New toy - Tek TLA614 Logic Analyzer

[nctnico]

Well you could load the program in which you enter the upgrade key into a debugger and see what it does...

[TerraHertz]

Do you know if the capture memory is fixed and real or is it software disabled? It would be great if people could start figuring out how to unlock tek equipment! They wanted to charge me 10K to unlock more capture memory on my Oscope when it is already there just disabled. I hate that!

I hate that too. It's an attitude that takes the basic irrationality of belief in ownership of data to an extortionist extreme. So, Tek 'owns' the short string of data required to 'unlock' a feature, and those few bits are 'worth'$10,000?

Anyway, yes, as more of this gear is in the hands of people who bought it second hand and relatively cheaply, with no warranty or expectation of any manufacturer support, let's hope we'll see some reversing going on.
The TLA 614 options are:
Base config: 64K depth at 100MHz.
1S: 256K depth at 100MHz (mine is this)
4S: 64K depth at 200MHz
5S: 256K depth at 200MHz

I haven't yet had a close look at the capture board, but once I have this thing 'data secured' to my satisfaction, I will be.
The fact that there's a Dallas 'RAM in a box' soldered into the capture board, and I can't see any reason to make such a deeply stupid design decision other than to hold 'option enabling' data securely, suggests it may well be one board for all options.

[Edit: that there is a flash memory chip on the board near the Dallas DS1225AD makes that suggestion even stronger. What data could be unsuitable to be kept in the bare, exposed flash chip, other than very 'expensive' data? That Dallas Ram-in-a-box is soooo getting unsoldered soon. I probably have a few old ones lying around, that I can dissect to map their guts. Ideally I'd like to end up with a replacement DS1225 in a socket on the board, and with a 3V coin cell socket hot-melt glued on top, replacing its internal battery via engraving tool excavations of the potting.]

When I get to backing up that Dallas thing, I'll be paying close attention to what it contains.  Maybe, maybe... I end up with 256K @ 200MHz memory. Hope so.

In the meantime, I'm done with backing up the 6GB 'IBM Deathstar' HD from the TLA614. Took a little while, since it's been years since I bothered to do any low-level work on a HD, and all the tools I used to use are since dead. Norton Disk Ed...
So it was also an exercise in finding new tools.

With something like this, where there's a worry the critical application software _might_ be node locked to the hardware in any of several ways, it's essential to do more than just simple file backups. Safest is to capture:
* A whole disk image file. For this I used PowerQuest Drive Image 7.
* A direct file tree clone of the drive, for easy examination. For this I used ZTreeWin V2.2.19, zipped with ALZip.
  Also Overdisk http://overdisk.en.softonic.com/ is extremely useful for examining drives.
* Binary file copies of the drive's Master Boot Record (actually entire track 0), and Partion boot record. This is very useful for checking/modifying things like the HD serial number - a common node-lock reference. Also there can be keys stored in the unused track zero sectors, etc.
This last one was the difficulty, since I found I no longer had any working low level HD raw sector hex editor. Off on the freeware utilities hunt. Which always devolves into trying multiple utilities that don't quite do what you want.
In the end I used Active@ Disk Editor http://www.disk-editor.org/ which is really quite nice. Except it can't directly save selected sector data as a file on another HD. But it can copy to the clipboard, so together with a good Hex editor - ICY Hexplorer http://hexplorer.sourceforge.net/ it does the job.

[TerraHertz]

while the TLA614 case only has an opening for one of the USB connectors.

Same here on CSA7404.

It seems they used the first USB port for something, perhaps the front panel keyboard. It's jumpered to the MB NLX connector.

Yes, those NVRAMs easy to loose data on overheat...and even no overheat. Who knows how alive that battery inside it already.
But this NVRAM crap was already the only reason for my luck with CSA7404 scope Bought it "dead" from previous owner, and disregarding HDD with bad sectors and bad smart, main failure was NVRAM battery. After replacing HDD and installing fresh Win2000, all software for scope and drivers - it did not work, because
of vxWorks trickery for PowerPC board. After i finally figured that out - i got everything working flawlessly and now upgrading scope in various ways (replace Celeron with PIII, add memory, already swapped CDROM to DVDRW, now changing LEDs on front panel and fixing some minor knob issues)

Extracting & cloning the NVRAM is my next task with this thing. I'm quite nervous about this, as I can see there's a high chance of disaster. For instance the LA has option 1S, and by now I'm sure the NVRAM is where the option keys are held. I can imagine what my chances of getting a replacement 'upgrade kit' (encrypted key) from Tek would be.
But it has to be done. The Dallas DS1225AD guaranteed battery life is 10 years. The LA is over that already. So it's going to lose data, and soon.

For instance where did you get the various versions of the TLA App?

From tek website. I can upload everything i have onto my public redmine project, if u want. I have TLA714 with 7AA4 module tho, not sure if everything work directly on your 614.

After having struggled with the Tek website for a few hours now, and failing to find anything but a few scraps of random versions of the TLA application - no sensible 'here are all the old versions on one page, and which hardware models they work with' - I'd like to take you up on that offer please.

It's like Tek are being managed by monkeys these days. Oh wait, they are:
https://www.eevblog.com/forum/chat/things-don't-sound-good-at-tektronix/msg173467/?topicseen#msg173467
http://www.oregonlive.com/silicon-forest/index.ssf/2012/12/tektronix_five_years_after_sal.html
Tektronix, five years after sale to Danaher, continues to shed jobs and struggle
Five years, and five rounds of layoffs. That's the tale of Tektronix under Danaher Corp., the East Coast conglomerate that paid $2.85 billion for Oregon's signature technology company in the fall of 2007.
Danaher Corp http://www.danaher.com/
Executives http://www.danaher.com/leadership
https://www.eevblog.com/forum/chat/looks-like-tek-are-gertting-really-desperate/

Also CMOS battery on PC board just died on my LA, so will tear it apart again

At least I remembered to replace the coin cell. With the power on, and butter-fingers protection. See pic.

Something I did forget (to expect) - all my other PCs are WinXP (it's a deliberate choice.) And I *know* what a pain getting non-XP PCs to file-share on a LAN with WinXP machines always is. So now... once again the horrible battle with Microsoft's idea of how to f*ck their users over and force them to upgrade all their PCs in unison. Of which this incompatibility is just one example. Can't make Win2000 use 'simple file sharing', so all the other PCs have to do it the Win2000 way. Urrrgh! Poisonous.

I'm still hoping to switch the LA to WinXP. Not possible according to Tek, but it's looking hopeful that I can find WinXP drivers for the Intel SU810 motherboard. Leaving only the Tek TLA Application compatibility issues.
If that doesn't work, I guess I will be trying out Samba after all.

[nctnico]

Before you start digging in the NVRAM make a backup, clear the NVRAM, place it back on the board and see if the option changed. If not, the options are not in the NVRAM. AFAIK it should be possible to run WinXP on this logic analyser. Check the user forum on Tektronix' site.

[madsci]

I see an IBM Deathstar... may want to get on that

It appears to be a TravelStar 2.5" IDE disk...far more reliable than their 3.5" counterparts. With that said, age alone is a worry.

The 60GXP and 75GXP series 3.5" drives were the ones that were the baddies. They suffered from a media delamination problem. A chunk of media would flake off the disk, get stuck under the head and the drive would scour the disk bare leaving nothing but the clear glass substrate behind. Truly a sight to behold!

I kept my 60GXP that suffered this fate. I should post pics of it next time I go home from college.

[echen1024]

Huh. I would LOVE to see a photo of that

[TerraHertz]

Huh. I would LOVE to see a photo of that

Me too. I regularly strip old hard drives, mostly for the magnets. No idea how many I've pulled apart, but I've _never_ seen any with glass disk substrates.

[alm]

http://www.astro.ufl.edu/~ken/crash/

[SeanB]

You cannoy tell they are glass until you try to bend it and it shatters. Look for an old IBM drive that is between 10-20 years old and it likely will be one with glass platters.

[Seg]

Ummm yeah. IBM Travelstars are some of the most reliable hard drives the world has ever seen. Don't replace it if you don't have to. It's the Deskstars that were junk.

I have a 20gb IBM Travelstar that is way past all its SMART wear thresholds, has been knocked around plenty and dropped once or twice, has spindle bearings so worn it makes a loud whine that annoys everyone in the room, and performance has dropped quite a bit, but it has yet to lose a single bit of data. Which is more than I can say for every Maxtor and WD drive I have ever owned, every single one of which have managed to corrupt files after a mere year of gentle use.

I keep it running as an experiment at this point...

No idea if this extends to Hitachi Travelstars, but so far so good. Ask me in another decade or so...

[nctnico]

The biggest enemy of a hard drive is heat and most PC enclosures don't provide any airflow along the hard drives. In my PCs I have mounted the hard drives in 5.25" bays and milled slots in the plastic covers so there is airflow around the hard drives. In some servers I have had hard drives running for over 10 years 24/7. Turning such a machine off/on is always a white knuckle ride but I never had a problem.

In test equipment I always use compact flash cards. I don't like the whining of a hard drive right in front of me.

[Fsck]

Ummm yeah. IBM Travelstars are some of the most reliable hard drives the world has ever seen. Don't replace it if you don't have to. It's the Deskstars that were junk.

I have a 20gb IBM Travelstar that is way past all its SMART wear thresholds, has been knocked around plenty and dropped once or twice, has spindle bearings so worn it makes a loud whine that annoys everyone in the room, and performance has dropped quite a bit, but it has yet to lose a single bit of data. Which is more than I can say for every Maxtor and WD drive I have ever owned, every single one of which have managed to corrupt files after a mere year of gentle use.

I keep it running as an experiment at this point...

No idea if this extends to Hitachi Travelstars, but so far so good. Ask me in another decade or so...

if you're comparing it to other spinners, sure. if you're comparing it to some industrial rated SLC with a correctly designed board/platform? not so much.

[madsci]

http://www.astro.ufl.edu/~ken/crash/

That's EXACTLY what mine loos like inside, clear disks and dust.

[Oracle]

I found that keyboard really uncomfortable, maybe I'm wrong, anyway I'm wondering what would you do whit so many channels. Anyway nice toy.

[TerraHertz]

  the 'see through' disks. I hope I find one among my buckets of old disks (literally buckets) as I'd like to fully clean some of the platters, just to get cool-looking and possibly useful precision glass disks.

Oracle: yeah, I used a TLA 614 in a job once, and yes the keyboard is unbearable. Simply plug in a real keyboard. Or as some suggested here, and I've yet to try, use a remote desktop system.

I want to apologize for my silence on this project. I've been snowed under.
Firstly I decided to upgrade part of my work area before digging into the TLA 614, hence everything is now in a pile. Also my main computer was dismantled. I'm using a small backup PC. Then my 86 yo mum had a fall and broke two ribs. Was in hospital for a while. Then the cable modem died, so no net at all. Once that was fixed, I discovered my main PC had developed a problem, which prompts me to finally do the switch to a better PC I've been putting off for a while. So still no main PC to access net.
Also my best surviving old friend is now dying of liver cancer. So I'm not very cheerful myself.
Meanwhile I've been trying to keep track of multiple overseas large purchases via that USA reshipper, using improvised net connection and web-mail, urgh!
And not finished the building improvements, so still chaos here.

On the bright side, I bought some more toys, including another Tek LA :

Tektronix TLA 721 Logic Analyzer Mainframe Dual Monitor w/ Win 2000 Pro WORKING
Item 121152614886

HP 80000 (E2900A) 1GHz Data Generator Mainframe + E2902A & E2903A Modules
Item 300694263268

So now I have  Tek TLA 614 and TLA 721 (which takes modules), and my friend (the one dying slowly) has a TLA 715. I'm definitely going to make a serious attempt at cracking the Tek options software keying system.

I have teardown pics of the HP 8000, and will also be dismantling the plugin modules to see if the damaged SMA connectors can be replaced. Likely not... But I expected that. Already have one other one on the way, and they're common enough on ebay.
Once my workspace is functional again,  I'll post the pics. Ditto of the TLA 721.

Oh and TiN, I'll finish fetching those files once the new PC is set up. In a few days.

Pics below are of a couple of mangled module-retaining screws on the HP 8000. Can anyone suggest where I might find replacements? The broken eject levers I can hopefully find, as they have a manufacturer name and PN. (Schroff 80817 135)
With the screws, in worst case I can probably make them, from standard screws of the same thread.
But it's academic anyway, till I determine if the damaged SMAs are replaceable. It's such a shame - the modules appear to work, except most of the output physical connectors are rooted.

[echen1024]

Sorry to hear about your friend  . How much should I expect the TLA's to go for?

[TerraHertz]

Sorry to hear about your friend  .

Thanks.
I've been trying to get him to join this forum for ages, but he's just not a public-text-talkative guy.

How much should I expect the TLA's to go for?

Err... you mean mine? I don't intend to resell them, if that's what you were wondering. The whole work area rebuild is so I'll have room to use these conveniently, and some other stuff.

On ebay there are many. Searches:
  tektronix tla7 module
  tektronix tla

There's the usual huge price range. Patience...
I've been pretty lucky lately with making offers well under listed prices, even when sellers didn't say 'make an offer'.

[Hamster]

i just thru away a bunch of those discs, i was destroying them, and did not expect the platters' to be glass and shot up and hit me in the face, good thing i had my safety glasses on!

as for logic analzers, i love my HP 16700B ( 16534A (500mhz dso) x1, 16750A ( 400mhz  La) 2x, 16522 ( Pattern Generator) 1x ) , i also have a spare 16700A ..

[MadTux]

Anyone succeeded in running a Tualatin PIII-S in a TLA 721 module?

Memory seems limited to 512MB because the Intel 82815 chipset only supports that much:
download.intel.com/design/chipsets/datashts/29835102.pdf
Installing two 512MB PC133 SDRAM SODIMMs results in a non booting machine that gives BIOS beeps.
So no more memory upgrade for me 

Next thing I tried was installing a 1400MHz PIII that I had tested successfully in a DL380 G2 server before. Also resulted in a non booting machine without BIOS beeps. The machine has the latest 1.00.02 BIOS and the Intel 82815 is the SL5NQ stepping that should theoretically support the latest P3 Tualatin CPUs.

Any ideas why it doesn't work? Missing microcode for Tualatin? Locked CPUIDs? Haven't that much experience with BIOS hacking, so I'll probably just use the system as it is, before I break anything.

[TiN]

I think you will just need hardware hacking

Like this: https://www.eevblog.com/forum/testgear/tek-csa7404-repair-project/msg301679/#msg301679

There are some straps and resets rerouted on CPU package to allow III-S 133MHz work as 100MHz in old boards. At least 1400 III-S shown on photo linked working like a charm as 1050MHz in i810 board of my Tek CSA7404. You can google about these mods to get Tualatin support on non-Tualatin ready boards.

[nctnico]

Still it makes more sense to run the software remotely. A faster CPU is not likely to make the software work any faster. I have a feeling the TLA software is slowed down on purpose; it never uses the CPU for 100%.

[MadTux]

It seems like TLA-721 is limited to Coppermine Pentium III 

Tried Tualatin pin surgery described by TIN, worked better than expected, used a very sharp soldering tip made from thick 3-phase wire, a soldering iron at max heat and some isolated enamel copper wire salvaged from broken CPU fan motor:


Epoxied the wire to substrate to prevent short circuit by exposed CPU socket pins (maybe use a bit less, stupid me used far too much):


After that I cleaned those pins close to epoxy sites with sharp knife/tweezers from stray epoxy, checked it under magnifying glass and put it into socket. Because Tualatin has IHS on top of die, the CPU is a bit higher and was strapped down with gummi rings for testing:


Unfortunately all that effort didn't help, system still doesn't boot.

Now I've installed a 1000/256/133 1.7V coppermine PIII, that works flawlessly @ 133MHz FSB:





As far as I understand that pinmod:
Top left pins:
   No RESET2# pin => no reset possible from mainboard/BIOS
   No DYN_OE pin => select CPU voltage by VID pins
   VTT_PWRGD--1k--VTT => VID/BSEL at correct state
Top right pins:
   VID25mV to VSS  => lower Vcore by 25mV
Bottom right pins:
   NCHCTRL to VTT => previously reserved pins that might not be connected in Tek TLA mainboard?

Any reason why VID0, VID1, VID2, VID3 pins aren't hardwired to VSS/VCC to set core voltage to a clearly defined value? I've seen other Tualatin pinmods where some more VID pins were wired to VSS. Just wondering.

BTW, Thank you very much TIN for great homepage with firmware revisions for TLA7AA4 modules, impossible to find in the wild   

nctnico: Probably best to operate it remotely (also because of limited memory), but faster CPU with double L2 cache would be IMO still a nice thing to have, especially if it's free from trash

If anyone wants that modded Tualatin for their TDS7000 for a few €, write me a PM. My DL380 doesn't like it any more and gives a CPU fail, but I'm quite certain that it's still OK.

[nctnico]

I just bought a TLA715 (closely related to a TLA721) but I think I leave the CPU as it is.

Edit: It seems I'll be needing a BIOS update/reload... none of the hard drives I tried is recognised properly.  Somehow the LBA translation doesn't work properly.

edit2: ofcourse it is just a bad contact... the removable drive bay has 5 connectors between the motherboard and the hard drive 

[TerraHertz]

Update: My TLA 614 got killed by a lightning strike. Now repaired, and the Dallas NVRAM read and replaced.
Writeup here: http://everist.org/NobLog/20160331_lightning_luck.htm#tla614