<azonenberg>
Tested two more probes, also good. Definitely more ripple than I'd like around 3 GHz, need to investigate that more thoroughly
<azonenberg>
So yield is actually not looking too bad
<azonenberg>
7 of 10 good so far
Degi_ has joined #scopehal
Degi has quit [Ping timeout: 265 seconds]
Degi_ is now known as Degi
<azonenberg>
Oook so final yield: 12 good, 8 bad, 2 untested (on the NIC) right now. Of the "good" ones there's also a few that i'm not thrilled about due to weird peaks, but they meet the 6 GHz BW target
<bvernoux>
It seems Rigol Engineer are working on a new firmware for MSO5xxx
<azonenberg>
Nice
<bvernoux>
to improve SCPI ;)
<azonenberg>
I just got confirmation from the dealer that my scope is shipping monday
<LeoBodnar>
is this Eric Bogatin's piece? Eric talked to me about this some time ago, he said he uses it in his classes and wanted a bit more background about how it came to be
<bvernoux>
They shall share news soon
<azonenberg>
LeoBodnar: I don't know if he wrote that particular article but he does a lot of signal integrity educational content for LeCroy
<bvernoux>
They have clearly see the different issues with my detailed test with results on different SCPI commands
<azonenberg>
so i'm sure he had a hand in it
<azonenberg>
I'm excited to try your SMA pulse gen with my new scope when it comes in
<LeoBodnar>
he's their signal integrity tsar - if I remember right
<azonenberg>
Yeah
<bvernoux>
LeoBodnar, I have used your amazing SMA puls gen (the 2.92mm version) on my Rigol MSO5074 ;)
<azonenberg>
I wanna see if my SDA 816Zi can accurately measure the ~30ps rise time of the SMA version
<bvernoux>
The bandwidth is amazing (with unlocked scope)
<LeoBodnar>
:) cool, thanks
<azonenberg>
What's the BW of the scope you use during manufacturing test?
<azonenberg>
iirc it was a tek CSA of some sort
<bvernoux>
LeoBodnar, Your pulse gen just confirmed what I have measured with my RGSigGen (max freq -3dB)
<bvernoux>
With FallTime of 700ps & RiseTime of 690ps
<bvernoux>
So more than 550MHz BW
<bvernoux>
for a 1Keuros oscilloscope it is very nice ;)
<bvernoux>
shall be good enough to validate USB HS 480Mbps with eye diagram
<bvernoux>
But it requires special probes for that as default probe are limited to 350MHz or max 400MHz
<azonenberg>
bvernoux: have you tried hacking up 50 ohm feedthrough terminations and using transmission line probes with it yet?
<azonenberg>
The AKL-PT2 should be available for sale in the next few weeks
<bvernoux>
I shall test with my DC/1GHz adaptor
<bvernoux>
and my RF Probe to see ;)
<bvernoux>
Will do the test this evening with my RF sig Gen to check if I can recover >500MHz
<bvernoux>
I need to finish my transmission line test board also need to populate the SMA "26.5GHz"
<bvernoux>
On the OSHPark 6 layers test batch
<azonenberg>
Nice
<azonenberg>
That reminds me i need to make another test fixture for the AKL-PT2
<azonenberg>
my current ones are wearing out and i want to make a few tweaks to make them easier to solder to
<bvernoux>
yes i have seen the details on Twitter
<bvernoux>
it is not easy to have reproducible soldering/mount of SMA... especially to have flat response up to 6GHz+
<bvernoux>
the SMA connector you are using (pratically identical to mine) does not help ...
<azonenberg>
Yeah. But for something "disposable" like a solder-in probe i feel like using a high grade connector is a bad idea
<bvernoux>
So it is tricky to solder them perfectly with a good yield
<azonenberg>
i guess i could explore using a compression mount one
<bvernoux>
especially when we want signal to be very good to more than 6GHz
<azonenberg>
and yeah at this point what i want to do is understand the variation better
<bvernoux>
yes for disposable probe it is maybe too much to use such SMA
<azonenberg>
Any of these probes would work fine for passing digital data or narrowband RF, etc
<azonenberg>
The trick is super flat response
<bvernoux>
yes
<bvernoux>
A very nice things will be to build passive probe (up to 6GHz+) which can be connected to PCBite ;)
<bvernoux>
Will be a must have
<bvernoux>
Anyway I'm not sure they could reach 6GHz when not soldered to probe circuit ...
<azonenberg>
I actually already have a PCBite probe design that works well to about 3 GHz then peaks a bit. Similar problem to the AKL-PT1
<azonenberg>
needs more tweaking before it will be usable
<azonenberg>
I'd love to build an active diff probe for the system but that will be seriously challenging
<bvernoux>
it is very good if it is stable up to 3GHz just with "pogo pin" and without the need to solder it
<azonenberg>
WRT variable tip spacing
<azonenberg>
Good grounding is one of the things i am still working on. but that's the plan
<bvernoux>
yes good grounding + good center pin contact
<azonenberg>
anyway it's a work in progress, i want to get the AKL-PT2 launched and ready for orders before i spend toooo much time on other probe designs
<bvernoux>
to work fine from DC to 3GHz+
<bvernoux>
with PCBite ;)
<_whitenotifier-4>
[scopehal-apps] sfmth opened issue #309: Creating an AUR package for this repository - https://git.io/JmX5F
<azonenberg>
And before I do a PCBite differential probe I want to get a working normal one
<azonenberg>
I think what I might do is shelve the solder-in diff probe for a little while and try to make a handheld one first
<azonenberg>
that will avoid the long "tail" which has been such an EMI problem
<azonenberg>
But the mechanical engineering for that will be decidedly nontrivial
<azonenberg>
I think I might try to hack up a flex-rigid combo at oshpark for that probe
<_whitenotifier-4>
[scopehal-apps] azonenberg commented on issue #309: Creating an AUR package for this repository - https://git.io/JmXNw
<_whitenotifier-4>
[scopehal-apps] sfmth commented on issue #309: Creating an AUR package for this repository - https://git.io/Jm1Jd
<_whitenotifier-4>
[scopehal-apps] azonenberg commented on issue #309: Creating an AUR package for this repository - https://git.io/Jm1Uu
bvernoux has quit [Quit: Leaving]
Implant has joined #scopehal
Implant_ has quit [Ping timeout: 246 seconds]
electronic_eel has quit [Ping timeout: 264 seconds]