1
\$\begingroup\$

I would like to know how do manufacturers limit the bandwidth through software options in oscilloscopes? If i have a MSO with 1GS/s, but e.g. just have 70Mhz out of possible 200Mhz (via software option), do the digital channels sample at 1GS/s or are they as well slowed down? I always read of bandwidth upgrades, but the sample rate seems not to be affected. Is this correct? And how do manufacturers limit the bandwidth? Just by not enabling the necessary horizontal scale or how does it work?

Thank you!

\$\endgroup\$

closed as off-topic by RoyC, Finbarr, Sparky256, PeterJ, laptop2d Mar 5 '18 at 17:25

This question appears to be off-topic. The users who voted to close gave this specific reason:

  • "Questions on the use of electronic devices are off-topic as this site is intended specifically for questions on electronics design." – RoyC, Finbarr, laptop2d
If this question can be reworded to fit the rules in the help center, please edit the question.

  • \$\begingroup\$ With a lowpass filter and sampling, depends on the manufacturer \$\endgroup\$ – laptop2d Mar 5 '18 at 17:24
7
\$\begingroup\$

You should realize that sample rate (1 GS/s) and Bandwidth (70 MHz) are different things!!!

They are related in that a certain sample rate dictates the maximum bandwidth of the signals which can be sampled accurately. This is set by the Nyquist frequency

The Bandwidth of the oscilloscope is most often limited in the frontend of the oscilloscope. The frontend is the input amplifier including protection circuits and range switching (which changes the voltage gain of the frontend). There might also be an anti-aliasing filter (a low-pass filter) present.

On way to make the frontend's Bandwidth changeable by software is by simply switching on/off a capacitor. This is done in the Rigol DS1054Z as shown by Dave from the EEVBlog in this video. That capacitor can simply be part of an RC Lowpass filter (the anti aliasing filter!) which sets the Bandwidth.

It is theoretically possible to (also) limit the sample rate and/or do post-processing to limit the bandwidth but this can result in aliasing effects and requires processing power. Switching a capacitor is much, much simpler. Also, that would limit the oscilloscope's bandwidth in the same way as it always has in analog oscilloscopes. You can view a 100 MHz signal on a 70 MHz oscilloscope but the 100 MHz will be attenuated. So you might measure 1 Vpp while the signal is really 1.5 Vpp for example.

\$\endgroup\$
  • \$\begingroup\$ In the high end capable units from Tek, the front end has been 63GHz capable for many years. \$\endgroup\$ – Peter Smith Feb 28 '18 at 17:15
1
\$\begingroup\$

The bandwidth is limited by a low pass filter. There is a hardware low pass filter to prevent aliasing, probably with its -3dB point at 200MHz. That low pass filter must attenuate everything above 500MHz (the Nyquist frequency) enough to prevent aliasing. The sample rate is always 1GSa/s. Then there is a software low-pass filter to limit the -3dB bandwidth to what you have purchased.

You can still see signals above 70MHz, they're just attenuated.

\$\endgroup\$
0
\$\begingroup\$

Modern Digital oscilloscopes have several fundamental blocks that process probe input data in a chain before the trace gets displayed.

  1. Front End - analog circuitry with programmable attenuation and offsets etc. Depending on gain settings, it might have somewhat different bandwidth. And "bandwidth" is a stretchable concept, the transfer function can have gradual decline, not just "-3dB cut-off", which can be corrected later in the processing chain.

  2. Sampling/ADC unit. In many cases the ADC samples the signal at constant (and rather high) rate, above the Nyquist frequency of anti-aliasing filter of the Front End. So the signal is usually oversampled. The rate of data storage however can be "decimated" in the process.

  3. Data storage (memory) for raw data. Fast memories are required to store the input stream from ADC unit.

  4. Display Unit.

Before the data are displayed, modern scopes have the signal digitally processed. So you can correct the uneven input characteristics, put interpolations, adjust scales in all directions, and run various measuring algorithms. This is all in post-processing software, to put a nice picture on LCD.

So there are many options for software/firmware to change/extend basic scope characteristics, depending on how much you are willing to pay. Most notably the "software upgrades" are used in configuration of depth of data storage. Scopes might have the super-fast memory soldered down for the maximum already, but software enables only certain portion of it. And to get full memory, you might need to purchase a special license to use it, and it might be sold on expiration basis.

Regarding bandwidth "upgrades", if the software-based "upgrade" is offered, then the scope has a full-featured front-end that meets highest advertised parameters. However, good quality analog circuitry in high-MHz area is expensive. Manufacturers of analog-to-digital components usually have their ICs binned to different grades, and prices vary substantially. It is possible that less expensive version of a scope has the binned-down front-end components, which can be upgraded only by upgrading the hardware module.

\$\endgroup\$
-2
\$\begingroup\$

There are several methods that can be used:

  1. Limit the clock of the ADC.
  2. Limit the number of samples read from the ADC per unit of time.
  3. Limit the number of reported samples.

It is really up to the developer to choose. Usually, there are other hardware-based limitations, such as filter bandwidth and ADC grade.

\$\endgroup\$
  • 1
    \$\begingroup\$ What you suggest is changing the sample rate, that does not influence the usable (for reliable measurement) bandwidth of the oscilloscope. Changing the sample rate but not limiting the bandwidth of the ADC's input signal can result in aliasing products and give unwanted artifacts on the screen. \$\endgroup\$ – Bimpelrekkie Feb 28 '18 at 14:17
  • \$\begingroup\$ Those are the options available, I did not claim they are optimal. \$\endgroup\$ – Lior Bilia Feb 28 '18 at 15:32
-2
\$\begingroup\$

Since DSO's are not all the same but can have ADC's up to 10 Gs/s to support high resolution with various memory capacities and ratios of sample rate to time resolution to support high resolution vertical traces. Obviously 2:1 ratio is min. for Nyquist fundamental capture, but signals have vertical resolution that may span N bits and thus this ratio needs to be >=10 to xxx for accurate measurements.

How do manufacturers limit the bandwidth through software

Possibly an encrypted parameter to define the attribute for max PLL multiplier value and host of other parameters associated with that.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.