Product : FAQs


FAQs
DIGITAL STORAGE OSCILLOSCOPES


What do those frequency specs really mean?

There are several different measurements that may be quoted as the oscilloscope bandwidth or speed. It's important to understand each type and know which is quoted.
  • ANALOG FRONT END BANDWIDTH. The fastest signal that can be seen by the scope before conversion to a digital signal.
  • DIGITAL BANDWIDTH. This is a frequency limit based on the sampling rate of the A/D converter. The actual limit is 1/8 to 1/10 the sampling rate for linear interpolation and up to 1/4 for Sin(X)/X interpolation. Signals faster than this limit will lose amplitude and detail information. Note that sample rates drop as sweep time (time/division) increases and long memory lengths help maintain high sample rates.
  • MAXIMUM SAMPLE RATE. This is simply the maximum clock rate for the A/D converter. It determines the limit of the digital bandwidth as described above.
  • REPETITIVE, EQUIVALENT TIME, RANDOM REPETITIVE SAMPLING. This is an effective sampling rate resulting from building up a waveform in several successive acquisitions. Unfortunately, this only works with waveforms that repeat exactly the same for 10 to 100 sequential acquisitions. This therefore is not useful for events that occur only once or infrequently like glitches, dropouts, or noise.
Why does sampling rate matter?

The sample rate determines the digital bandwidth of the scope. High sample rates must be maintained to provide a good representation of the input waveform. Signals faster than the digital bandwidth limit will lose amplitude and detail information. Note that sample rates drop as sweep time (time/division) increases. The same amount of memory must store longer periods of time, meaning fewer samples can be taken for a given time period. Only longer memory lengths can overcome this limitation and maintain higher sample rates.
What's the difference between single shot Vs repetitive sampling?

The two terms describe the technique used to digitize the input waveform.
  • SINGLE SHOT captures the waveform in one pass, capturing events that happen once or only infrequently. This includes most signals like glitches, dropouts, noise and digital signal streams that are not by nature repetitive. A high single shot sampling rate is important for these types of signals, and the repetitive sampling rate is useless.
  • REPETITIVE SAMPLING builds the waveform in multiple passes which are synchronized to a repeating trigger event. Because the event must reoccur for each pass, only repeating waveforms can be sampled with this technique. You must be careful when using this technique because you can easily create imaginary or garbage waveforms.
Why can't a 1 kHz scope capture my 15 MHz signal?

The speed rating of many scopes is the repetitive mode limit. However the single shot limit can be much lower, such as 10 MHz. This prevents the acquisition of signals faster 2.5 MHz. Your evaluation of digital oscilloscopes must incorporate each instrument's analog bandwidth, digital bandwidth, and single shot limit, in addition to its repetitive mode limit.
How fast a scope do I need?

This depends on the characteristics of the signal you are trying to view. How fast is the signal? For digital, You need a scope that samples from 4 to 10 times faster than the fastest clock rate or glitch that you want to see. For analog, the scope needs to be faster than 1.4 / (risetime).

Rise Time
Sample Rate
1ns 1.4GHz
10ns 140MHz
1µs 1.4MHz
1ms 1.4KHz

Is the signal single shot or repetitive in nature? If the signal is repetitive, then use the scope's repetitive sampling rate to judge. Or, if the signal is single shot, like most glitches, noise, dropouts, etc., then you have to look at the scope's single shot sampling rate, which is many times lower than the repetitive. For example: on the HP54503A which has a repetitive bandwidth of 500 MHz, the single shot sampling rate is only 20 MHz yielding bandwidth of only 2 MHz!

Read specs carefully because they often hide the real numbers. And make sure you understand the specs and know which are important to your application.
Can your software perform math functions on the acquired waveforms?

Yes, unlike many stand-alone scopes, where math functions are either not available or extra-cost options, our scope software includes many math functions standard. These include: addition, subtraction, multiplication, averaging and inversion.

No instrument can have all of the functions possible, but with ours, the full power of the PC is available. You can save your captures in standard file formats readable by PC math packages and thus perform any further analysis desired.
A time saver is automatic waveform measurements in real time on live waveforms. What measurements does your scope provide?
I would like to use my scope in a production test and debug environment. Can your scope perform GO/NO-GO testing based on waveform measurements?

We have an extensive list of parameter measurements available, and in addition to displaying the measurements, pass/fail testing can be performed on the results. Measurements include: area under the waveform, cycles, voltage and time measurements (relative and absolute), delay time, duty cycle, falltime, risetime, frequency, mean, average, median, max, min, peak to peak, period, rms voltage, standard of deviation, and pulse width. You can set pass and fail limits for each measurement, and, on pass or fail, the scope can stop acquisitions, store the capture to a file, beep to alert you, or a combination of these three.
Is this just an analog-to-digital converter board with fancy scope software?

No. Unlike most other PC based "scope" boards, this is not based on a simple A/D board. Other boards leave out input signal controls that real oscilloscopes have. Our product has input protection and conditioning circuitry like that found on conventional stand alone DSO's:
  • Input coupling selectable between DC, AC, and GND
  • Gain stages that range from 10 mV to 2 V per division (with 10 divisions) at the 1X probe setting, or 100 mV to 20V per division at the 10X probe setting
  • Settable DC offset for the input signal range of +/- 4.96V
  • Inputs are rated at +/-50 VDC continuous and +/- 150V (DC + AC < 10KHz) transient
Does the speed of my PC make a difference in acquisition speed?

Our instruments sample independently from your PC. The PC is used as an display interface and all the high speed sampling is done on our instrument. A faster PC will draw the screen faster, but it won't affect sampling rate. If you PC can adequately run the Operating System it should run out products well also.
How can a PC based instrument sample so fast?

Our instruments have high speed sampling and buffer circuitry that allow them to sample and record at high speed and then to transfer it to the PC after the buffers are full.
Why do I need a long memory buffer?

With longer memory, you can both maintain high sampling rates, giving you better waveforms, and see longer duration events.

Memory is used to store the input waveform. Consequently, the longer the memory, the longer the period of time you can acquire. Or, more samples can be taken within the same period of time at a faster rate.

Often, the known error occurs some time after the actual cause. Longer memory allows you to trigger on the known error, but also capture the sequence of events back to the root of the problem. A long buffer allows you to see both the problem and its cause.

With a given memory size there is always a tradeoff between sample rate and the max time that can be acquired. Longer memory minimizes this problem. Without a large memory you can't use fast sample rates with real world events.

Event Duration
Memory Size
Sampling Rate
1µs 1k 1GHz
1µs 64k 1GHz
10µs 1k 100MHz
10µs 64k 1GHz
(10x faster than 1k)
1ms 1k 100KHz
1ms 64k 5MHz
(5x faster than 1k)



FAQs
LOGIC ANALYZERS


When do I use timing or statelist modes?

Timing mode is used to make real time measurements of your signals, independent of your circuit's own clock. Usually, sampling is done at a rate much higher than your clock to measure details like noise, glitches and crosstalk.

Statelist mode is synchronized to your circuit's clock to record the sequence of events in your circuit. Typically, you are trying to verify the correct sequence or are trying to figure out the actual sequence. Often this is done when there is software control of the hardware, or external data streams are fed into your circuit. The display is a list of decoded values in hex, ASCII, binary, decimal, or user defined.
With my last logic analyzer, if I captured in timing mode I could not view it in statelist mode, or vise versa.

Our product does not suffer from that problem. After you capture you can view it in timing, statelist, or both at the same time.
How do I connect your  Logic Analyzer  to my circuit?

With the provided hardware accessories, you connect via:
  • Individual double hook test clips to chip pins
  • Individual header wires directly onto header pins and strips
  • Header pins on the Logic Analyzer pods to your ribbon cables
Can I group my signals together and view them as bus structures?

Yes, in both timing and statelist modes, you can assign channels to groups decoded into binary, decimal, hexadecimal, ASCII, or user defined values. This is especially useful with data and address busses that should be viewed as one multi bit entity instead of as independent signals. Each group can have a user defined label to make the display easier to understand.
Why do I need a long memory buffer?

With longer memory, you can both maintain high sampling rates, giving you better waveforms, and see longer duration events in their entirety. Memory is used to store the input waveform. Consequently, the longer the memory, the longer period of time that can be acquired. Or, more samples can be taken within the same period of time at a faster rate. Often, the known error occurs some time after the actual cause.

Longer memory allows you to trigger on the known error, but also capture the sequence of events back to the root of the problem. It is a long buffer that allows you to see both the problem and its cause.

With a given memory size there is always a tradeoff between sample rate and the max time that can be acquired. Longer memory minimizes this problem.
Does the speed of my PC make a difference in acquisition speed?

Our instruments sample independently from your PC. The PC is used as an display interface and all the high speed sampling is done on our instrument. A faster PC will draw the screen faster, but it won't affect sampling rate. If you PC can adequately run the Operating System it should run out products well also.
How can a PC based instrument sample so fast?

Our instruments have high speed sampling and buffer circuitry that allow them to sample and record at high speed and then to transfer it to the PC after the buffers are full.