Atten ADS-1102CML Low Freq + Volts Question

[ACvolts]

I am thinking of buying the Atten ADS-1102CML, but I really need to have the
actual lowest Hz + Volts you can get out of it...
The Spec-pdf of the Atten 1102CML states its = to or < than 10 Hz
but it's not exactly clear.

Could someone who owns a Atten ADS-1102CML or same brand of Oscope
do a test for me and let me know the exact Hz + Volts that it will display a
Trigger and also Trigger into Long Memory?

I know that the Rigols can do low Hz's and Volts, there is a video on Youtube
somewhere and I saw a 1102e display 2 Hz with around 2 Volts!  So I might
have to buy that one instead.   

Thanks in advance!

ACvolts

[Herman]

in the picture, the signal is 2Hz and 2V, and it's trig'd.
"its = to or < than 10 Hz"   it's digital frequency counter    f<10Hz
atten is OEM by siglent.

[valentinc]

   I have a Siglent SDS1072CML, it's the same as the atten (except the badwidth)...

   It works ok at low frequencies, I tested it from 50 Hz down to 1 mHz (signal form a Siglent SDG1020 AWG). The amplitude is exactly as it should be (4 Vpp).

   Note that below 50 ms/DIV the trigger runs in SCAN mode. Above that the normal trigger works just fine.

   Also, the Long Memory cannot be used in SCAN mode (and you won't need it anyway). The frequency counter (which is based on the trigger, it's not an actual hardware counter) also doesn't read anything below 10 Hz (because it can't in SCAN mode)...

   About the voltage, below 30 mVpp it tends to become a little noisy, I will post some pictures to make an idea ... But note that below 10 mVpp (or so, I don't remember exactly) the function generator is a little noisy too (I tested it with my Tektronix 2235 analog scope), so the picture with 4mVpp it's not very relevant about the noise performance of this scope. But still, you can make an idea ...

The actual V/DIV are /10 (what is on the screen), it was set in the x10 probe in the menu ...

4mVpp


10mVpp


100mVpp


1mHz


What exactly is your need with this oscilloscope ? If you have more questions, let me know.

[ACvolts]

in the picture, the signal is 2Hz and 2V, and it's trig'd.
"its = to or < than 10 Hz"   it's digital frequency counter    f<10Hz
atten is OEM by siglent.

If it displays the waveform than I could always check the Frequency with a DMM.
I'm testing low frequency waves on plants thru electronics.  The waveform displayed
is the most important thing I guess.  The fact that the Atten has a long screen and
has 2 meg memory is a big plus.  It's also the way the Atten is layed out.  I looked
at Siglent and Atten and decided on the 1102CML.  I also looked at GW Instek
and Rigol but price and features is important here.  The Rigol 1102e only has 1
meg memory and the GW Instek has a small screen and slow in the FFT.  But the
electrician component parts in the GW Instek was really what I am looking for.
Due to the fact that it has FFT RMS, Displays by Ground/Center and it also has
the capability to use current probes and that maybe why it takes so long to update.
GW Instek needs to work on the speed of their display before I buy the product.

So it's the Atten.  The Siglent is good but I need the physical layout of the Atten.

Thanks alot for your help.  Instead of the <10 they should actually put the Hz
though.  Maybe it's an issue from patent law which is confining the display to this
readout.  Lawyers!  Humph!

ACvolts

[valentinc]

     The fact that the oscilloscope doesn't display frequency below 10 Hz, like I said, is because it's based on the trigger data. It's not a real counter. And below 10 Hz the trigger runs in SCAN mode (meaning that it doesn't draw a full screen at one trigger, and will "scan" the screen like an analog oscilloscope) and that's why in this mode it doesn't know the frequency ...

     In this case the only solution is a scope with a real frequency counter ... But none in this price range has that.

     By the way, if you say that price is important and you work only at low frequencies why don't you go for the Atten ADS1072CML (70 Mhz bandwidth) ?

[ACvolts]

     The fact that the oscilloscope doesn't display frequency below 10 Hz, like I said, is because it's based on the trigger data. It's not a real counter. And below 10 Hz the trigger runs in SCAN mode (meaning that it doesn't draw a full screen at one trigger, and will "scan" the screen like an analog oscilloscope) and that's why in this mode it doesn't know the frequency ...

     In this case the only solution is a scope with a real frequency counter ... But none in this price range has that.

     By the way, if you say that price is important and you work only at low frequencies why don't you go for the Atten ADS1072CML (70 Mhz bandwidth) ?

Yeah that's right..These scopes don't have an >actual< counter!..
Why the 100mhz?..
Room to grow, the 1102cml is good to have for the rainy day that
you would say, "Oh shhhh...I should've bought the 100mhz"!

I found the ADS-1102cml at deal express for: $532.00 + Free Shipping!
After I buy it, it's broke city for a bit for me.  I have a old Tek 2232 that's
broke.   

But, if someone knows a more reputable dealer in the USA, feeeeeel free to
drop in a URL.  It maybe a good idea to have some security. 

I almost bought the Rigol 1102e but with the 7" screen and the 2meg memory
I'm set in a better position.  Plus the interesting plugged software some
kwel user has made for Linux using the Atten is great! 

ACvolts

[valentinc]

What problem has your Tektronix 2232 ? 

[ACvolts]

What problem has your Tektronix 2232 ?

I bought it used and it worked a bit before it didn't display a waveform properly.
Numerous issues.  Eventually I'll look into it.

ACvolts

[Deckert]

I am thinking of buying the Atten ADS-1102CML, but I really need to have the
actual lowest Hz + Volts you can get out of it...
The Spec-pdf of the Atten 1102CML states its = to or < than 10 Hz
but it's not exactly clear.

Thanks for your direct emails - as per your suggestion I'll put my tests into this thread as well.

Tests were done on a ADS1102CML with firmware 3.01.01.31R16 and hardware 6-41-2.0.

Note that the scope switches to scan mode only if in Auto-trigger mode. If triggering in normal mode it won't scan, but instead will measure the whole waveform, then update the display with the newly measured wave. Using the measure control button I added a few software measurements to the right of the display. The frequency value was updated correctly regardless of whether it was in scan or normal trigger mode.

At 250ms/div it scans through the wave every 4.5s and also updates the frequency on the right every 4.5s. 0.97Hz is the lowest I can go on my crusty old signal generator.

Incidentally, I believe the frequency counter ("f<10Hz" at the bottom) is in fact a hardware counter. As long as it triggers, the counter works fine. I've measured frequencies of up to 158MHz, but that's the highest frequency source I had. I suspect it will go all the way to 200MHz (which is the top-end of this series).

The noise in the 100mV screenshot is due to me not using coax from the sig-gen to the 'scope. The sig-gen was set to 100mV RMS, so it shows the scope also correctly measures/calculates that value.

--deckert

[valentinc]

Note that the scope switches to scan mode only if in Auto-trigger mode. If triggering in normal mode it won't scan, but instead will measure the whole waveform, then update the display with the newly measured wave. Using the measure control button I added a few software measurements to the right of the display. The frequency value was updated correctly regardless of whether it was in scan or normal trigger mode.

I can confirm that on my Siglent, I didn't notice that because haven't used the scope at so low frequencies...

Incidentally, I believe the frequency counter ("f<10Hz" at the bottom) is in fact a hardware counter. As long as it triggers, the counter works fine. I've measured frequencies of up to 158MHz, but that's the highest frequency source I had. I suspect it will go all the way to 200MHz (which is the top-end of this series).

If it was a hardware counter its displayed value did not depend on the trigger... Also note that at high frequencies (above 1 Mhz or so at my Siglent) when the displayed signal tend to bounce around the screen the actual decimals in the counter changes with the  "jitter" of the signal: that concludes to only one thing: the Hz displayed on the screen is calculated from the image on the screen, not the actual signal itself (which I measured with a frequency counter and its frequnecy does not change at the first decimal places as the scope indicates)

About the 158 Mhz measured, you can measure that because the roll off of the front end amplifier is not very steep beyond de 100 Mhz -3dB point. And maybe all the models share the same front end with and with a shift register and a varactor circuit they set the bandwidth in software ... (like in the Rigol DS1052E)

I bought it used and it worked a bit before it didn't display a waveform properly.
Numerous issues.  Eventually I'll look into it.

It's worth reparing that scope. It's much more stable at displaying waveforms than this low-end digital scopes (I have a Tek 2235). It's good to have one in the lab

[Deckert]

If it was a hardware counter its displayed value did not depend on the trigger...

From my ADS1000 series manual:



I'm pretty sure they won't have that and also a separate "software" frequency counter if both was done in software. Either way, I'm very happy about how the frequency counter works - it's accurate and quite stable for me.  I rarely use my Racal-Dana frequency counter these days.

--deckert

[rf-loop]

About the 158 Mhz measured, you can measure that because the roll off of the front end amplifier is not very steep beyond de 100 Mhz -3dB point.

Siglent SDS1102CML:
PCB version 11 and FW version 5.
With 50mV/div
Signal from HP8657B with 1m 50ohm coaxial and 50ohm HP terminator in scope end using T connection.

With EqualTime capture
compared to 1MHz as reference level 6dic p-p on scope screen.

-3dB around 145MHz
-6dB around 185MHz
-10dB around 240MHz

with 300MHz signal still trig nice ok if use E.T. (can not trig using ral time mode)
Signal 300mVpp. With 300MHz it still show 50mVpp.
So roughly -16dB


With real time capture
-3dB around 132MHz
-6dB around 165MHz

-------------------
About counter. It take signal from trig line afaik and count these events. This trigger pathway starts before ADC from front end analog amplifier.  Whi it starts from around 10Hz, I do not know. (but perhaps I can quess but just becouse it, I do not tell it. Facts are more nice.)

[valentinc]

-3dB around 145MHz

 I wonder if it can do 0 dB at 100 Mhz or around ... If you can, please do a measurement at 100 Mhz (in the same conditions mentioned above) and post the results.

 The roll off is better than I expected anyway

[rf-loop]

-3dB around 145MHz

 I wonder if it can do 0 dB at 100 Mhz or around ... If you can, please do a measurement at 100 Mhz (in the same conditions mentioned above) and post the results.

 The roll off is better than I expected anyway

Later as I have time  I will do better tests.

Last time I looked it looks like around -1dB at 100MHz

But whole this test was not very reliable due to Signal pathway (T connected terminator and also HP 8657B have not checked for flatness. Next time I will test using HP8644B what is also checked for flatness. (it need remember, external 50ohm terminator on oscilloscope input is NOT 50ohm impedance due to oscilloscope input itself reactance. But under 200MHz not very extremely bad mismatch))

[rf-loop]

Here is Siglent SDS1102CML   frequency response picture from sweep.

Now generator used i test is HP8644B (and this flatness is checked so that I can trust it - enough accurate for this test and in better condition than HP specs limits)

50ohm good cable and Tektronix 50ohm  feed thru terminator.
(this is not perfect becouse its impedance over thius freg span is not 50ohm due to oscilloscope input reactance - but it is acceptable also becouse this is how so many peoples do connection  in practice. Now there can see some amount of this "wave" in shape but not so much... I think no one can calculate my cable lenght looking this picture... so... )

Oscilloscope in  peak mode and display vectors and hor speed 10s/div.

It starts so that Generator first run 1MHz with level -3dB below 300mVp-p displayed level on oscilloscope. Then 3x 1dB step up and from 1MHz line start linear sweep 1MHz to 300MHz in 100s sweep time. (after stop, horizontally zoomed to 5s/div and freg scale adjusted exatly for scope divs)


Result: Not so bad...



Later after I get ready more tests with SDS1072CML, SDS1102CML I will then look the appropriate place to throw them to see.

[tinhead]

Siglent SDS1102CML:
PCB version 11 and FW version 5.
With 50mV/div
Signal from HP8657B with 1m 50ohm coaxial and 50ohm HP terminator in scope end using T connection.

With EqualTime capture
compared to 1MHz as reference level 6dic p-p on scope screen.

-3dB around 145MHz
-6dB around 185MHz
-10dB around 240MHz

with 300MHz signal still trig nice ok if use E.T. (can not trig using ral time mode)
Signal 300mVpp. With 300MHz it still show 50mVpp.
So roughly -16dB


With real time capture
-3dB around 132MHz
-6dB around 165MHz

i wish some day Hantek will understand it and code proper working EQU sampling, like Siglent did it now.

About counter.

It take signal from trig line afaik and count these events. This trigger pathway starts before ADC from front end analog amplifier.  Whi it starts from around 10Hz, I do not know. (but perhaps I can quess but just becouse it, I do not tell it. Facts are more nice.)

it is AC coupled and these frontends are designed to work from >10Hz, there is nothing wrong about.

[rf-loop]

it is AC coupled and these frontends are designed to work from >10Hz, there is nothing wrong about.

It was my quess that it is AC (but I did not know it sure) coupled to counter but then why becouse trig pathway can handle DC.

But then I think also counter resolution and counting gate open time and then counter resolution becouse my quess is it is just pulse counter - not expensive (complex) time interval machine.
But yes, of course nothing wrong. And who really need <10Hz counter if its resolution is 1Hz or something like this.

And if in other hand think scope input for waveform capture and display and AC coupling there... around 5- 20Hz corner  is normal typical compromise what  typically all manufacturers use for AC coupling. DC coupling limit is 0Hz but if go to very deep in thinking ... we can ask what is frequency of DC ... is it 0 Hz  or 1/infinite Hz.

And if in some cases really need example DC block but very very low frequency.. example for very  low freq noise measurements. It is simple. Scope input DC and external capacitor for DC block.

[valentinc]

i wish some day Hantek will understand it and code proper working EQU sampling, like Siglent did it now.

I wish that too, as in my Hantek DSO1060 the Equivalent Time Sampling does not work at all in most cases...

It starts so that Generator first run 1MHz with level -3dB below 300mVp-p displayed level on oscilloscope. Then 3x 1dB step up and from 1MHz line start linear sweep 1MHz to 300MHz in 100s sweep time. (after stop, horizontally zoomed to 5s/div and freg scale adjusted exatly for scope divs)

   The result isn't bad at all indeed. I'm wondering about another thing regarding the probes of these scopes: mine are markes 70 Mhz, I wonder if they bother to sell different probes for let's say the 70 and 100 Mhz version as they sell cheap probes with the scope anyway (I can't see a correct shaped square wave of above 2-3 Mhz with original probes, but with some chinese cheap probes that were advertised as 100 Mhz -3 dB bandwidth I can).