DSO Recomendation

[JPortici]

and correlation in time domain between wiggles and events!

[bsas]

Thanks again for the advises!

Another newbie question, sorry about that:

- Is it possible to use the DS1054Z to plot a frequency response graph? For example, if I use a sweep frequency generator from 100Hz to 20kHz, can I setup the DSO to collect all and plot a historic graph? If not, is there any DSO under $1000 that can do that?

Thanks!

[Fungus]

- Is it possible to use the DS1054Z to plot a frequency response graph? For example, if I use a sweep frequency generator from 100Hz to 20kHz, can I setup the DSO to collect all and plot a historic graph? If not, is there any DSO under $1000 that can do that?

No, but you can use (eg.) an Analog Discovery for that:

https://www.eevblog.com/forum/projects/analog-disovery-as-audio-analyzer/

Plus: The Analog Discovery is a really useful toy when you start with digital electronics.

They normally cost about $200.

[Dan Moos]

Yes, you can do a frequency response graph (bode plot), but you are gonna have to do a little extra work. Dave did a video in fact.

The gist is, you send the trigger pulse from your function gen to one channel on your scope, and the swept signal to another. If you set your scope's time base properly, you should have a single full sweep steady on screen. It will look like the bode plots you're used to seeing.

OK, here's where the extra work comes. To find the -3db break point, do the math and set your cursor. To find the frequency at any point on the graph, it's gonna be easy or hard depending on whether you swept with log, or linear. Also, you need a function gen that can have the sweep set exactly. But it works.


Sent from my E6830 using Tapatalk

[rstofer]

- Is it possible to use the DS1054Z to plot a frequency response graph? For example, if I use a sweep frequency generator from 100Hz to 20kHz, can I setup the DSO to collect all and plot a historic graph? If not, is there any DSO under $1000 that can do that?

No, but you can use (eg.) an Analog Discovery for that:

https://www.eevblog.com/forum/projects/analog-disovery-as-audio-analyzer/

Plus: The Analog Discovery is a really useful toy when you start with digital electronics.

They normally cost about $200.

I wish I could have bought one for just $200 but, alas, us really senior citizens don't seem to get the student discounts (if such even exist for this product).  Just another example of age discrimination (joking...mostly).

One of the very neat features of the Analog Discovery is the Network tool.  Sweep over some range of frequencies and get back amplitude and phase.

If the Analog Discovery was the only piece of test equipment I had, I could still do a lot of electronics.

http://store.digilentinc.com/analog-discovery-2-100msps-usb-oscilloscope-logic-analyzer-and-variable-power-supply/

Attached is a Network sweep of a simple RC low pass filter using a 6800 Ohm resistor and a 10 nF capacitor.  We expect the 3 dB point to be somewhere around 2350 Hz and it measures 2545 Hz.

For those working on tuned circuits, this tool is outstanding.

[MrW0lf]

I would pay $2000 for a 4-channel, 50MHz, hackable to 100MHz scope with 1% DCV accuracy.

Bit OT but since you asked:

https://www.picotech.com/oscilloscope/5000/flexible-resolution-oscilloscope
5443A, 4ch, 64MB, 1GSa/s
US$2,055.00 / €1,745.00

DC accuracy

±50 mV to ±20 V ranges,  >= 12-bit mode: ±0.25% typical @ 25°C (±1% of full scale max @ 20 to 30°C)
±50 mV to ±20 V ranges,  8-bit mode: ±1% typical @ 25°C (±3% of full scale max @ 20 to 30°C)
±10 mV and ±20 mV ranges: ±2% typical @ 25°C (±5% of full scale max @ 20 to 30°C)

However must be aware of sampling rate limitations in 12-bit mode:
any 1ch 500 MS/s
any 2ch 250 MS/s
any 3ch 125 MS/s
4ch 125 MS/s
(twice high in 8bit mode with ±1% typical @ 25°C)

But not hackable, it's already 100MHz So accurate scope is not some pipe dream, quite affordable if need this sort of thing. Tek/Agilent take about as much money for "regular" 8bit scopes.

[nctnico]

A $10 multimeter can usually give you better than 1% DCV accuracy at room temperature, right out of the box.
I would pay $2000 for a 4-channel, 50MHz, hackable to 100MHz scope with 1% DCV accuracy.

This is a contradiction since DC accuracy says nothing about AC accuracy and oscilloscopes are typically used for looking at AC (-ish) signals. 

[MrW0lf]

This is a contradiction since DC accuracy says nothing about AC accuracy and oscilloscopes are typically used for looking at AC (-ish) signals. 

Possibly this calls for some testing. Should be easy - can use high quality square. Look for overshoot at transitions. DC level can be confirmed on second channel + multimeter and compared to square tops. DC - peak = AC error. Or something similar. Sounds valid procedure? Sadly do not have 5000 but could test with latest 2000 soon.

[JPortici]

Yes, you can do a frequency response graph (bode plot), but you are gonna have to do a little extra work.

not a bode plot, just frequency responce. you don't have any information on the phase

i think that with a picoscope you could have both amplitude and phase (by using the math channels. you literally enter the function you want to plot)

[bsas]

For my current need which is audio applications (specifically guitar pedals and pre-amps) sounds like the "Analog Discovery 2" looks better for me.
But it is $279 and I still need to use my computer.

In the other hand, the Rigol DS1054Z is $399 (just $120 more) and it is a full independent DSO, and can be used with my computer too, right?

My new questions are:

1) How much more precise and better the Analog Discovery is compared to the 1054Z? And how much flexible it can be for all other applications that the 1054Z can provide me?
2) I can see in the future having both units, but for a beginner like me that I will be learning basic components like opamps or basic transistor circuits, which on will be better to start with?

Thanks!

[JPortici]

1) How much more precise and better the Analog Discovery is compared to the 1054Z? And how much flexible it can be for all other applications that the 1054Z can provide me?

a lot. you don't care about bandwidth. you care about resolution and noise (more bandwidth <-> more noise)
also you have a signal generator, pattern generator, ...
I agree that it's annoying to use it with a computer but while you are designing/programming you probably have your computer besides you anyway

2) I can see in the future having both units, but for a beginner like me that I will be learning basic components like opamps or basic transistor circuits, which on will be better to start with?

i think the AD is still better because you have more tools like signal generator and you still don't need to care about the bandwidth.
If the choice has to be between these two and you don't have other equipment i'd be inclined to say go with the AD

[bsas]

What is the difference between the analog discovery 1 and 2?

[rstofer]

For my current need which is audio applications (specifically guitar pedals and pre-amps) sounds like the "Analog Discovery 2" looks better for me.
But it is $279 and I still need to use my computer.

In the other hand, the Rigol DS1054Z is $399 (just $120 more) and it is a full independent DSO, and can be used with my computer too, right?

My new questions are:

1) How much more precise and better the Analog Discovery is compared to the 1054Z? And how much flexible it can be for all other applications that the 1054Z can provide me?
2) I can see in the future having both units, but for a beginner like me that I will be learning basic components like opamps or basic transistor circuits, which on will be better to start with?

Thanks!

The AD has a 14 bit A/D converter versus 8 bits for most scopes including the DS1054Z.  That's a HUGE difference in resolution!  The DS1054Z will have a higher bandwidth and 4 channels but, in my view, 4 channels are better used in the digital domain and the AD has 16 bits of digital IO.  That means I get the full SPI transaction with 12 channels left over versus using all 4 channels of the DS1054Z.

The AD is perfect at the frequencies most used with op amps and transistor circuits.  I was fooling around with some transistor logic circuits the other day using the digital IO but then I turned around and built up a properly biased Class A transistor amplifier which I looked at with the Network tool of the AD.  You just can't do that with the DS1054Z.

I would highly recommend starting with the AD because it is a pretty complete electronics lab in a hand held box.  In my work area, I have several computer workstations.  No reason, really, but I'm never more than 3' away from a computer.  I haven't done it but I could certainly justify using LTspice to model a circuit and the AD to see how it really works.

Having the two arbitrary waveform generators, something the DS1054Z doesn't have, is perfect for providing signals to breadboard circuits.  Usually the waveform will be a simple sine wave but there are square, ramps, triangle and a few others plus you can invent your own.

I could see a college student buying one of these gadgets and using it throughout the EE program.  Add in a laptop with LTspice and Maxma (wxMaxima) and I think you have everything necessary to do the work.  OK, I would add Eclipse (the IDE) and GNU Fortran but that's just me...

I think the BNC adapter board is a necessary add-on:
http://store.digilentinc.com/bnc-adapter-board-for-the-analog-discovery/

As are two scope probes (maybe 4 if you want to use probes to inject signals):
http://store.digilentinc.com/bnc-oscilloscope-x1-x10-probes-pair/

Those can probably wait because, more often than not, I just use pin headers on the solderless breadboard and connect the signals with the flyleads on the header that comes with the AD.

I like the gadget a lot!  Nevertheless, it isn't my money and it isn't my decision.  I have both the AD and a DS1054Z plus there is a Tek 485 under the table.  I bought the AD years before I bought the DS1054Z and, to be honest, I only bought the scope because I wanted to see how they worked.  I could very well have just used the AD for the lower frequency stuff and the 485 for everything else - which is exactly what I did for quite a long time.

Download the software and play around with the Demo device.  Read the specs, see the tutorials, learn all you can about the gadget before you spend your money.  All in, the AD plus attachments will cost just about as much as the DS1054Z.  And you might consider the box of analog parts that go with the AD

http://store.digilentinc.com/analog-parts-kit-by-analog-devices-companion-parts-kit-for-the-analog-discovery/

[rstofer]

What is the difference between the analog discovery 1 and 2?

The 2 version has adjustable supplies and the ability to provide more power by adding a wall wart among other things.  For details, see here:

https://blog.digilentinc.com/a-comparison-of-the-analog-discovery-and-analog-discovery-2/

[mtdoc]

What is the difference between the analog discovery 1 and 2?

Different case.
Adjustable 0-5V power supply on the 2 versus fixed 5 V on the one - with the added ability on the 2 to power it external from the USB port to put out up to 2.1Watt per channel instead of 250mW.

Other than the power supply - no differences in the hardware or performance.

[MrW0lf]

i think that with a picoscope you could have both amplitude and phase (by using the math channels. you literally enter the function you want to plot)

Someone made nice app for that (phase and amp), but works with internal function gen, so limited by its max frequency.
https://bitbucket.org/hexamer/fra4picoscope/wiki/Home

Without phase could do sweep with (external) gen using FFT persistence mode.
Currently not sure if could do phase and amplitude plot with just math channels, would be interesting puzzle to solve sometime.

1) How much more precise and better the Analog Discovery is compared to the 1054Z?

AD2:

https://reference.digilentinc.com/reference/instrumentation/analog-discovery-2/reference-manual?redirect=1#features_and_performances

Resolution:   14-bit
Absolute Resolution(scale ?0.5V/div9)): 0.32mV
Absolute Resolution(scale?1V/div10)):   3.58mV
Accuracy (scale?0.5V/div, VinCM = 0V): ±10mV±0.5%
Accuracy (scale?1V/div, VinCM = 0V): ±100mV±0.5%

If try calculate for 5Vpp range:
(0.01 / 5 * 100) + 0.5 = +-0.7%

DS1000Z:

http://beyondmeasure.rigoltech.com/acton/attachment/1579/f-04cf/0/-/-/-/-/file.pdf

DC Gain
Accuracy[3]
<10 mV: ±4% full scale
>=10 mV: ±3% full scale

So in different ballpark.

[Fungus]

1) How much more precise and better the Analog Discovery is compared to the 1054Z?

In your frequency ranges (up to 100kHz)? A lot. The 14 bit DAC will make a huge difference in FFTs, etc.

Plus the AD works as a signal generator and you can easily automate things like frequency response testing using it.

for a beginner like me that I will be learning basic components like opamps or basic transistor circuits, which on will be better to start with?

Analog Discovery.

[DimitriP]

For my current need which is audio applications (specifically guitar pedals and pre-amps)

...and while you are deciding .... you can check out   
http://audio.rightmark.org/products/rmaa.shtml

and  http://www.daqarta.com/

[bsas]

Awesome feedback guys, thanks!

I am decided on the AD2. Just waiting for next paycheck in 2 weeks
Quick last question about it: what is the difference between "WaveForms" and "WaveForms 2015"?

[rstofer]

Awesome feedback guys, thanks!

I am decided on the AD2. Just waiting for next paycheck in 2 weeks
Quick last question about it: what is the difference between "WaveForms" and "WaveForms 2015"?

WaveForms 2015 is the newest version and the one I am using.  It has a lot more capability than WaveForms.  However, there are a number of college programs built around WaveForms so I suspect it is still available for those specific users.

I think you'll like the AD.  Breadboard some circuits and learn to use the device.  Something as simple as a low pass filter takes on a real life when you can get the Bode' plot with very little effort.  All of a sudden, the equations make sense!

The lowly transistor Class A amplifier with the emitter bias resistor will come alive with the Network gadget or even just the dual channel scope watching input voltage and output voltage.  You can expand the idea by using an op amp as a x10 inverting amplifier (for example) and plotting input/output.  The equations for gain vs feedback resistors rock!

Don't forget, you can use the Digilent Adept library to create custom instruments
https://reference.digilentinc.com/_media/waveforms_sdk_reference_manual.pdf

There is a lot of educational material related to AD on the Digilent site.

[Fungus]

I would pay $2000 for a 4-channel, 50MHz, hackable to 100MHz scope with 1% DCV accuracy.

But not hackable, it's already 100MHz

Not 100MHz with 4 channels on. 

[MrW0lf]

Not 100MHz with 4 channels on. 

Might just barely make it in 8 bit mode, ±50 mV to ±20 V range if room is typically at 25C

[kcbrown]

Not 100MHz with 4 channels on. 

Might just barely make it in 8 bit mode, ±50 mV to ±20 V range if room is typically at 25C

Why aren't other 8 bit scopes able to achieve the same thing?

[nctnico]

Not 100MHz with 4 channels on. 

Might just barely make it in 8 bit mode, ±50 mV to ±20 V range if room is typically at 25C

Why aren't other 8 bit scopes able to achieve the same thing?

Regular oscilloscopes have a much larger range and allow higher input voltages. This translates to a less accurate input divider.

BTW note that the jitter spec on the Picoscope 5000 series is severely inadequate for the high resolutions/samplerate they specify. See www.linear.com/docs/25374

[kcbrown]

Might just barely make it in 8 bit mode, ±50 mV to ±20 V range if room is typically at 25C

Why aren't other 8 bit scopes able to achieve the same thing?

Regular oscilloscopes have a much larger range and allow higher input voltages. This translates to a less accurate input divider.

I don't follow.  That would make sense if you were limited to a fixed voltage range, but if that were the case, then with an 8 bit ADC you wouldn't be able to see signals down to 2mV with a simultaneous maximum of 30VRMS on the input.  The smallest voltage change you'd be able to see in that case would be about 330mV.

If you're not limited to a fixed voltage range, as seems to be the case (indeed, when you're playing with the vertical scale knob on a DS1054Z, you can hear a relay switch over when transitioning between 5V/div and 2V/div) then you'd be able to get 8 bits worth of resolution in whatever range you're actually recording, as defined by how the scope maps the vertical settings to the internal gain and offset.  Given that, why is the Picoscope better?