Well, I did something impulsive

[slburris]

Time will tell if I did something impulsive *and* stupid. :-)

Having just gotten paid, I decided it was time to finally get a better
handheld scope to replace an ancient Velleman HPS5.  The HPS5
samples at a whopping 500Khz, so it's only useful for the most basic
of things.

So I went out and ordered a Hantek DSO-1060, a 150Ms/sec
scope with a 60Mhz bandwidth.  The folks over at RCgroups said
nice things about thefamousbrandsoutlet.com, so in spite of the
name, I ordered from them.  Excellent communication on their part,
and the scope arrived last night, just as they said it would.

http://www.hantek.com/english/produce_list.asp?unid=79

I haven't had time to do much other than calibrate a scope probe
and so some basic tests of the multimeter.  Using the calibration signal,
the trace looks a little noisy, but I'm not sure if that the calibration signal
itself, the scope, or if this is what to expect with an 8-bit ADC.  Certainly
it's head and shoulders above the old Velleman unit.

Some reports of a related model, the DSO-8060 have said that it becomes
more voltage inaccurate with signals above 10Mhz.

Lots of testing ahead of me.

Scott

[slburris]

I took it apart, at least until I got stuck.

First, you have to remove the battery.  It's a 7.4v 3.5Ah unit.
Next, look at the ends.  There are 4 roughly rectangular covers
that need to be pried off.  You'll find 2 screws under each cover.
After you take those out, 8 screws totals, you can remove the
rear cover.

I tired to remove this PCB and failed.  There is a screw underneath
the BNC panel and I couldn't find a way to get access to it.  But I was able
to peek underneath.  The backside of this PCB is packed with parts, but
I couldn't get a good enough look at it.  If I end up breaking my scope,
then I'll make a bigger effort.


There is also a second PCB which handles the LCD and switches underneath
this one.

Looking at the PCB we can see, you'll see the usual ICs that have been scratched
off.  I think we're looking primarily at the multimeter section of the scope on
this side.  Also the PCB looks well made, the multimeter section is not
impressive.  There's little in the way of input protection.  Yes, they
are fuses, but they are standard, not HRC.  Note that the 10 amp current
shunt runs right past one of the fuse holders.  Someone has thoughtfully
put some heat shrink on the shunt so it doesn't short out.  I'd rather the
shunt had been placed father away so this would be an issue.

Would I use the multimeter to measure the mains?  Probably not.  But it should
be OK for measuring low voltage circuits.  In any case, I think of the multimeter
as a bonus, not as a required feature of this scope.

So I put the scope back together, and it still powers on :-)

Next up, I'll do some functional testing.  PCB photo in next post.

Scott

Th

[slburris]

Here's a photo of the PCB.

Scott

[Simon]

photo missing,

you will get the noise, consider the amount of pixels the screen has, then consider that the 256 levels have to be interpolated to the vertical amount of pixels that actually shows the waveform, then think that any ADC is accurate to +/-1+ bits and you will see that you will be getting at least +/-1 pixel of "noise" needless to say any real noise

[slburris]

Well, my first attempt to test out the scope was a mess.
Why?  I thought I saw a problem where as the frequency
increased, the amplitude of the displayed waveform
would change (someone in RCgroups reported this).

Well, I thought I was doing a valid test with my trusty
WaveTek 288 signal generator, but I double checked what
I was getting with my Tektronix 2236 scope, and saw
the same behavior on that scope too.

So, I think my WaveTek isn't outputing a constant amplitude
as I increase the frequency.  It's supposed to, so there
must be something wrong with it.  So now I have to
fix another piece of test equipment in order to test
some test equipment!

I've got an HP 3336 generator somewhere.  If I can get that
working, and it behaves well, I'll use that for the test.
Otherwise, it's off to fix the 288, then back to the test.

Scott

[NiHaoMike]

How about just check the relative amplitudes?

[Simon]

yea,  compare it to the scope you already have

[slburris]

yea,  compare it to the scope you already have

As far as I can tell by eyeballing it, the scopes show identical waveforms at
various frequencies.  I think I'll warm up the Rigol, then I can post some screen
captures comparing the Rigol to the Hantek.

Scott

[Mechatrommer]

i wish you can show me picture on ampere graphing, i never saw that, pls give me a chance at least once in my life plsss.

[saturation]

Thanks for the review and tear down.  All you need do is pass a high frequency square wave to the scope and see how much distortion it has, that can qualitatively say a lot.  Then compare its basic Vdc and Vac accuracy, and that's it.  Everything else is optional but those are key features.

Even if you don't have a comparison against a know working scope, it would help.  You can compare the Vdc and Vac against your high end DMMs.

[slburris]

In the photos below, I put a 1Mhz square wave into both the Hantek and the Rigol scope
and set them both to measure Vpp.  The Hantek seems to consistently read a little
lower than the Rigol.

I didn't even have to read the manual for the Hantek, as they use essentially an identical
user interface to the Rigol.  Who *really* wrote the software for these scopes?

The Hantek seems much more optimistic about its frequency counter than the Rigol, but
if you ignore the last few digits, they are quite close.

Next, I'll post some 5 and 20Mhz tests.

Scott

[slburris]

Here's the same test for 5 and 20 Mhz.  I'm not sure if I believe my WaveTek 288 is
putting out a decent square wave at 20Mhz.  I looked at the waveform with my
Tektronix 2236, and it still looks mostly like a sine wave.  Both the Tek and the Rigol
are 100Mhz (well, the Rigol is modified).  I'd expect rounding of a 20Mhz signal, but
something that looks so much like a sine wave?

So I pulled out the best analog scope in my collection, a Tektronix 2465 300Mhz.
Indeed, the waveform still looks like a sine wave.  My conclusion is that my signal
generator isn't generating very good square waves at higher frequencies.

With that caveat, here are the photos.  Note that the Hantek seems to roll off
on its Vpp measurement vs the Rigol at 20Mhz.

Scott

[Mechatrommer]

so far i think the Hantek is good enough! any chance measuring Amperage? to see say amperage oscillation from mcu wakeup to sleep mode? how about battery life?

[saturation]

These are great shots.  The Wavetek is 20 MHz generator? so sq. waves at 20 MHz will put out harmonics above 180 MHz, it might be off spec. 

For a quick test, Wavetek can be expected to put out good sq waves at freq/9, or say ~ 10, to make it easy, reflecting the highest harmonic.  That would be just ~ 2 MHz sq waves.  At least its freq/5, at the 5th harmonic, or ~ 4 MHz. 

The Rigol shows a decent sq wave at 5 MHz from the Wavetek, and the Hantek is rolling off as expected as a 60 MHz scope, the ends of the wave represent about 45 MHz or so, it would be good to see where the Rigol alone begins to show roll off using sq. waves from the Wavetek.   My guess you can output near 10 MHz sq waves from the Wavetek before a roll off is visible.

The Hantek has good response!

[slburris]

The Wavetek 288 is a 20Mhz generator.  Internally it generates triangle waves.
There are shaped into sine waves or sent to a discrete comparator which switches
on the top and bottom of the triangle to convert to square waves. 

I guess I would have expected a 20Mhz square wave to look more like a square wave
on the 300Mhz scope, but maybe the output of the wavetek is having trouble passing
the higher harmonics?

Scott

[saturation]

Try comparing the Hantek and hacked Rigol to a 10MHz sq wave from the Wavetek.  I think you should see the Hantek really roll off on the corners and the same start to happen on the Rigol.  You can manually sweep up from 10Mhz and see how the changes occur from there.

The Wavetek 288 is a 20Mhz generator.  Internally it generates triangle waves.
There are shaped into sine waves or sent to a discrete comparator which switches
on the top and bottom of the triangle to convert to square waves. 

I guess I would have expected a 20Mhz square wave to look more like a square wave
on the 300Mhz scope, but maybe the output of the wavetek is having trouble passing
the higher harmonics?

Scott

[slburris]

Here are the results for the Hantek and Rigol with a 10Mhz square wave.

I discovered that the Hantek doesn't like FAT formatted USB sticks, so
those tiny 32MB sticks won't be useful with it.  It likes FAT32 formatted
ones just fine.  The Rigol doesn't seem to care.

shafri: I will do a current test, I just haven't had time yet.

Scott

[slburris]

Just for fun, I pulled a 50Mhz TTL crystal oscillator out of my stock,
a Fox Electronics F110E series part.

Attached are the Hantek and Rigol response to this output.

Scott

[saturation]

Excellent, that is a near textbook response.   However, we can't tell if the limitation is due to the Wavetek or the Rigol, regardless, both are good with sine waves to 100MHz, since the Hantek is clearly rolling off while the Rigol still looks great. 

Here are the results for the Hantek and Rigol with a 10Mhz square wave.

I discovered that the Hantek doesn't like FAT formatted USB sticks, so
those tiny 32MB sticks won't be useful with it.  It likes FAT32 formatted
ones just fine.  The Rigol doesn't seem to care.

shafri: I will do a current test, I just haven't had time yet.

Scott

[saturation]

Its certainly doing well for both, except the Rigol seems to have some minor distortion at the peaks.  This probably is reflected on the sq wave tests, there is some overshoot at the edges.  This could be however, be from the probes.  Check the probe compensation, or switch probes to see if the Rigol waveform improves. 

Just for fun, I pulled a 50Mhz TTL crystal oscillator out of my stock,
a Fox Electronics F110E series part.

Attached are the Hantek and Rigol response to this output.

Scott

[Nounours18200]

Hi slburis,

You have written :

"Using the calibration signal, the trace looks a little noisy, but I'm not sure if that the calibration signal itself, the scope, or if this is what to expect with an 8-bit ADC.  Certainly it's head and shoulders bove the old Velleman unit."

did you make your final opinion on this point ? I ask you this question because on my DSO8060 I also noticed a noisy signal when I measured a squred signal generated by the DSO8060 itself.

Thanks for your reply ,
Nounours

[Nounours18200]

You have written :

"Using the calibration signal, the trace looks a little noisy, but I'm not sure if that the calibration signal itself, the scope, or if this is what to expect with an 8-bit ADC.  Certainly it's head and shoulders bove the old Velleman unit."

did you make your final opinion on this point ? I ask you this question because on my DSO8060 I also noticed a noisy signal when I measured a squred signal generated by the DSO8060 itself.

Thanks for your reply ,
Nounours

[KTy]

Any updates on 1060 accuracy / noise / distortion ?

I also have 8060 (they share same hardware except signal generator) and I am facing some offset + interference issue, even after calibration 
See here: https://www.eevblog.com/forum/testgear/hantek-dso-8060-noisy-channels/

Thanks,