AnsweredAssumed Answered

how to measure time interval using 53220A series counter

Question asked by tamourjaved on Apr 12, 2012
Latest reply on Jan 6, 2014 by jdwindrvr

Like • Show 0 Likes0
Comment • 12

Hello,
   I have a 53220A counter/timer, I want to measure the time interval between two channels, the way I want it to work is that the the channel1 should wait for a particular trigger than start the timer and when a particular trigger is reached at channel 2 the timer stops and also the counter stops i.e I need only one time interval measurement and not the continuous.

   Now I am trying to read the manual and follow the instructions but its my first time using these devices so I am facing some difficulties. In particular I am having difficulties how to setup triggering and gating options, I want my channel 1 to start triggering and channel 2 to stop the measurement cycle. There are several triggering options, like internal external and manual.

    So can any one help me with setting up my measurement cycle?

Thanks,

Edited by: tamourjaved on Apr 12, 2012 8:42 AM

No one else has this question

Outcomes

12 Replies

lhornburg Apr 13, 2012 2:15 PM
Mark Correct
Correct Answer
When making time interval measurements, the "triggers" on the input start and stop channels are input signal levels of a given polarity (rising edge or falling edge). The diagram below shows the concept of a two channel (1-2) time interval measurement.

Below is the set up for the measurement:

Preset
Function: Time Interval 1-2
Channels 1 and 2:
   Auto Level: Off
   Start Channel Level: 500 mV   (Ch. 1)
   Stop Channel Level: 500 mV    (Ch.2)
   (Start and Stop) Slope: Pos
   Coupling: AC
   Impedance: 50
   Range: 5V

You may need to experiment with the start and stop levels etc. for your particular signal. As you need to make only a one time measurement you can set the trigger source and readings per trigger as follows:

Trigger Source: Manual
Rdgs/Trigger: 1

Each time the trigger key is pressed a single time interval measurement is taken when the start and stop edges and levels are received. The counter’s gate time is not used as the start edge starts the measurement which continues until the stop edge is received.

One final item to consider is the measurement time out setting (Utility   Instr Setup   Time Out Settings). This value should be longer than the expected time interval so that the measurement completes before the time out occurs.

To save some typing, attached is a counter state file with the settings listed above (including a 2s time out). Extract the state file from the zip file. To load the state, store the extracted state file (STATE_TI.sta) on a USB thumb drive and insert the thumb drive on the front panel of the counter. Press the following keys:

Utility   Store/Recall   Recall State

Using the knob on the front panel, highlight the state file (STATE_TI.sta) and then press ‘Select’.

Good luck.
Attachments
- interval.jpg18.6 KB
Like • Show 0 Likes0
Actions
- lhornburg @ lhornburg on Apr 13, 2012 2:16 PM
  Mark Correct
  Correct Answer
  Forgot to attach the state file. here you go
  Attachments
  - STATE_TI.zip227 bytes
  Like • Show 0 Likes0
  Actions
- tamourjaved @ lhornburg on Apr 14, 2012 1:56 AM
  Mark Correct
  Correct Answer
  Dear lhornburg ,
  
  Thank you so much for the detailed reply, this helped me a lot. But the problem still is not resolved because I don't want to manually trigger the counter, I want the counter to start triggering when the signal comes at channel 1 and stop triggering when the signal comes at channel 2 without me pressing the trigger button. I don't know what exactly will be my experimental time(experimental time= apparatus setup time + test time), I know my test time will be around 20-30 msec, but the apparatus setup time can take around ten minutes, so does this mean that I can set the time out to around 10-15 mins ,than push the trigger button ,than setup all my experiment and than once my channels starts producing the signal during the test time they will automatically trigger the counter and hence give me the required time interval?
  
  Sorry to bother you but I haven't used these counters before so hope you can bear with me. Hope to hear from you soon.
  
  Thanks.
  Like • Show 0 Likes0
  Actions
  - lhornburg @ tamourjaved on Apr 18, 2012 9:56 AM
    Mark Correct
    Correct Answer
    Thank you for your question. It's not a bother. What is important is that you are able to use the counter to make the measurement you need to.
    
    If set up time and test (measurement) time must both be considered, you can always turn the measurement time out 'off'. In this case the instrument will not time out, but will wait indefinitely until the measurement completes. This may be good or bad. If you don't want the counter to hang due to a measurement that doesn't complete, go ahead and set a time out greater than 10-15 minutes in case set up time takes longer than expected.
    
    In regard to the set up I sent you, change the trigger source from 'external' to 'immediate'. A time interval measurement starts when the signal on channel 1 passes through a specified level (amplitude) in the specified direction (rising - POS, falling- NEG). The measurement stops when the signal on channel 2 passes through its programmed level and direction. With the trigger source set to immediate, the measurement will not start (or stop) until the signal conditions (channel 1) conditions are met. You do not have to press the 'Trigger' key.
    
    Let us know how this works and if you have any more questions.
    Like • Show 0 Likes0
    Actions
buzz38 Dec 23, 2013 9:43 PM
Mark Correct
Correct Answer
I have a similar need as the OP, measuring interval between start and stop.   My questions pertain to the 53230A, which has a 20ps resolution.

First question. What is the minimum interval? I'm confused about this. On the datasheet under "Range, Time interval A to B, B to A" it reads "-0.5ns to 100,000s (min)". It has a negative minimum? I didn't know negative time was even possible.

In keeping with this question, say we're measuring the propagation time for a signal. For example, see this video. In this video they are using A/B to measure the propagation time, and consequently the length of the cable connecting A to B. They measure 4.616ns. So I take it that means the cable is about 4.54 ft long?

What if the cable was only half a foot long? This would be about 0.508 ns or 508ps, which is well above the 20ps resolution. Would the device be able to measure this?

http://www.youtube.com/watch?v=eJ_IbJuI5TE

Second question. What is meant by "single shot" resolution? This term implies that it happens only once. My understanding is that this value, 20ps in the case of the 53230A, is meant to represent the minimum accuracy between channels. So if we were to freeze frame all the channels at the same point in time, then measure the difference. So it can measure down to a 20ps difference between A/B/C. And this means that any value which comes back is accurate to +/- 20ps. So in the case of measuring the cable for example, that means the figure of 4.54 ft could be off by .02 ns or 0.0197 ft. So really it could be between 4.52 and 4.56 ft.

Third question. If the device is set to trigger starts/stops as described in this thread, can it handle multiple starts/stops simultaneously? So could I do two starts, then two stops? So effectively every time I "start" it launches a new timer. ?? And then paired with this, what is the minimum interval between successive timers? I'm assuming that this is where the frequency limitations come in. A/B is limited to 350MHz, which translates to 2.8ns. So I would have to wait at least 2.8ns before starting another timer?

What if I use the RF channel though? Say I have the 6GHz option. Does this mean I would only have to wait 166ps between start triggers? And if so, I assume the problem then becomes that my stop trigger is still limited to 2.8ns.

Fourth question. Can you start/stop on the same channel? Like could I define a threshold of >=3V = start and <=-3V = stop? If so, this would allow me to overcome the disparity between the high frequency RF channel and the low frequency A/B channels. However, I assume this also means I'm increasing my minimum interval... to 166ps in this case.

Fifth question. Is the raw data available through USB or other IO? This is a huge issue I have with the Tektronix function counters. They won't give me the raw data, ie X,Y axis of time and voltage levels. Instead they will only provide the by-product of the data after the meter analyzes it. So I can get a CSV of time intervals, etc. Can I get the raw data from the Agilent units? Allowing me to run my own math if so desired?
Like • Show 0 Likes0
Actions
- buzz38 @ buzz38 on Dec 26, 2013 5:12 PM
  Mark Correct
  Correct Answer
  *bump. Looking to make a quick purchase before end of the year. Any application engineers working right now?
  Like • Show 0 Likes0
  Actions
  - buzz38 @ buzz38 on Dec 30, 2013 6:55 AM
    Mark Correct
    Correct Answer
    Bump
    Like • Show 0 Likes0
    Actions
jdwindrvr Dec 30, 2013 5:42 PM
Mark Correct
Correct Answer
Hello:

+First question. What is the minimum interval? I'm confused about this. On the datasheet under "Range, Time interval A to B, B to A" it reads "-0.5ns to 100,000s (min)". It has a negative minimum? I didn't know negative time was even possible.+

The 53230A by definition is a universal counter and not a time interval counter, but it is able to make negative time interval measurements to -3 ns ( -0.5 ns specified). A negative time interval measurement means that for a two channel time interval measurement, an edge was received on the ‘stop’ channel before the edge on the ‘start’ channel.

+In keeping with this question, say we're measuring the propagation time for a signal. For example, see this video. In this video they are using A/B to measure the propagation time, and consequently the length of the cable connecting A to B. They measure 4.616ns. So I take it that means the cable is about 4.54 ft long?+

+What if the cable was only half a foot long? This would be about 0.508 ns or 508ps, which is well above the 20ps resolution. Would the device be able to measure this?+

Cable delays are generally specified and tested relative to the velocity of light. In a vacuum the velocity of light is 3x10^8^ m/s or approximately 1 ft/ns. Velocity factors of the cable (impedance, dielectric constants) and the cables being outside of a vacuum limit the propagation velocity to approximately 2/3 the speed of light. Performing the measurement shown in the video, I measured a time interval of 6.1 ns for a cable from channel 1 to channel 2 that is just over four feet in length. Assuming a velocity of 0.5x the speed of light (1.5 ns/ft), the calculated length is: 6.1 ns/1.5 ns = 4.06 ft.

The 53230A can measure cable lengths well below ½ foot.

+Second question. What is meant by "single shot" resolution? This term implies that it happens only once. My understanding is that this value, 20ps in the case of the 53230A, is meant to represent the minimum accuracy between channels. So if we were to freeze frame all the channels at the same point in time, then measure the difference. So it can measure down to a 20ps difference between A/B/C. And this means that any value which comes back is accurate to +/- 20ps. So in the case of measuring the cable for example, that means the figure of 4.54 ft could be off by .02 ns or 0.0197 ft. So really it could be between 4.52 and 4.56 ft.+

Single shot resolution (20 ps) is the resolution that is available with a single time interval measurement. That is, the measurement is not made using time interval averaging, for example. Averaging increases the resolution to well below 20 ps as indicated in the video.

+Third question. If the device is set to trigger starts/stops as described in this thread, can it handle multiple starts/stops simultaneously? So could I do two starts, then two stops? So effectively every time I "start" it launches a new timer. ?? And then paired with this, what is the minimum interval between successive timers? I'm assuming that this is where the frequency limitations come in. A/B is limited to 350MHz, which translates to 2.8ns. So I would have to wait at least 2.8ns before starting another timer?+

I am not certain of the meaning of this question regarding simultaneous starts/stops. Per the data sheet (pg. 18), the minimum edge repetition rate is 6 ns ( 160 MHz):

The 6 ns (160 MHz) limitation provides the counter time to complete the current measurement and prepare for the next measurement, and provides sufficient bandwidth to accurately measure the edges of non-sine wave input signals.

+What if I use the RF channel though? Say I have the 6GHz option. Does this mean I would only have to wait 166ps between start triggers? And if so, I assume the problem then becomes that my stop trigger is still limited to 2.8ns.+

Time interval measurements (2-channel and single channel) can only be made using channels 1 and 2. Time interval measurements cannot be made using optional channel 3. Channel 3 measurements include frequency, period, ratio, time stamp, and with Option 150, the microwave pulse measurements shown below:

+Fourth question. Can you start/stop on the same channel? Like could I define a threshold of >=3V = start and <=-3V = stop? If so, this would allow me to overcome the disparity between the high frequency RF channel and the low frequency A/B channels. However, I assume this also means I'm increasing my minimum interval... to 166ps in this case.+

The 53230A provides single channel time interval measurements and allows you to specify different start and stop thresholds and edges (positive/negative). Note that the 6 ns edge repetition rate also applies to single channel measurements.

+Fifth question. Is the raw data available through USB or other IO? This is a huge issue I have with the Tektronix function counters. They won't give me the raw data, ie X,Y axis of time and voltage levels. Instead they will only provide the by-product of the data after the meter analyzes it. So I can get a CSV of time intervals, etc. Can I get the raw data from the Agilent units? Allowing me to run my own math if so desired?+

The 53230A returns the measured interval. There is no processing/analysis of the data (assuming the math/statistics function is off). Readings (up to 1M) can be stored in memory and then transferred to the PC.

If you have more information the specific measurement(s) you wish to make (the end result), we may be able to provide additional counter details that can help you select the best counter for your application.
Like • Show 0 Likes0
Actions
- buzz38 @ jdwindrvr on Dec 30, 2013 9:13 PM
  Mark Correct
  Correct Answer
  Very much appreciated, this is exactly the information I was looking for. Some interesting things I wasn't expecting. A few more questions. My interest in the device is several-fold. The main uses are for very high precision time interval measurement, essentially just a stopwatch for measuring things like signal propagation time, as well as for measuring the effect of jitter in oscillating circuits, and for performing frequency analysis on some RF signals. I feel that you've satisfied my questions as far as the stopwatch is concerned.
  
  I want to jump in the jitter analysis a bit more. Based on what you've said regarding the inability to measure time intervals through the (optional) 6GHz channel (I wasn't expecting this), what are the limitations for measuring jitter? For example, does this mean the maximum clock speed we can analyze is 350MHz? So I could not measure a 1GHz clock directly? I could however still measure a process running on a 1GHz clock, so long as the duty cycle were at least 1/3. So like if I wanted to look at the time drift on an internal clock on a 1GHz ARM processor. And by internal clock I mean the timekeeping process programmed into the firmware which increments by 1 microsecond (or whatever the resolution is). This duty cycle would be well below the 1GHz oscillator so I should be able to reasonably measure that, correct?
  
  For the optional 6GHz frequency counter, what type of data is available there? Meaning is the raw data available and what does that data look like? Is it essentially just time on the X axis and voltage on Y, like an oscilloscope? What is the microwave pulse measurement, how is that useful? I think you meant to include something in your post but it seems to be missing.
  
  Thanks
  Like • Show 0 Likes0
  Actions
jdwindrvr Dec 30, 2013 5:50 PM
Mark Correct
Correct Answer
Hello,

Here are the images that were to accompany my reply.
Attachments
- edge repetition rate.jpg47.2 KB
- Channel 3 Measurements.jpg50.9 KB
- single channel time interval.jpg70.0 KB
Like • Show 0 Likes0
Actions
- buzz38 @ jdwindrvr on Jan 2, 2014 3:53 PM
  Mark Correct
  Correct Answer
  Bump
  Like • Show 0 Likes0
  Actions
jdwindrvr Jan 6, 2014 10:48 AM
Mark Correct
Correct Answer
Hello,

To confirm, time interval measurements are only available on counter channels 1 and 2. Frequency and period measurements are available on channel 3, along with the pulse microwave measurements mentioned. These include pulse repetition interval - PRI, pulse repetition frequency - PRF, burst carrier frequency, and burst width. These are indicated in the drawing below. These measurements require option 150 in addition to the channel 3 option.

Regarding jitter measurements, these measurements are generally made by taking single-channel period measurements or single-channel time interval measurements depending on the type of jitter analyzed (period, time interval). The specifications (minimum width, minimum edge repetition rate) indicate the limitations of the measurements. Note that the minimum edge repetition rate is 6 ns which limits the input frequency to approximately 166 MHz.

In case you have not seen these, the following link is a discussion of jitter measurements using a counter and the attached measurement brief also covers jitter measurements.

http://gpete-neil.blogspot.com/2012/05/jitter-measurements-with-counter.html
Attachments
- Channel 3 Measurements.jpg50.9 KB
- Single Channel T-I.png42.3 KB
- 5990-5648EN.pdf594.9 KB
Like • Show 0 Likes0
Actions

how to measure time interval using 53220A series counter

Outcomes

Attachments

Attachments

Attachments

Attachments

Related Content

Recommended Content