I have cought a 16-bit picoscope, I believe those were cancelled later. The cool thing is I can do FFT with good dynamic range and this has been very usefull a few times. Where does the noise come from? Fire up the FFT and you clearly see where it sits at.
Also check poScope, this is incredibly cheap and does some usefull stuff as well, such as serial protocol analysis (I2C etc). Also a logic analyzer, signal generator etc. I'm just not sure how fast as this is probably AVR based.
If you ask me I'd get a nice analog standalone scope, 200-300MHz and a PC based for those rare protocol debugging issues. Sometimes a brain fart prevents you from seeing a very obvious bug and they help save long hours.
That's good to know, Tom. I recently purchased one of those to supplement my ol' reliable Tek 2235 for my hobbyist work, but haven't had a chance to really shake it out yet.
From the little bit I've played with it, it looks pretty good, although it will never totally replace my Tek. I now need to figure how to best integrate it into the sort of work I do.
The logic analyzer adds new capability to my bench, which pleases me immensely. I'm hoping that it will make my micro experiments easier and my time more productive.
Without which many newbies would never learn the basic rules for Usenet use. Your addition did nothing to improve the newsgroup and was totally useless.
--
[mail]: Chuck F (cbfalconer at maineline dot net)
[page]:
Try the download section.
** Posted from http://www.teranews.com **
You'll find those kids on Google Groups, you know? The main benefit of delivering discussions via NNTP rather than through a "web board" interface is the much greater efficiency (far fewer bytes transferred per message -- so much snappier, regardless of the connection type) and the significant feature of having filtering performed on the client side using however fancy of a filter you'd like to use rather than on the host side using what's almost always a very, very limited set of options (e.g., just try getting Google Groups to not display posts from a particular other NNTP server!).
Most "hard core" development still takes place on straight mailing lists of Yahoo! groups anyway -- the Linux guys alone have thousands of them.
What's involved in building a simple 50 MHz oscilloscope? I have a spare Xilinx Spartan DSP board, and an unused LTC 105 MSps A/D chip in the part box. The board's 125 MHz clock is different enough from the sampling rate, I think, to jitter the trigger point for meaningful averaging. It seems almost too simple. Aside from a low noise amp, what am I missing? The board has a VGA port, 128 MB 5ns DDR2, and tons of I/O.
That certainly builds a decent simple 'scope (and the amp doesn't even have to bet that low noise!). To compare with contemporary off-the-shelf offerings you need...
-- A large amount of software to create a nice GUI, control the time base, make measurements, do FFTs, display cursors, etc. Overall implementing the software will probably take longer than implementing the hardware.
-- Decent triggering abilities. "Edge" and "level" are easy, but most scopes today also do video triggering, serial protocol triggering, missed/runt pulse event triggering, etc. Additionally, getting the display data to precisely line up from one trigger to the next requires the ability to reset your clocking system on a trigger event or some fancy software and a high-speed timer to figure out when the "real" trigger event happened and shift the ADC's data (in time) accordingly. Without this, you'll always have (upwards of) one sample time worth of jitter on your signals -- many dirt cheap digital 'scopes do suffer from this problem.
-- Some "smart" interface to the probes, if desired. Thees days all Tek and Agilent scopes have a little serial protocol used to interrogate the "fancier" probes so that the probe can tell them what scale and units to use (e.g., "This is a current probe, it produces 100mV/Amp.")
-- Connectivity to the outside world via USB, Ethernet, or (very old school) RS-232 to save/print traces, data, etc.
A "simple" oscilloscope is easy, real easy, as you suspect. Buffer/amp
ADC + memory + basic sample control + PC interface + a few lines of code to display samples.
Making it more useful on the other hand is much more work. You need wide range input attenuators and amps plus triggering and protection stuff. Lots of switching control stuff involved too. The control side gets more complicated with pre/post triggering, buffering and other niceties.
A decent DSO project involves a lot of software, and this is where many of the cheap Chinese PC based DSO's fall down.
BTW, 50MHz analog bandwidth is going to need a lot more than 105MS/s to be useful in single shot mode.
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.