20080311

oscilloscope/oszilloskop and DSO

Separate workstation with it's own CRT/LCD, one which is PC based or build-it-yourself PC based? Does it need to handle the task of logic analyses to any degree (hence, data logging and advanced trigger options)? Should it be a brand easier to resell (picoscope, tetronix)?

  • ebay OSD standalone €444 (new, welec W2022A, 200Mhz, color LCD, 16kb mem)
  • ebay scope standalone €400? (used, tektronix 2430A, 40Mhz, CRT)
  • ebay scope standalone €100? (used, hp 1740A, 100Mhz?, CRT)
  • Picoscope 3100 range PC scope with greater input range (€580 50Mhz, €870 100Mhz, €1160 200Mhz)
  • Picoscope 2200 pocketable PC scope with advanced triggers 25Mhz €430 (5Mhz with standard triggers is €230)
  • Picoscope 2100 single line handheld PC scope (€300 for 25Mhz, €200 for 10Mhz)
    (fyi the picoscopes can also be purchased at Conrad.de for similar prices)
  • Testec 75Mhz single BNC line PC scope €360
  • Voltcraft 40Mhz €200 (really only 0 to 30 +V input? not +/-? 60Mhz is €360, 100Mhz is €550)
  • Bitscope PC scopes and logic analyzers (LA), pseudo open source and combine both analog and the logic lines into one view around €400 for 100Mhz. The LA requires another 70 for the probes and supports 3.3V/5V CMOS & TTL, buffered inputs. It does not appear to give packet analysis for the number of busses that the $1.5k USBee DX would (I2C, SPI, Async, USB, CAN, 1-Wire, Sync Serial, I2S, and PS2) or even the $150 Logic (I2C, Serial, SPI), but its enough to build your own.
  • Owon PDS6062 is a stand alone 60Mhz DSO for 350.

Do-it-yourself scopes are interesting and I've seen some based on FPGA's, sound cards and game ports. The latter two have extreme input limitations (see the wikipedia article on oscilloscopes). I've also seen one based on arduino here ("int val_read = ((analogRead(0)*5) / 1024.0) you can see the signal as you want") and even better directions for using arduino with the Liberlab python analog display software (comprehend?). The ADC pins on the Atmega128 used on the Arduino board take +0-VCC DC, which is 5V. Meaning +0 to 5VDC. The sample rate is around 10,000 per sec at an accuracy of 4.9mV (hence, no ECG monitoring of the heart pulse which is around 0.1mV to 1.5mV). However, these are averages the Arduino community report and in truth it appears with an external clock and VCC one could get much higher rates. I'm curious at what frequency the standard Aruduino configuration can give. Regardless, FPGA based seem the only robust option but not sure i want to dive into dealing with software that isnt widely used or standardized yet.

Update: arduino v11 lets one control the adc input reference voltage either switching it to a 1.1V internal reference of the atmega chip or supplying an external reference with the aref pin on the board. see analogReference()

The PC based scopes have the advantage that some initial logic analyses can be done. However, because they are all Windows based, it will require a restart or virtual machine that a standalone wouldn't need. Also, those that do not have external power and rely on USB power can run into ground issues. But considering my living situation is typically flexible carrying a standalone cross-continent is a pain. Plus who doesnt want an oscilloscope on them at all times. So for me the question now is how much portability do i need, and what are the trade-off's. I would like to sample up to 10Mhz so a 20Mhz scope should be my lower threshold. I'm split on having an isolated ground or having the flexibility of powering of USB. would love one of the Picoscope 2100's or Testec that are truly portable. But having the reliability of the bitscope 100U is also desired. If i had endless cash id get both. Instead I will probably purchase a cheap old standalone used on ebay or go for the Bitscope 100U. The latter having the advantage of some Linux support and having logic analyzer and scope inline.

Update: someone mentioned the Simple Software Radio Peripheral as another possible option and completely open source. More versatile and meant for much more than just a scope (it is related to the GNU Radio project). That said, I havent found software for visualizing the wave as you would get with a standard PC scope, yet. Perhaps Liberlab (mentioned above) can be used? This route requires the LTC1264 ADC module and maybe the USB2.0 board which is bus compatible (or could I use the FTDI standard RS232 to USB boards/chips?). Though, im not sure the modules can be purchased. If not I would need to have the PCB's made.

3 comments:

Unknown said...

Hi, Nice write up. I'm seriously considering the GNU Radio Hardware as a flexible solution for a oscilloscope. Did you ever make any progress down this path? Kind regards. Nathan

cyphunk said...

well now that you mention it the SSRP gnu radio solution has gone up in value imo. its used for other other hardware projects such as the open GSM hacking and even the recent tempest proof of concept attacks. so, perhaps it is worth it. since writing that post i got 1 click away from purchasing a bitscope and backed off because the sampling rate is pretty bad there has been little progress made on the device in over a year.

are you going the GNU route? do you have other documentation on setting it up? you have my curiosity

cyphunk said...

its, have you considered the USRP? its not open, but is actually the device people are using for other projects i mentioned (GSM, tempest).