Oscilloscope as Spectrum Analyzer/Frequency Response Analyzer

[sameerdhiman]

Hi,

Does someone know about any DIY adaptor to turn Analog Oscilloscope X-Y mode into Spectrum Analyzer and Frequency Response Analyzer ?

It would be great help for poor hobbyist like me who can purchase Spectrum Analyzer in dreams only.

 

[Svein_B]

You do not have to buy a spectrum analyzer if you have a PC with soundcard near your workbench.
Several free or inexpensive PC software do the job nicely.
A lot of user ARTA: ARTA Home

You will have to make a small interface to adapt the power and signal level to the input of your soundcard.

Svein.

 

[sameerdhiman]

Thanks Svein,

I have the setup (PC+ARTA) but I am not satisfied with the speed of screen refresh. It is most problematic when I have to tune notch filter's pot, adjusting a pot takes ages to be updated on screen.

If I can use my Analog Oscilloscope as Frequency Response Analyzer It would be TRUE REAL TIME Analyzer.

 

[gpapag]

As Svein has pointed out, you can use your sound card.
You can then record the signal using a wave editor free software and save it in “wav.” Format. This “wav.” File can be later analysed (FR/Spectrum/Distortion) by importing it either to ARTA or RMAA or some other SW package.

Your main question though is for the hardware part, that is, what physical connections (wiring, and/or adaptor) you will use to bring the signal to the input of your soundcard and to an appropriate amplitude level.
This seems to be a well thought interface:

Soundcard Interface


If you are using specifically your oscilloscope for the “front part” of signal probing, the “Y Out” of the oscilloscope is a convenient pick up point for the input of your soundcard. See here
http://www.diyaudio.com/forums/power-supplies/188975-lm317-experiments-measurements-2.html#post2573695

[Note: If your oscilloscope has a “Y Out” you can use it to feed your soundcard.
The probe will be connected to the Y Input of the oscilloscope, so effectively you are using the scope’s preamplifier/adjustable attenuator as a good quality adjustable preamplifier.]

and here:

http://www.diyaudio.com/forums/power-supplies/188975-lm317-experiments-measurements-3.html#post2577515

If your oscilloscope does not have a “Y Out” then, you may have a look here (it is an "ongoing" project, as I have to clear up some issues )

http://www.diyaudio.com/forums/equipment-tools/189545-implementing-y-out-oscilloscope.html#post2581456

Regards
George

 

[gpapag]

Sameer
There is somethink that I overlooked in your first post.
Are you specifically interested in capturing and analysing the X-Y mode signal of your oscilloscope?

If this is the case, then the "Y Out" will not do the job.

You have to tap both the CH1 and CH2 outputs from inside your oscilloscope and feed them to your sound card's Lin and Rin respectively.

Then manipulate them with software.

Regards
George

 

[sameerdhiman]

Thanks George,

Well the concept is having standalone adapter

Sweep Generator --> Adapter --> DUT --> Demodulator --> Y input
|--> Ramp Voltage --> X input

Sweep Generator can be stand alone or PC's sound card. Ramp Voltage will be provided by an adapter.

 

[macboy]

You don't need to use the XY mode to do this. You can do it with normal triggered sweep mode. The trick is to set the horizontal of the oscilloscope to a timebase that will give a full sweep in the same amount of time (or slightly longer) as the sweep generator takes to do a full sweep. The trigger output of the sweep generator can be sent to the trigger input of the scope, or just trigger off of the signal. One caveat is that you need to use a somewhat long sweep time (certainaly no less than 1 or 2 seconds) so it is beneficial to use a digital storage scope. Otherwise you depend on your persistence of vision to see the full sweep. You don't even need a demodulator to get a nice looking display. Another caveat is that the vertical is linear not logarithmic (dB).

 

[sameerdhiman]

Thanks macboy

Is there any IC available which can convert linear to log voltage ?

If I use PC as a sweep generator then from where I shall get the trigger signal ?

 

[SY]

Sound Cards

I use AudioTester and set the average number to 1 for real time measurements. With a 16k sample length, and using the old POS computer in my lab, the refresh rate is at least once per second.

 

[ej25awd]

Common IC used by hobbyist is the AD8307 log amplifier Got excellent dynamic range and comes in through-hole package.

I've used it for filter sweeping together with an HP 8601A sweep generator and a scope (with X-Y) capability.

Thanks macboy

Is there any IC available which can convert linear to log voltage ?

If I use PC as a sweep generator then from where I shall get the trigger signal ?

 

[gpapag]

You may build a log converter with common components ,which you may already have

http://www.national.com/an/AN/AN-30.pdf

Regards
George

 

[geraldfryjr]

What Frequency range?

 

[gpapag]

I haven’t built it.

From the pdf (page 2, left column):

Both of the transistors are used in the grounded base connection, rather than the diode connection, to eliminate errors due to base current. Unfortunately, the grounded base connection increases the loop gain. More frequency compensation is necessary to prevent oscillation, and the log converter is necessarily slow. It may take 1 to 5 ms for the output to settle to 1% of its final value. This is especially true at low currents. The circuit shown in Figure 2 is two orders of magnitude faster than the previous circuit and has a dynamic range of 80 dB. Operation is the same as the circuit in Figure 1, except the configuration optimizes speed rather than dynamic range.

If settling speed has anything to do with highest frequency, then, for Fig.2 circuit, using f =1/T, we have f= 1/0.00005sec, that is f=20kHz worst case and f=1/0.00001sec=100kHz best case

 

[ej25awd]

I know that building your own stuff from scratch is fun and fulfilling but sometimes you have to draw the line somewhere.

Re: NS AN-30, where can you find the thermistor used in the example? Is there a reference board that you can copy to save time? Is the person that will be building this log amplifier have enough equipment and resources to validate the finished unit's operating performance?

Personally, I'm not willing to spend days, weeks and even months designing something that already existed, tried, tested and the best part of all cheap!

Not here to argue, just pointing out the obvious

 

[gpapag]

Not here to argue

Hitachi_nut

No problem from my part.
I just hope you didn’t take my intervening as an attempt to argue with you either.

It was only that with my metrics, 12 UKP (from RS Components) isn’t cheap.
The penalty I have to pay (and gains I may get)for choosing the cheap components route though, is well described in your response.

I have nothing to confront your solid logic.
On the contrary, to add to it, from the AD8307 data sheet
http://www.analog.com/static/imported-files/data_sheets/AD8307.pdf
Fig 19, I can concur that the response time has nothing to do with freq.response.
Got me a null score here too.

Therefore, the question

What Frequency range?

Has to be answered by a knowledgeable member. Sorry geraldfryjr

Regards
George

 

[sameerdhiman]

What Frequency range?

Audio Band

 

[ej25awd]

AD8307 frequency range as specified on the datasheet -> DC-500MHz, not enough

Dynamic range is better than 88d B, Settling time < 500nsec

Almost everything you need to know is on the AD8307 datasheet. Analog Devices even offer samples for crying out loud. That's how I got mine before

AD8307 | Low Cost, DC to 500 MHz, 92 dB Logarithmic Amplifier | Log Amps/Detectors | Specialty Amplifiers | Analog Devices

There's also a guy on eBay selling them for <$10 shipped. There several ways to get the chip "cheaply" if look hard for it.

What Frequency range?

Audio Band

 

[ej25awd]

George,

I have no desire getting a score with this thread. No offense.

I just have first hand knowledge in the topic and I'm glad to lend a helping hand to DIY out there

Cheers!

Hitachi_nut

No problem from my part.
I just hope you didn’t take my intervening as an attempt to argue with you either.

Fig 19, I can concur that the response time has nothing to do with freq.response.
Got me a null score here too.

Regards
George

 

[gpapag]

hitachi_nut

It was my pleasure reading your posts here.



George