FNIRSI-1013D "100MHz" tablet oscilloscope

[Kosmic]

Hi!

Did anyone check if it was capable of 100Mhz? Or where its limit is?

Considering that the FNIRSI-5012H is not 500MS/s sampling rate and 100Mhz bandwidth. I would be really surprise if this one is 1GSa/s and 100Mhz. The frontend will probably have the same variable impedance (depending on the range) problem but the FPGA will certainly help with sampling rate.

[gfmucci]

I notice that this unit is sold under three names:  Daniu,  Fnirsi, Elikliv.

Daniu is sold at Banggood for $148.  Fnirsi is sold at Ali Express for $138. Elikliv on Amazon for $190.  But new customers at Banggood get a $20 first time discount (I think - I haven't tried it yet  - could be a come on - or maybe only for items $5,000 or more .)

Does anyone know the difference between these three brands?  Probably no difference, right?  Does someone else make them besides the names on them?

Would this unit be:  1) Overkill  2) About right, or 3) Inadequate for a noob as an electronics teaching/learning tool, working through a Make: Electronics book and kit or planning to explore Arduino in the near future.  I'm already using a couple of DMMs.

I'm thinking that even 1/2 the value of the specs. on this unit would still be decent for my purposes.  The touch screen interface seems to be more intuitive than most.

[boggis the cat]

Hi all.

I got one of these to test out.  Results so far are OK-ish – i.e. it does work.

• My unit is way out on vertical accuracy, yielding around 212 mVpp for a 200 mVpp 1 kHz square wave.  Specification is 2%, which seems optimistic, but 6% high is fairly poor.
• No way to store or restore setups, and fairly limited measurement options.
• I am not sure if it is set to a sinc mode or not – no setting for this.
• Waveform capture is two horizontal screens wide, so you can shift the window that much.  No hold-off or other such niceties.

It comes with quite two decent-looking older HP type probe copies.  I have not checked these as yet.


I have an oscilloscope calibrator here at work, so give some suggestions as to what I should verify and I will try to get it sorted.

This unit also came without a front bezel, from the 'Official FNIRSI' store on AliExpress.  Make of that what you will, but it suggests some form of re-work without proper QC to me...

[Fungus]

Hi all.

I got one of these to test out.  Results so far are OK-ish – i.e. it does work.

Cool. I don't expect it to be perfect at this price point but is there anything you'd say was a deal breaker? How's the touch screen? Good update rate? How easy it is to set horizontal/vertical scale? Does it crash and need rebooting?

Are you "This is OK...", or "What a pile of garbage!"?

• My unit is way out on vertical accuracy, yielding around 212 mVpp for a 200 mVpp 1 kHz square wave.  Specification is 2%, which seems optimistic, but 6% high is fairly poor.

That wouldn't worry me at all. "About 200mV" is fine.

(and yeah, 2% is never going to happen with an 8-bit ADC on a device like this)

• No way to store or restore setups, and fairly limited measurement options.

Sooooo annoying if it doesn't power on to a known state (preferably the same as power-off).

• I am not sure if it is set to a sinc mode or not – no setting for this.

It will be obvious if it doesn't have it: Zoom in on a rising edge and look for Gibbs phenomenon, ie. a sinc will show ringing before the signal starts to rise.

• Waveform capture is two horizontal screens wide, so you can shift the window that much.  No hold-off or other such niceties.

Not ideal, but not a deal breaker.

I have an oscilloscope calibrator here at work, so give some suggestions as to what I should verify and I will try to get it sorted.

The main thing at this price point is being able to see wiggly lines on screen properly and not have any deal-breaking fails/annoyances.

I'd go to the extremes of horizontal scale and see how badly it aliases high/low frequencies. See if you can get it to display correct-looking waves with completely wrong frequencies.

How good is the FFT? I have a cheapo device with an awesome FFT (much better than the FFT on low end Rigols/Siglents).

[Kosmic]

One of the problem of the FNIRSI-5012H was software triggering + slow sampling rate. Hopefully this is now fixed with the newly added FPGA ? in any case I would make sure to test triggering on slow or non repetitive waveforms.

[boggis the cat]

Cool. I don't expect it to be perfect at this price point but is there anything you'd say was a deal breaker?

Nothing like that, for the cost.  It is 'feature poor', compared to a 'real' 'scope, but for the price...

How's the touch screen? Good update rate? How easy it is to set horizontal/vertical scale? Does it crash and need rebooting?

Good update rate, seems responsive.  No crashes yet, but the AUTO SET function can be a bit hit or miss.  (But it is fast, so just tap it more times.)

Are you "This is OK...", or "What a pile of garbage!"?

So far, it seems usable.

The biggest issue may be that 50 mV/div lowest amplitude setting.

That wouldn't worry me at all. "About 200mV" is fine.

(and yeah, 2% is never going to happen with an 8-bit ADC on a device like this)

Sure.  If you know how inaccurate it is, then it's less of a problem.

Sooooo annoying if it doesn't power on to a known state (preferably the same as power-off).

Sorry: it does power-up as last set.  No way to store another setup, though.

It will be obvious if it doesn't have it: Zoom in on a rising edge and look for Gibbs phenomenon, ie. a sinc will show ringing before the signal starts to rise.

Probably not, then.  (See attachments.)

The main thing at this price point is being able to see wiggly lines on screen properly and not have any deal-breaking fails/annoyances.

I'd go to the extremes of horizontal scale and see how badly it aliases high/low frequencies. See if you can get it to display correct-looking waves with completely wrong frequencies.

How good is the FFT? I have a cheapo device with an awesome FFT (much better than the FFT on low end Rigols/Siglents).

FFT included in attachments.  It seems viable for simple requirements.  Update rate is fast.

Attachments:

Pic 1: 200 mVpp @ 1 kHz (centred)
Pic 2: 200 mVpp @ 100 kHz (centred)
Pic 3: 3 Vpp @ 50 kHz sine (centred) – horizontal jitter roughly ± 0.1 div
Pic 4: 3 Vpp @ 50 kHz sine (centred) – trigger shifted low
Pic 5: 3 Vpp @ 50 kHz sine (centred) – trigger shifted high
Pic 6 & 7: 3 Vpp @ 50 MHz sine (centred) – amplitude ~2.3<->2.5 Vpp
Pic 8: 3 Vpp @ 10 MHz sine (centred) – stable amplitude (becomes significantly unstable around 20 MHz)
Pic 9: 3 Vpp @ 10 Hz sine (centred) – roll mode (automatic - cannot be selected)
Pic 10: 3 Vpp @ 50 kHz sine (centred) – aliased signal (AUTO SET not used, but shows aliasing is possible)
Pic 11: 200 mVpp @ 1 kHz (centred) – FFT
Pic 12-14: 20 Vpp @ 1 kHz (centred) – FFT, shifting horizontal scale

Notes on pictures and waveforms saved.

This is fast and seems to work well – but does not show the FFT in the thumbnail.  (No idea why not.)  There is no way to organise them; last saved is at top left.


Edit: forgot to add CH impedance characteristics as follows:

CH1: 1.015-1.016 MOhm // 30.0-30.1 pF
CH2: 1.012-1.013 MOhm // 30.5 pF.

[Fungus]

FFT included in attachments.  It seems viable for simple requirements.  Update rate is fast.

Definitely something weird in the FFT.

Attachments:

Pic 1: 200 mVpp @ 1 kHz (centred)
Pic 2: 200 mVpp @ 100 kHz (centred)
Pic 3: 3 Vpp @ 50 kHz sine (centred) – horizontal jitter roughly ± 0.1 div
Pic 4: 3 Vpp @ 50 kHz sine (centred) – trigger shifted low
Pic 5: 3 Vpp @ 50 kHz sine (centred) – trigger shifted high
Pic 6 & 7: 3 Vpp @ 50 MHz sine (centred) – amplitude ~2.3<->2.5 Vpp
Pic 8: 3 Vpp @ 10 MHz sine (centred) – stable amplitude (becomes significantly unstable around 20 MHz)
Pic 9: 3 Vpp @ 10 Hz sine (centred) – roll mode (automatic - cannot be selected)
Pic 10: 3 Vpp @ 50 kHz sine (centred) – aliased signal (AUTO SET not used, but shows aliasing is possible)
Pic 11: 200 mVpp @ 1 kHz (centred) – FFT
Pic 12-14: 20 Vpp @ 1 kHz (centred) – FFT, shifting horizontal scale

Thanks for all those.

How do you adjust the horizontal/vertical? Does it use gestures like pinch-zoom, etc.?

[gfmucci]

What are your thoughts about ease of use of the interface, especially compared to traditional knob and button layouts?  Simpler?  Confusing?  More or fewer steps to accomplish intent?  How useful is the manual for experienced?  For noobs? 

[tunk]

Would it be possible to test it with higher frequency square waves, e.g. 10-100MHz?

[boggis the cat]

Definitely something weird in the FFT.

Yeah.  I suspect it’s working on the (full?) bitmap.  Not sure how useful it is.  I noticed it looks like it overlaps / interleaves when the signal becomes less ‘noisey’, so I tried to capture that behaviour.

How do you adjust the horizontal/vertical? Does it use gestures like pinch-zoom, etc.?

Horizontal (which took me a while to figure out) is just tapping left or right side on the display.  Vertical is by the CTRL item; opens a menu with V+ / V- items for each channel.

Functional enough, if a bit clunky.

[boggis the cat]

What are your thoughts about ease of use of the interface, especially compared to traditional knob and button layouts?  Simpler?  Confusing?

For people used to tapping phone screens, probably just as easy to use.  Cost is likely lower for capacitive touch screens now.  The screen is quite responsive.  Edit: seems to be single point / single touch only — no pinch type behaviour.  (I believe this is cheaper.)

More or fewer steps to accomplish intent?

It’s very basic.  Think of it as the equivalent of a simple, 3.5 digit multimeter.  The ease of capturing screenshots and waveforms is a positive.  Battery life of four hours seems feasible, too.

How useful is the manual for experienced?  For noobs?

Utterly hopeless.  But this is a pretty basic instrument, so I’m not sure if a decent manual would add much.

[boggis the cat]

Would it be possible to test it with higher frequency square waves, e.g. 10-100MHz?

The Wavetek calibrator tops out at 100 kHz for amplitude-controlled square waves.  I was more focused on the vertical (in)accuracy, as the timebase seemed reasonably accurate up until the trace became unusable (jitter / trigger issues render the trace ineffective by around 50 MHz).

I can check the higher frequencies with a nominal square wave using the timebase function later today.

It looks like the effective bandwidth is way less than 100 MHz, but I can check using the supplied probes as the manual appears to claim that you need to use the 10x probe setting to achieve the claimed bandwidth.

Jitter starts getting significant past 10 MHz for sine waves, so the triggering is a bit dubious.  You have the option of a set 50% trigger (in a settings menu) or can manually move the trigger point.  It certainly isn’t as stable as a decent ‘scope, but would be fine for gross estimations of signal presence and approximate shape.

[Fungus]

the manual appears to claim that you need to use the 10x probe setting to achieve the claimed bandwidth.

That's true on all 'scopes.

Jitter starts getting significant past 10 MHz for sine waves, so the triggering is a bit dubious.

This is much more worrying. 10Mhz is in the "Arduino range".

[Fungus]

I just watched the video on this page, I think it gives a pretty good idea of what it does and how it does it.

https://www.aliexpress.com/item/4000861098295.html

[boggis the cat]

the manual appears to claim that you need to use the 10x probe setting to achieve the claimed bandwidth.

That's true on all 'scopes.

Not with an oscilloscope calibrator.  Or at least I have not found this to ever be the case.

The bandwidth is determined using direct connection from an Active Head.  (Usually 3 Vpp @ 50 kHz as reference level, so I did that here, but didn't bother to adjust for an accurate 3 Vpp reference indication because this isn't really a 'performance' instrument, so meh...)

So this instrument seems a bit odd, if it only functions correctly with the probe.

Jitter starts getting significant past 10 MHz for sine waves, so the triggering is a bit dubious.

This is much more worrying. 10Mhz is in the "Arduino range".

There isn't a lot in the front end.  A gradual deterioration in stability seems to occur, somewhere past 10 MHz for a sine wave.  Because this starts at 50 mV/div, perhaps it would perform better with a higher amplitude than 3 Vpp.

[tautech]

the manual appears to claim that you need to use the 10x probe setting to achieve the claimed bandwidth.

That's true on all 'scopes.

Not with an oscilloscope calibrator.  Or at least I have not found this to ever be the case.

The bandwidth is determined using direct connection from an Active Head.  (Usually 3 Vpp @ 50 kHz as reference level, so I did that here, but didn't bother to adjust for an accurate 3 Vpp reference indication because this isn't really a 'performance' instrument, so meh...)

So this instrument seems a bit odd, if it only functions correctly with the probe.

Jitter starts getting significant past 10 MHz for sine waves, so the triggering is a bit dubious.

This is much more worrying. 10Mhz is in the "Arduino range".

There isn't a lot in the front end.  A gradual deterioration in stability seems to occur, somewhere past 10 MHz for a sine wave.  Because this starts at 50 mV/div, perhaps it would perform better with a higher amplitude than 3 Vpp.

So its amplitude rolloff is not consistent through the V/div ranges and BNC cable connection produces different results ? 

[maginnovision]

Sounds like it's still a toy scope, not a serious instrument, and the bandwidth is still a lie. As expected.

[gfmucci]

Sounds like it's still a toy scope, not a serious instrument, and the bandwidth is still a lie. As expected.

Is the Hantek linked below a less enough of a "toy" to be worth $70 more?
https://www.amazon.com/Hantek-DSO5072P-Digital-Oscilloscope-Bandwidth/dp/B00RJPXB6Y/ref=sxin_12_ac_m_pm?ac_md=5-1-QmV0d2VlbiAkMTAwIGFuZCAkNDAw-ac_d_pm&cv_ct_cx=oscilloscope&dchild=1&keywords=oscilloscope&pd_rd_i=B00RJPXB6Y&pd_rd_r=a2fcc83f-174f-4d8b-8b42-a9c39cadd7d8&pd_rd_w=Z1GDp&pd_rd_wg=4wViO&pf_rd_p=14719209-e059-46b3-be1a-859d16f01f7e&pf_rd_r=ZQK2WZNQ2XPTRBDYA52V&psc=1&qid=1590459641&sprefix=oscillo&sr=1-2-3f74d940-526b-436f-a83d-f486577edf4d

[maginnovision]

Yes, that's not a toy. It's a real oscilloscope and while its specs aren't very good at least they're honest and I'm reasonably certain it can do whatever they claim.

[boggis the cat]

Some more results.

15_c.png through 25_c.png are a 1 Vpp sine wave at settings from 20 ns/div (50 Mhz) through to 10 ns/div (100 MHz) in -1 ns/div steps.  Notice that the 100 Mhz trace has essentially died.  But...

26_c.png and 27_c.png are an 11 ns/div then 10 ns/div trace.  This was done by slowly ramping from 20 ns/div down to 11 ns/div, then 10.9 ns/div down to 10.0 ns/div.  So, a stable 100 MHz trace – if you can slide onto it gently...   

28_c.png and 29_c.png are part of a PAL TV signal check at white and mid-grey levels respectively.  It shows the cursors in use.  There is a "move fast" / "move slow" setting (right of the s/div) that alters the sensitivity of dragging movements.  "move fast" is too fast, while "move slow" is frustratingly slow.   

[boggis the cat]

So its amplitude rolloff is not consistent through the V/div ranges and BNC cable connection produces different results ? 

It appears to be a weird hybrid / fast logger design to me, repurposed through software to give 'scope-like functionality.

My positive takes are: it is cheap, claimed to handle 400 Vpeak directly (yeah... I could check, but...), responsive UI, makes taking screen / waveform captures easy (but limited options if you want to do anything on the device), and it does yield a sensible(ish) trace up to 50 MHz+.

Negative: not really a handheld 'scope, IMO.

I think it is more of a "basic 'scope" / pretty inaccurate DMM hybrid.  (And I suspect it's really a data logger.)


Should you buy it?

Well, will it do what you would want it for?  Should be useful to quickly probe something before cranking up the real 'scope, as an alternative to the cheaper not-quite 'scopes, or for certain 'niche' uses.  I intend on adding it to my gear for the Airsoft club, where it could be handy to check for battery/motor type faults where a DMM doesn't tell you enough.

[tautech]

So its amplitude rolloff is not consistent through the V/div ranges and BNC cable connection produces different results ? 

It appears to be a weird hybrid / fast logger design to me, repurposed through software to give 'scope-like functionality.

My positive takes are: it is cheap, claimed to handle 400 Vpeak directly (yeah... I could check, but...), responsive UI, makes taking screen / waveform captures easy (but limited options if you want to do anything on the device), and it does yield a sensible(ish) trace up to 50 MHz+.

Negative: not really a handheld 'scope, IMO.

I think it is more of a "basic 'scope" / pretty inaccurate DMM hybrid.  (And I suspect it's really a data logger.)


Should you buy it?

Well, will it do what you would want it for?  Should be useful to quickly probe something before cranking up the real 'scope, as an alternative to the cheaper not-quite 'scopes, or for certain 'niche' uses.  I intend on adding it to my gear for the Airsoft club, where it could be handy to check for battery/motor type faults where a DMM doesn't tell you enough.

Cool, thanks for your checks and the GUI looks not too bad but shame it's not a real 100 MHz scope.

[Fungus]

Sounds like it's still a toy scope, not a serious instrument, and the bandwidth is still a lie. As expected.

Anybody could have told you that just from the price, no need to read the thread or get snobbish over it.

[gfmucci]

Sounds like it's still a toy scope, not a serious instrument, and the bandwidth is still a lie. As expected.

Anybody could have told you that just from the price, no need to read the thread or get snobbish over it.

From some of these comments it appears there is little "modulation" of opinion.  If it isn't a scope that is on target with the specs or they haven't mastered the use of yet or it isn't an instrument that they would feel comfortable using, it's a "toy."  I see the word "toy"  (or "shit" from Dave to emphasize his frustration with something) thrown around quite often. 

It would be more helpful to describe what the "scope" would be good or not so good at doing along with the extent to which it misses its stated specs. I might have missed the definition of "toy", which apparently is "Any device which misses its stated spec or behaves in a manner in which the operator does not expect or doesn't meet his own specialized need."

Every measuring/diagnostic device has its best uses and limitations and requires significant user familiarity with its best useage, quirks and anomalies.

[Fungus]

Every measuring/diagnostic device has its best uses and limitations and requires significant user familiarity with its best useage, quirks and anomalies.

No arguments there.

Compared to what's gone before, this seems like a big step towards being a contender for being a "real" 'scope. The touch screen is a good feature and seems intuitive enough, I'd have no problem buying one if I needed something cheap/portable and I knew I was going to work at frequencies I knew it could handle.