Imagine a box with a USB interface, that is full of DACs, a mux'd high-quality DVM, a bunch of logic i/os, I2C and SPI, some SSRs, a few programmable power supplies, all easy to control from various programming languages.
An internal FPGA would be programmable through the USB port. The standard config would do the standard i/o, but it would be user reprogrammable to do special stuff, with the VHDL source provided as a starting point.
All the i/os would be hardened against electrical damage, and the box would have some sort of provisions for self-test.
Connectors would be suitable for cables or a baby-board test fixture, which might include special circuits.
This would be a universal test set for production electronics. It should cost something in the $1K to $3K range.
I'd buy several of these if it existed.
--
John Larkin Highland Technology, Inc
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
Sort of what you get when a Raspberry Pi slips into a nearby phone booth to change identity. ;)
You can get all that by putting a bunch of National Instruments boxes together with duct tape, but not for $1k!
I'm just finishing up that boxcar lock-in gizmo, and decided to put in an ADS8331 16-bit ADC. Seems like a very nice part, but we'll see.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
Maybe even USB powered, possibly at reduced capability? Add a wall-wart for full power. (Or a super hydra USB cable to get it 2.5 watts at a time...)
It needs some blinky LEDs too.
Advanced mode: device has a slot for an SD card or USB flash stick/ hard drive and a battery holder. Plug it into a PC, set it up, unplug it, and then let it run by itself, logging data. (Or maybe put the setup file on the SD card or USB stick and have the gadget read it at boot.)
Maybe ship a few different FPGA binaries with the device, and make it easy to trade FPGA binaries on the web. Some people want a little customization but don't have the kilodollar VHDL tools.
Maybe have the I/O buffers/clamps/whatever in sockets or on a little daughterboard or something, so they are easy to change when somebody inevitably blows them up.
In super perfect world, you'd be able to loop things around internally to do this. A loopback cable that plugs into the output would be simpler.
You might also have a calibration cable/jig that allows the user to easily supply their own externally-defined voltages and signals to the device. This could be something as simple as an AA or CR2032 holder that the user could measure with their Fluke at the same time the test gadget is measuring it.
0.1" spacing pin headers. The cables are available worldwide. It is probably a little late to use a 34-pin one to take advantage of disused floppy cables, but a 40-pin one would breathe life into all those old IDE cables sitting in boxes. Really high-frequency stuff would need something different, but I don't think you're talking about GHz here.
You might have to include some kind of "calibration lock" or "verify calibration against yesterday" feature in the software, to make it acceptable to the QA department.
I can think of a few boards that each do some of this stuff, but not one board that does all of it. A Saleae Logic is very much in this direction, but I think all of their models are input only at the moment.
What I had in mind was a sort of universal fixture for production testing of electronic gadgets. It wouldn't be portable and would potentially use a lot of power.
Oh, it needs some thermocouple channels, too.
A lot of my customers insist on all that cal and traceability and SPC sort of stuff.
--
John Larkin Highland Technology Inc
www.highlandtechnology.com jlarkin at highlandtechnology dot com
Precision electronic instrumentation
Sure, but if it's really handy, the field techs might want one too. Maybe there is a CPU board with some I/O, and then an I/O board with all the higher-power stuff. The "test stand" version has both boards in a tabletop box, while the "field" version has just the CPU board in a smaller box.
I am always happy to provide ground-breaking ideas. :)
Seriously, I have been the guy that gets to go out to production and figure out why they can't build or test something before. A non-trivial amount of the time, it is due to power-over-air or signal-over-air problems. Having a few blinky LEDs that indicate power, signals going out, signals coming in, can be handy.
Understood. I was just thinking that there would probably be a small CPU in the gadget, and if you've got that, you're probably not too far away from being able to have local mass storage on the gadget.
You might even be able to sell that as a QA thing: one can get the golden USB sticks from QA in the morning, plug them into the test gadget, and be assured that it is running the right test program, as opposed to the production bench PC that is exposed to 3v1l h@x0rs On The Internets.
Sure. For a lot of people the built-in test will be enough, but some need traceability to some external reference.
Not a bad choice, either - widely available and relatively easy to work with. Use a D-sub on a ribbon cable if you need all the signals, or a solder-cup D-sub if you just need a few. (You do probably want to make sure the pins that are relevant for RS-232 can withstand RS-232 voltages, against the day that someone plus a live serial cable into it.)
With pin headers you can do tricks like install long-tailed sockets and stack up the daughterboards, PC-104 style. To do the same thing with D-subs, you sort of have to "daisy chain" them end to end.
I was thinking more in the direction of a checksum of the setup file that is recalculated at every power-up, or something like that.
Sure and sometimes you need to run it far away from any power. Good battery packs (and low power design) could be really handy. In fact maybe a (semi-)modular design, plug together the special modules (perhaps self made; open interface) you need for the offhand task.
Well at least to set them up. After that being a bit more portable can have huge advantages.
a
Sure. Not all users would need to have certified calibration, but providing for that is a real nice touch.
Hmmmph. You haven't seen the gorilla techs i have. Hell, i have even seen some gorilla "engineers".
We keep designing custom test fixtures; just got out of a design review for another one. Sometimes the test fixture is as much work as the product.
We could design this, but it would take a bunch of work, especially on the software side, and then it would need to be marketed and supported. Single-unit sales imply lots of customers, hence lots of support.
I think it curious that nothing quite like this exists.
--
John Larkin Highland Technology, Inc
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
[about a general-purpose test/measure/switch as a USB peripheral]
Surely you know of laboratory cards/educational boxes with similar characteristics! Labjack comes to mind.
Mainly, though, the problem is: if you can't visually trace the automated wiring, how do you know why your frammis just fried? (if you CAN trace it, it isn't all tucked inside the box)
Secondary concern: the connectors are all wrong for my frammis!
Tertitary concern: who does the calibration once a year, and how?
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.