...or, Amazon's, as may be...
Having determined that the plumbing expenses Shall Be Paid For Out Of The Household Emergency Reserves and not out of Christmas money, I'm back to pondering capital equipment wherewith to celebrate the new year productively.
A few years ago, I started seeing Mixed-Signal Oscilloscopes - digital scopes with a modest number of logic-analyzer channels built in - in the Tektronix booth at trade shows. Being as how a lot of my work is of a mixed-signal character, I've naturally been drooling, but of course the price of a serious MSO remains more than I can justify, and I'm not sure I want to fool around with one of the no-name USB toys.
Well, Rigol has been turning out some nice-looking little scopes with decent specs, and I've been hearing good things about them, and... well, looky here. While the higher-range units are big bucks, the entry-level 1000 series looks plausible, and even the top of that line could fit my budget.
So, take inventory. Discounting things in storage, such as the old storage-tube scope, I have two useful scopes: a 2-channel, 100 MHz Tek portable, with a monochrome screen and the option module for RS-232 and CF card, and a 4-channel, 1 GHz (sort of) LeCroy, with an orange CRT, RS-232 and HP-IB interfaces, a floppy drive (that doesn't work; I need to replace it again) and some sort of PCMCIA slot.
Hm. A portable of decent capability, with some digital channels, a color display, and modern interfaces would plug a gap indeed.
So I look at the MSO1xx4Z line. In practical terms, 70 MHz analog bandwidth is probably good enough for a general-purpose scope; if I'm trying to spot distortion in high-speed signals, overshoot on clock signals, and suchlike, I can always resort to the bench scope. But... the 100 MHz model isn't that much more. Downside: allow 4 to 6 weeks for delivery.
Check Amazon. Oho! Someone has the MSO1104Z-S (includes built-in 25 MHz signal generator) in stock, for delivery by the end of the year! So I can record the date placed in service as being the same year the money was spent (and I even have a legitimate immediate use for it, so there).
So I guess that's the major item. Then there's some Digi-Key shopping to be done, too: latest Atmel debugging pod, for example; I'll probably be needing to debug an ATmega168 application (that uses the serial port for application purposes) long about February or March, and I still can't find my AVR Dragon board, besides which I'll be wanting to program ATtiny9 parts (for which I'll be also needing a SOT-23-6 ZIF-socket-on-a-board, also from Atmel).
I'll put off getting a better Kinetis debugging pod (as compared to just using the OpenSDA interface on a FRDM board) until there's a more concrete requirement.
So here we go... aw, expedited shipping to get the scope this year costs a bunch extra; I'll settle for Dec 31 - Jan 5. Put it on the business credit card. Done! Now for the small fiddly stuff from Digi-Key.
Oh, and the carrying bags for the 1000-series scopes don't appear to be in stock anywhere, and Rigol shows a very long delivery time. Maybe OK; I don't expect to need it at any client sites real soon. I need to swing by a mil-surp store sometime and look at tool bags anyway - been carrying my field-service soldering iron around in a cardboard box for far too long. Maybe I'll find something scope-sized, with suitable pockets for probes and whatnot. If not, I can always order the official bag later.
Update: Looks like I may be needing that Kinetis debugging pod after all, depending how some upcoming hacking goes (when time allows). Turns out that you can't just use a FRDM-KL27Z's OCD to debug a board with a KL27Z256 chip... unless there's something in the project settings causing the error message I'm getting (and it's not the target device setting in the debug configuration; I can change that, and it still happily talks to the on-board KL27Z64 (with the jumper re-installed, of course) while giving the mismatch error when connected to my new target board. It's complaining of a mismatch, so I'm assuming it communicates OK... which may not be a valid assumption. Could be that the FRDM board needs a firmware change to make it more general.
Nothing's simple... in this case, probably due to third-party licensing issues.
Update 2: Yeah, third-party license issues. There are open-source firmware packages for the FRDM's debug MCU, but I'm not immediately seeing a convenient way to get one of those nicely integrated with KDS. So, maybe I have to buy the real pod. Which puts somewhat of a crimp in my product-line concept; I want to use the KL27Z, because of the handy ROM bootloader to allow easy field firmware changes, but I need (or at least may need, someday, for future requirements) the top-of-the-line part. I figure some users will want to develop their own firmware, and maybe they'll want to do low-level debugging. Oh, well: maybe if they want better-than-Arduino debugging capability, they'll have to spring for the couple hundred bucks for a real debugging pod, or work out a way of using the FRDM board with alternative firmware. And I guess I should, regardless, include a 3-pin header with UART0 brought out as a debug console, for good ol' printf debugging. (And consider using the 10-pin, 0.05"-pitch header, instead of a 6-pin Tag-Connect footprint.)
And why doesn't Freescale have a cheap (under $50, or hand-out-at-trade-shows-cheap) pod that handles all the Kinetis parts? Seems like a good use of the promotional budget, right? (Well, I guess that's in NXP's hands now; wonder if KDS and KSDK will survive, and whether they'll be expanded to include the LPC1xxx line.)
Update 3: It appears that USBDM is the firmware/software solution for turning a FRDM board into a multi-purpose debugger/programmer. Unfortunately, I can't get it to run on my workstation - something to do with my not yet having updated Debian to the current version, so a lot of the libraries are out of date.
I should probably just buy the dang pod, and maybe fiddle with alternative debugging options later.
Update 4: Crud. Come the Monday, Amazon tells me my MSO order has been canceled because Out Of Stock. May end up ordering from the factory and waiting until February to get my hands on it.
Update 5: Another Amazon vendor shows the MSO I want being in stock. With a not-too-hefty expedite charge, I should have it next week. Meanwhile, I can be working on the current live project for a client, and maybe fiddling with the Kinetis debugger matter.
Update 6: Fiddled with the Kinetis debugging situation some more. Installed the CMSIS-DAP OpenSDA firmware. Found the instructions for replacing two broken plugins to make OpenOCD debugging work, for sufficiently oh crap, it frotzed the security fuse settings and now the OpenOCD can't talk to my chip anymore values of "working".
Gave up and ordered the dang P&E Micro (/Freescale/NXP) MULTILINK pod. It's legal for real development purposes. (So's the CMSIS-DAP thing, but it suddenly becomes useless if you get the fuses wrong (to include not setting them at all), and maybe the commercial pod will allow a mass erase of a secured chip...?)
Oh, and I also did a few hours of real work today (being Monday). Imagine that!
Recent Comments