruments, activate demo mode with an easy button, etc
<azonenberg>
But i've been too busy with other stuff to touch that
<azonenberg>
sam210723_: also with those new commits are you ready for me to review and merge your contributions?
<azonenberg>
or are there still open action items you want to address?
<Famine->
azonenberg, any suggestions for cheap/semi cheap 1000:1 probes ? (low energy, maybe ~200uJ)
<azonenberg>
1000:1? how high voltage?
<azonenberg>
and are you looking for 1M outputs or 50 ohm?
<azonenberg>
a 49.95K resistor in series with 50 ohm coax would form a 1000:1 transmission line probe, and in theory you could use it up to 5 kV with a normal 50 ohm 5V scope input. But i'd be worried about the resistor arcing over and frying the scope and/or me
<Famine->
~50kV
<azonenberg>
So you definitely want a 1M ohm probe then
<azonenberg>
alternatively you could do a 10K:1 (or more) resistive divider probe
<azonenberg>
the other problem with a resistive divider at these extreme values is fast edge
<azonenberg>
since you mentioned uJ i assume this is for pulse applications
<azonenberg>
even a few fF of shunt capacitance across the resistor would allow high voltage transients through
<Famine->
yea, ford coil pack testing. the lovely design of the PCM means if the coil pack dies then the driver fets fry and its $2500 for a new pcm lol
<azonenberg>
if my mental math is right with a 1000:1 probe you've got 30 dB of attenuation, so you'd need *extremely* flat frequency response out to well past the highest freq in the signal to avoid overvolting the scope
<azonenberg>
6 dB of peaking doubles the voltage the scope sees
<azonenberg>
Which is not that much of a non-flatness, proportionally speaking, on a 30 dB probe
<Famine->
ugh lovely and i'll probably have harmonics >500MHz
<azonenberg>
have you ever looked at the derating curve for a typical scope probe?
<azonenberg>
they reduce the max voltage greatly as freq goes up, in part because of this
<azonenberg>
High voltage *and* high frequency is hard
<Degi>
Also the loading can be relatively high from probe capacitance
<azonenberg>
You could probably add lowpass filtering to remove some of those harmonics
<Famine->
good to know. i thought oh this will be easy i only need 10MHz
<azonenberg>
but you can't use a normal matched 50 ohm filter
<azonenberg>
because 50 ohms to DC with this kind of power level would be insane :p
<Degi>
Like a pF at 10 MHz and 10 kV can do 6.2 kVAr load
<azonenberg>
Famine-: Data point: a 500 ohm 0603 resistor is 474 ohms at 500 MHz
<azonenberg>
and that's a flip chip RF resistor
<azonenberg>
the FC series 1K ohm with wraparound terminals is 894 ohms at 500 MHz
<azonenberg>
So you're down by 10% which is not quite as bad as i thought. but still, as resistance goes up with the same shunt capacitance
<azonenberg>
the proportional change goes way up
<azonenberg>
typical resistors in the megohms will... not be megohms at hundreds of MHz
<azonenberg>
now, going with big fat carbon composition or even wirewound (deliberately increasing L) will definitely improve that vs an 0402
<azonenberg>
but this is not something you throw together casually without a LOT of testing and measurement
<azonenberg>
high power RF is something you can easily screw up quite badly