<_whitenotifier-3>
[scopehal-apps] azonenberg labeled issue #360: Allow historical waveforms to be displayed in the plot, used as filter inputs, etc. - https://git.io/JGkK3
<_whitenotifier-3>
[scopehal-apps] azonenberg opened issue #360: Allow historical waveforms to be displayed in the plot, used as filter inputs, etc. - https://git.io/JGkK3
<_whitenotifier-3>
[scopehal-apps] azonenberg commented on issue #360: Allow historical waveforms to be displayed in the plot, used as filter inputs, etc. - https://git.io/JGkKS
<GenTooMan>
hmm you may wish to consider SLA investment casting mold use for some of your scope design. Having "taken apart" a 6Ghz network analyser the use of cast metal for both mechanical and RF shield has potential for you.
<azonenberg>
For production i would probably do CNC'd alumium plate for that kind of stuff
<azonenberg>
aluminum*
<azonenberg>
I don't have the facilities to do casting in house
<azonenberg>
and CNC is easier to outsource
<GenTooMan>
I've seen CNC usages mostly for finishing but casting for the majority of mass production. Just an FYI not a "tell you how it's done" but perhaps there is a good reason for using cast aluminium instead thought.
<azonenberg>
There is a good reason. It's cheaper in volume
<azonenberg>
I'm not making the kind of volume R&S or Keysight are
<azonenberg>
High O(1) cost will dominate the savings in the O(n) term
<azonenberg>
A lot of the tradeoffs are different
<azonenberg>
3D printed metallized plastic for prototyping, or for stuff like the diff probe where flexibility is an application requirement
<azonenberg>
then CNC'd aluminum for small run production
<GenTooMan>
hmm well I've seen it done for 1 off production too (cast aluminium) so I was wondering what was different.
<azonenberg>
The cheap way to do higher volume parts is to CNC a mold, cast aluminum or zinc alloy into it
<azonenberg>
then do minimal CNC to finish them off where you need super flat surfaces, threads, etc
<azonenberg>
Less machine time = less O(n) cost
<azonenberg>
In any case the probes will be plastic all the way to production because flexibility is a design requirement
<azonenberg>
And i don't want the probe exterior conductive to avoid shorting out the DUT if you drop it etc
<azonenberg>
So it's going to be plastic with a metallized coating on the inside, and carefully masked during the spraying step so no metal gets on the outside
<GenTooMan>
for the probe well understood that part. plus their is the flexible circuit for connection of the probe to the scope.
<azonenberg>
Yeah
<azonenberg>
Construction of the later instruments i'll deal with when the time comes
<GenTooMan>
:D
<GenTooMan>
tackle one problem at a time they say
<azonenberg>
oh i'm pipelining like crazy
<azonenberg>
because of manufacturing lead time etc
<azonenberg>
Right now i'm waiting on one CNC job, three SLS prints, and a PCB
<azonenberg>
For three different projects
<GenTooMan>
hmm investment casting sleeps quite an RF doable feature I watched someone use it to make a feed horn for micro wave and satellite stuff.
xzcvczx has quit [Read error: Connection reset by peer]
<azonenberg>
TSP has done some good work on microwave cavity filters etc out of metal spray coated FDM plastic
Degi has quit [Ping timeout: 252 seconds]
Degi has joined #scopehal
_whitelogger has joined #scopehal
_whitelogger has joined #scopehal
_whitelogger has joined #scopehal
_whitelogger has joined #scopehal
bvernoux has joined #scopehal
_whitelogger has joined #scopehal
_whitelogger has joined #scopehal
someone--else has joined #scopehal
someone--else has quit [Quit: Connection closed]
_whitelogger has joined #scopehal
someone--else has joined #scopehal
_whitelogger has joined #scopehal
_whitelogger has joined #scopehal
_whitelogger has joined #scopehal
<azonenberg>
Initial AKL-AD3 hinge prototype came in. Less stiff than my calculations predicted by far
<azonenberg>
So i need to make the spring beam a little bit thicker
<azonenberg>
Second iteration is already printing, i already know it will be too floppy since i made that one less stiff, but it's got some other stuff on it that i wanted to test too so not entirely a waste
<azonenberg>
SMA test board is expected to come this afternoon
<azonenberg>
but not showing out for delivery yet (at local fedex warehouse) so we'll see, today or tomorrow
_whitelogger has joined #scopehal
_whitelogger has joined #scopehal
_whitelogger has joined #scopehal
ericonr has quit [Ping timeout: 272 seconds]
<azonenberg>
SMA test board arrived
bvernoux has quit [Quit: Leaving]
ericonr has joined #scopehal
<GenTooMan>
azonenberg, :D "we don't got it" "then why did you deliver it?" "duh?"
<azonenberg>
lol
<azonenberg>
Ok so, the 901-10510-2 connector doesnt fit well
<azonenberg>
the footprint is too far in from the edge of the board (had to sand it a bit)
<azonenberg>
and the drill holes are undersized. It fit but i literally had to pound it into place with a punch
<GenTooMan>
well at least you know what you have to adjust :D
<GenTooMan>
hmm for the holes you followed the diagram for the connector correct? It specifies 1.14mm diameter and an offset from the board edge of 1.75mm both are +-0.05mm
<azonenberg>
actually hmmm
<azonenberg>
I think the holes are actually fine
<azonenberg>
and the one i pounded might have just not been sanded far enough on the edge
<azonenberg>
this one on the other side i sanded a bit more and it went in fine
<azonenberg>
so i think its probably just edge routing tolerance that's off
<azonenberg>
Once you get it aligned it solders really well though
<GenTooMan>
just checking looks like they have a recommended via fence layout as well. Obviously you aren't doing 80GHz 5G with that. :D
<azonenberg>
I didn't use the copper layer layout because i was tuning to the oshpark stackup
<azonenberg>
not whatever they did dev on
<GenTooMan>
I've not used Oshpark maybe I should check them out might be good for stuff I want in a week instead of months.
<azonenberg>
I use them for pretty much all of my prototyping
<azonenberg>
then go to a particular chinese fab for high end stuff or larger volume