<azonenberg>
ok so, updates... i had lain collect some data for me on the probe with her 5 GHz scope
<azonenberg>
There is still some ringing in the ~1 GHz band on the rising edge that i believe is due to ground lead inductance of the wire-based ground, we're going to try again with some less inductive ground accessories shortly
<azonenberg>
but we won't have final results until the new probe PCB spin comes back, with the updated tip design so i can use a shorter, less inductive tip
<azonenberg>
overall it looks like we have a solid ~2 GHz probe right now and there's potential to push it further
<azonenberg>
here's the current design, i still have to do the impedance matching of the tip socket
<azonenberg>
the probe body and attenuator are going to be much higher bandwidth, right now the tips and grounds are likely to be our limiting factor
<azonenberg>
i expect the highest bandwidth is going to be the solder in probe, because it will be CPW *all the way* from sma down to the solder-in tip
<azonenberg>
but we'll see
<Degi>
cpw?
<lain>
coplaner waveguide
<Degi>
Ah yes
<Degi>
I got a fun idea, a waveguide probe with integrated downconverter
<Degi>
Like a free air waveguide in the 24 GHz range or so
maartenBE has quit [Ping timeout: 265 seconds]
maartenBE has joined #scopehal
<azonenberg>
ok, sonnet is running on the first tests to figure out the cpw layout for the tip
<azonenberg>
doing a few runs to get a rough idea then parameter sweep to fine tune
alexhw has quit [Ping timeout: 250 seconds]
alexhw has joined #scopehal
<azonenberg>
33 ohm impedance for the first thing i tried, that's lovely :p
<Degi>
hmm shouldnt it have higher impedance? Or is it a 50 ohm one?
<azonenberg>
well it's more, i was hoping i'd be closer to 50 on the first try
<azonenberg>
I have it up to 44 now
<azonenberg>
it looks like i will need a ground plane cutout under the tip connection, as well as pulling back the top layer CPW ground a bit
<Degi>
Hm will it be a 1:1 probe?
<azonenberg>
no it's 10:1. And i'm trying to match the impedance of the tip
<azonenberg>
this is before the first resistor
<Degi>
Hm why not make the tip Z 500 ohm then
<azonenberg>
do you know how hard it is to get a 500 ohm impedance on a stackup designed for 50?
<azonenberg>
the tip is wider than a 50 ohm trace
<azonenberg>
i'm having to fight to bring it DOWN to 50
<azonenberg>
well up i guess, but you know what i mean
<azonenberg>
i'm adding ground plane cutouts, pulling the ground back
<azonenberg>
and this lastest round only got up to 48
<Degi>
Well if you could get buried vias it could be possible lol
<Degi>
Hm 0.1 mm stackup?
<azonenberg>
0.508 mm RO4350B
<azonenberg>
but it's a 1.4mm wide footprint
<azonenberg>
making a 1.4mm wide trace 0.5mm above a ground plane be 50 ohms is not really a thing :p
<Degi>
Heh
<azonenberg>
right now i have the top layer ground pulled back 175 μm, instead of the.... 100 μm, i think, that the normal CPW uses
<azonenberg>
and the inner layer ground completely removed in that 175 μm gap, as well as under the trace itself
<azonenberg>
that should get close
<azonenberg>
but i'm running the model with a pretty small cell size to get more accuracy so it's slow
<Degi>
Hm I guess when you place the caps rly near to the front it should be fine
<azonenberg>
and using a thick-metal model that takes the cross sectional area of the trace sidewalls into account
<Degi>
Ideally something where you can continuously vary the dielectric thickness would be neat
<azonenberg>
keep dreaming :p
<Degi>
Hmm whats the dielectric loss of 3D printer resins like lol
<azonenberg>
uncontrolled and varaible? :p
<azonenberg>
anyway the other thing i have to work out is the ground/tip inductance.
<Degi>
Hmm yesterday you talked about solvent based epoxy, is that different from 2 component epoxy?
<azonenberg>
epoxy resins used as coatings are often solids that have to be dissolved in something
<Degi>
Ah
<azonenberg>
others "100% solids" use liquid components that react completely so there's not much outgassing
<azonenberg>
the antistatic epoxy i used on the floor of my lab is something like 30% alcohol by volume
<Degi>
Yes I have that at home
<Degi>
huhh antistatic epoxy
<azonenberg>
so when you're spreading several liters of it on the floor, there is substantial evaporation :p
<Degi>
Hm why arent surface coatings 2 component?
<azonenberg>
It's still 2 component
<azonenberg>
but one of the components is a solid dissolved in a solvent
<Degi>
Ah yes solvent and epoxy, but why not 100% solid
<azonenberg>
Some monomers aren't liquid at room temperature i guess
<azonenberg>
anyway it's mixed with some kind of carbon particles or something to make it weakly conductive
<azonenberg>
you put copper tape around the perimeter of the room and ground it
<Degi>
NNeat
<azonenberg>
then paint the epoxy over that
<azonenberg>
i did sheet resistance testing, it was a few M ohms per square iirc?