I think I may have found out why the CF isn't well recognized, this "mainframe" (I dunno how to call it to be honest) has a hard limit of 2GB (if I remember correctly) for the hard drive. Anything larger will not work well (the ROM will fail to recognize it or something like that).
@@JoshuaBoyd I've had other machines with similar limitations where it would only recognize one platter of a hard drive that was too large (eg. it would recognize ~680gb of a 2tb hard drive). I don't know exactly why it happened, but it's a good indication that this sort of issue isn't always as simple as just an easy cutoff.
The oscilloscope traces don't appear on X11 because it draws straight to the screen to speed up redraw. The confusing part is - only some models of scope card do this. The later scope cards work fine on X11.
I just came here to say that, as I found the list of supported cards in the LAN card manual. The only scope card listed is the 16532A Great video, thanks!
This was pretty common on scopes well into the 2010s. When I was working on the Tek DPO and MSO 3k/4k series back in the mid-2000s to early 2010s, those scopes all bypassed the standard UI rendering code and DMAed waveform data directly from dedicated acquisition memory to the video RAM. It made supporting screen capture and printing (one of the main areas I worked on) a huge pain because I had to do some work to transparently freeze acquisitions long enough to pull the current waveform data out of acquisition memory to a buffer, merge the UI and waveforms in software, then render the resulting image.
so the reason why oscilloscope probes can't be directly extended is that they are not low impedance low loss transmission lines, as you might expect, but are actually quite high resistance. the idea there is that because the attenuation is quite high, you can get away with being impedance mismatched to different sources, and the high-ish resistance of the probe lines will eat the reflections that happen inside the probe cable. the professionally way to extend the probes would either be with 1. a high frequency amplifier 2. active probes sadly both of these options are quite expensive my recommendation is to look online for regular passive probes that just have really long cables
I wonder if Dave over at EEVBlog knows more about this particular device? He definitely has seen a great deal of these kind of devices at least, it might be beneficial to see if he has any extra advice or tips.
according to what I just observed, as soon as the HP was used, around 2:50, the interface inside the Linux window shows _the_ _wrong_ _palette_. If that is the case, it accounts for why you can not see the trace lines.
I used some networked HP/Agilent Logic Analyzer remote displaying to Redhat Linux 9 or RHEL 3 circa 2003-2004. I don't recall the exact HP model used, it wasn't brand new, so it was probably a contemporary of the 16500C, but it definitely was lower end. If you have access to a suitably plain and older laptop or desktop, it might be worth going way back in Ubuntu or Redhat to see if that will fix the issue with display in X11. Of course, getting an old HP 9000/712 could be fun to.
I got a reply from someone that the way some of the slower earlier cards interface with the system are incompatible with X and the later ones do work properly. It could be that, so if I can get my hands on some better cards, that may solve the problem. I'm going to look around online for other people using these and try to see how they have them configured to see if I can get some confirmation on which ones will and wont work. I definitely have some older 2000's linux releases I could try for this though. That's the potentially free solution, so I will look into that as well.
you can extend BNC connections. ~1.5 dB of insertion loss per connector. You should make damn sure that the total cable length for each probe is identical and has the same number of connectors. You may have problems as you get close to the Scope card's max bandwidth and minimum input voltage. You may have issues with reflections and EMI on the probes.
On the CF card not working, in the past I've found that some systems are pickier about accepting CF as an IDE substitute than others. What I've had better luck with is using IDE DOM modules. Such a module plugs directly into the motherboard header and typically has a power cord running from it to a molex HD power connector. I think KingSpec 2GB modules were what I used to use the most back in the mid 2000s. Hopefully something like that will be easy to get in your area.
That's exactly what I'm going to be getting. I've got a 512MB DOM on the way that I should be able to just pop in and get going. I think there are a lot of little things working against this system for drive compatibility, so I'm hoping the DOM just works so I can be done.
You could do vnc but it would be pointless. It’s obvious that it’s written to the frame buffer. Really cool man thanks for sharing.. is it rack mountable?
You could probably modify the colours on the Linux machine using the .Xresources file. Been decades since I touched that, so I don't remeber the details.
Random thing, why not just image the hard drive and flash a smaller card instead? I'd use something like 1GB or 512MB Also Linux is cool, what's your favourite distro? I'm a lubuntu and zorin os guy.
Yes, I found out that this unit has a limit of 2GB or so (from what I heard in another video about this logic analyzer), I presume that it's a bug in the unit's "BIOS" or ROM.
A 512MB CF card would 100% work, there is definitely a limit somewhere around 2GB, but 512MB would probably be the safest. They should be pretty easy to find too, since small (64-512MB iirc) CF cards were standard in basically every piece of Cisco gear for a decade or two Edit: Yep, they're like $5 on eBay. Awesome
It's got that bowling alley scoreboard color pallet, classy
you are onto something there ...
@ungratefulmetalpansy I went into the replies because I KNEW that would be there.
I think I may have found out why the CF isn't well recognized, this "mainframe" (I dunno how to call it to be honest) has a hard limit of 2GB (if I remember correctly) for the hard drive. Anything larger will not work well (the ROM will fail to recognize it or something like that).
I thought that issue usually showed up as being recognized as 2GB or smaller instead of the real size, not completely refusing to work.
@@JoshuaBoyd I've had other machines with similar limitations where it would only recognize one platter of a hard drive that was too large (eg. it would recognize ~680gb of a 2tb hard drive). I don't know exactly why it happened, but it's a good indication that this sort of issue isn't always as simple as just an easy cutoff.
The oscilloscope traces don't appear on X11 because it draws straight to the screen to speed up redraw. The confusing part is - only some models of scope card do this. The later scope cards work fine on X11.
I just came here to say that, as I found the list of supported cards in the LAN card manual. The only scope card listed is the 16532A
Great video, thanks!
This was pretty common on scopes well into the 2010s. When I was working on the Tek DPO and MSO 3k/4k series back in the mid-2000s to early 2010s, those scopes all bypassed the standard UI rendering code and DMAed waveform data directly from dedicated acquisition memory to the video RAM. It made supporting screen capture and printing (one of the main areas I worked on) a huge pain because I had to do some work to transparently freeze acquisitions long enough to pull the current waveform data out of acquisition memory to a buffer, merge the UI and waveforms in software, then render the resulting image.
I love that style of hp old professional equipment, I myself have a hp 4951b protocol analyzer and it's beautiful
so the reason why oscilloscope probes can't be directly extended is that they are not low impedance low loss transmission lines, as you might expect, but are actually quite high resistance. the idea there is that because the attenuation is quite high, you can get away with being impedance mismatched to different sources, and the high-ish resistance of the probe lines will eat the reflections that happen inside the probe cable.
the professionally way to extend the probes would either be with
1. a high frequency amplifier
2. active probes
sadly both of these options are quite expensive
my recommendation is to look online for regular passive probes that just have really long cables
I wonder if Dave over at EEVBlog knows more about this particular device? He definitely has seen a great deal of these kind of devices at least, it might be beneficial to see if he has any extra advice or tips.
I like that idea, about Dave
according to what I just observed, as soon as the HP was used, around 2:50, the interface inside the Linux window shows _the_ _wrong_ _palette_. If that is the case, it accounts for why you can not see the trace lines.
I've watched 2 vids on this and still have no idea what it is. Have enjoyed them tho.
It's basically a computer for designing computer components. Like CPU's, GPU's, ect.
I wonder if Xephyr on the Linux side would work. It's basically a nested x server where you can set the bit depth and scaling and everything.
I used some networked HP/Agilent Logic Analyzer remote displaying to Redhat Linux 9 or RHEL 3 circa 2003-2004. I don't recall the exact HP model used, it wasn't brand new, so it was probably a contemporary of the 16500C, but it definitely was lower end. If you have access to a suitably plain and older laptop or desktop, it might be worth going way back in Ubuntu or Redhat to see if that will fix the issue with display in X11. Of course, getting an old HP 9000/712 could be fun to.
I got a reply from someone that the way some of the slower earlier cards interface with the system are incompatible with X and the later ones do work properly. It could be that, so if I can get my hands on some better cards, that may solve the problem. I'm going to look around online for other people using these and try to see how they have them configured to see if I can get some confirmation on which ones will and wont work.
I definitely have some older 2000's linux releases I could try for this though. That's the potentially free solution, so I will look into that as well.
you can extend BNC connections. ~1.5 dB of insertion loss per connector. You should make damn sure that the total cable length for each probe is identical and has the same number of connectors.
You may have problems as you get close to the Scope card's max bandwidth and minimum input voltage.
You may have issues with reflections and EMI on the probes.
Small footnote: If you need to scale X11 windows, use xpra!
you could try using renderdoc to troubleshoot the x rendering issues
On the CF card not working, in the past I've found that some systems are pickier about accepting CF as an IDE substitute than others. What I've had better luck with is using IDE DOM modules. Such a module plugs directly into the motherboard header and typically has a power cord running from it to a molex HD power connector. I think KingSpec 2GB modules were what I used to use the most back in the mid 2000s. Hopefully something like that will be easy to get in your area.
That's exactly what I'm going to be getting. I've got a 512MB DOM on the way that I should be able to just pop in and get going. I think there are a lot of little things working against this system for drive compatibility, so I'm hoping the DOM just works so I can be done.
To make an obscure Japanese game reference, the interface gives me Overblood ZEUS OS vibes
"Disengaging door lock."
Really nice piece of electronic.
keep them coming
Is there a version of this logic analyzer mainframe without the screen built in? Just a box that you can connect an external screen?
I believe there was an extender, but it was just extra card slots and a power supply. It still had to slave to the mainframe.
I love a good sequel.
Man, I want me a logic analyzer that has X built in...
Holy shit !!! Are you X11 forwarding?
You could do vnc but it would be pointless. It’s obvious that it’s written to the frame buffer. Really cool man thanks for sharing.. is it rack mountable?
Great video, as always.
You could probably modify the colours on the Linux machine using the .Xresources file. Been decades since I touched that, so I don't remeber the details.
You run kubuntu! I knew there was a reason I like your videos... 🤔😉
About the pods being removable, couldn't you just make it so they could be internally disconnected with pin headers and sockets?
Can it run doom?
That is quite a bit more advanced than my ML-4100! (But my ML-4100 still got the job done on my Tempest boards.)
it is amazing how backwards compatible X windows still is.
You could try running an older Unix/Linux in VirtualBox or such
Isn't the UI unusably slow, especially on the logic analyzer display ?
You'd think we would have gotten Dark Mode working after thirty years..
But does it run Cry... I mean Battle Zone?
Maybe it could run DOOM with a bit of hackery.
If only we had PEBCAK analyzers ;)
I'd invest in that! lol
Take my upvote sir, sign me up for a dozen or so of these.
When i started my new job my boss gave me one of these from his stack of 5 of them.
The osciloscope is superimposed in the graphic logic.
Random thing, why not just image the hard drive and flash a smaller card instead?
I'd use something like 1GB or 512MB
Also Linux is cool, what's your favourite distro? I'm a lubuntu and zorin os guy.
Yes, I found out that this unit has a limit of 2GB or so (from what I heard in another video about this logic analyzer), I presume that it's a bug in the unit's "BIOS" or ROM.
#LinuxMintMasterRace
A 512MB CF card would 100% work, there is definitely a limit somewhere around 2GB, but 512MB would probably be the safest.
They should be pretty easy to find too, since small (64-512MB iirc) CF cards were standard in basically every piece of Cisco gear for a decade or two
Edit: Yep, they're like $5 on eBay. Awesome
now p10k zsh? oh my .....
try on something that doesnt have a High DPI display
waw
HP computer
Perhaps a period of orientation would've been helpful before posting this.
such a waste.
First! Also great video!