Cheap 12-bit, 200 Ms/s arb--any wisdom?

Hi, all,

In my quest for an arb meeting the above specs, I ran across the Hantek

3x25,
formatting link
.

It looks about right, allegedly reads CSV waveform files, and generally looks like good medicine if it works reasonably well.

Anybody have experience with it?

Thanks

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
 Click to see the full signature
Reply to
Phil Hobbs
Loading thread data ...

No experience, but is 4K samples enough for you? At 200 MHz that's only 20 microseconds. Every time I've needed an arb I needed many milliseconds of waveform, so even at much slower clocks I needed much more memory.

----- Regards, Carl Ijames

Hi, all,

In my quest for an arb meeting the above specs, I ran across the Hantek

3x25,
formatting link
.

It looks about right, allegedly reads CSV waveform files, and generally looks like good medicine if it works reasonably well.

Anybody have experience with it?

Thanks

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
 Click to see the full signature
Reply to
Carl Ijames

I'm eyeballing the cheapest one (125Ms/s though):

formatting link

Lecroy also sells them under their own name.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
 Click to see the full signature
Reply to
Nico Coesel

Not for absolutely everything, but my usual need is to generate some a noisy tone burst or something like that, rather than a data stream. Besides, all my scopes are Tektronix, so how many points can I use? ;)

Thing is, older arbs hold their value better than most boat anchors--Tek AWG2020s seem to run north of $2k, for instance.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
 Click to see the full signature
Reply to
Phil Hobbs

Phil,

Just be aware, usually you get what you pay for.

I bought a $500 Rigol DG1022 func. gen. a while back, because I didn't want to pay the $3000 for the Agilent that has all the features I want but don't really need.

Well, the Rigol is buggy as heck and has all sorts of glaring glitches and limitations that are not made clear beforehand from reading the specs.

Well, it's still not bad considering the price relative to the "brand names." But it is essentially a POS. It will probably suffice for 90% of what I want to do at home. Fix the toaster, etc. :-D

OTOH, the scope I bought for home was a full blown Agilent 3000X series, and I don't regret a penny spent on that. At least if the measuring instrument is good, I can confirm whether the source is delivering an appropriate signal in a given situation.

But if I can't trust the scope, then I'm running blind. So at this point I'd still not consider buying a cheap Chinese scope.

That said, the Agilent new func. gen. series, what is it,...33522A that I bought at work a year ago was ALSO buggy as heck!

But at least Agilent will listen to me and take my bugs and work on them within 1-2 fairly short firmware rev. cycles. They have even implemented entirely new features in their scopes thanks to my suggestions.

For folks with the energy, skill, and time to spare, buying older high quality test gear and fixing up minor problems is probably a better approach than buying new Chinese cheap stuff, if budgets won't permit just buying the known good.

--
_____________________
Mr.CRC
 Click to see the full signature
Reply to
Mr.CRC

Interesting, thanks. The sweet spot seems to be the 20 MHz one (SDG1020) which is only about $70 more than the 5 MHz one, but the 50 MHz one is almost twice the price. I wonder if there's a Rigol-style hack to turn the slow one into the fast one?

Either way, $360 for a two-channel 125 Ms/s, 14-bit arb with 16k points is pretty attractive, even with fairly poor specs.

I also wonder if the bottom 4 bits are all zeros, or random, or what. Even at DC-1 MHz, the distortion spec is -60 dBc, which is a bit less than 10 bits' worth, and at higher frequency it goes down to 35ish dB, which is an ENOB of less than 6 bits, even assuming that the harmonics completely dominate the output spur spectrum.

Somewhat depressingly, it seems that one gets what one pays for, at least in arbs. Maybe I'll go console myself by buying another scope. ;)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
 Click to see the full signature
Reply to
Phil Hobbs

Note that these also suffer from leaky capacitors like the early TDS500/600/700 series. Even if you buy a 'calibrated' or 'certified' one with waranty you might be in for a nasty surprise and find yourself repairing PCB traces to get it working again. They don't get sold for $2k according to Ebay BTW :-)

The LW400 series of Lecroy are much more usefull. At least those can swing 20V p-p into a high-Z load.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
 Click to see the full signature
Reply to
Nico Coesel

I have the Rigal 20Mhz unit and I do like it. The software for the PC could be a little better through.

As for doing a software switch to bump it up, I don't think that is possible with the 1022. Being it's a DDS technology, you have to expect some issues when trying to generate AB's up to 20 mhz which is most likely why it only does it up to 5 mhz.

There are a couple of things I do not like about it, that is the back light operation behaves not the way I would like it and it seems for some reason it does not like to save my last set up by default on start up, I have to save it and reload it when I power up again.

Other than that, I've found the unit does as advertised with better results in performance than what I would have expected for the price I paid.

I will say this, if you are getting into a DDS system, you may want to consider getting a unit that is capable 2 x above your main interest range of use is. DDS systems tend to show their limits when you start reaching for their ends in exotic wave shapes.

Jamie

Reply to
Jamie

At the time I bought my Instek scope it clearly left a comparable Tektronix (at 1.5x the price) in the dust. It was engineered in Taiwan and built in China. Over 6x the sample memory, blazingly fast storage, less noise, and most of all less EMI. Tek had blown a few things such as USB storage. I simply don't have the time to wait many seconds to store

5k worth of screen plot data.

So yeah, I'd certainly buy it again, and clients of mine have.

[...]
--
Regards, Joerg

http://www.analogconsultants.com/
Reply to
Joerg

Well I guess it really just depends, huh?

I haven't been happy with Tek scopes for a while, though.

--
_____________________
Mr.CRC
 Click to see the full signature
Reply to
Mr.CRC

only 25MHz.

Now I know why the ones we were using were $15k each. but they were true synthesizers too.

And only a quarter that size (sma of course)

Reply to
Chieftain of the Carpet Crawlers

much better...

formatting link

but we gotz pocketz, so...

Reply to
Chieftain of the Carpet Crawlers

It's high time the management at Tek (and others) wakes up and realizes that their lunch is being eaten, every day some more. I see a similar trend as with the big three auto makers decades ago. They laughed and scoffed about foreign competition, and then one day ... "Oh s..t!"

--
Regards, Joerg

http://www.analogconsultants.com/
Reply to
Joerg

Why would the OP want that when even the new mid-range Agilent would exceed his specs for 1/10 the price:

formatting link

--
_____________________
Mr.CRC
 Click to see the full signature
Reply to
Mr.CRC

formatting link

I got my Chinese arb generator for around 350$. Ok, "only" 20MHz and only a small alphanumeric display but I kept over 1600 bucks of difference. And it works.

--
Regards, Joerg

http://www.analogconsultants.com/
Reply to
Joerg

I just took a look at the Instek scopes. Too bad they didn't put deep memory in their GDS-3000 series.

Is your scope the GDS-1000A-U ?

The comparable Agilent with similar memory would be a DSOX3012A, for about 3.87* the price of Instek. For the added $$$ you'd get 2Gs/s, high screen resolution, a high-resolution mode, 250ps (!!!) peak detect mode, statistical analysis on measurements (does Instek have that), and options to upgrade to MSO, advanced serial bus triggers and decoders, segmented memory, higher bandwidth, etc.

For the money, the Instek is pretty good. I really wish they would have put deep memory on the GDS-3000!

For the GDS-3000 series, the agilent is about 56% more $$$, but has 40x the memory, and what really sells Agilent for me is:

1000000 waveforms/sec.

Do you know how many waveforms/sec the Instek can do? I've found that having this Agilent capability has made it possible to find very rare glitches in systems very quickly. Invaluable.

--
_____________________
Mr.CRC
 Click to see the full signature
Reply to
Mr.CRC

One thing that can matter is whether the sample rate generator is fixed or variable. There are few truly arbitrary sample rate arbs. except for the high-end.

Agilent 3352xA series offers :

Sample rate 1 µSa/s to 250 MSa/s, 1 µSa/s resolution

Nothing else under $2000 can do this.

--
_____________________
Mr.CRC
 Click to see the full signature
Reply to
Mr.CRC

Mine is the GDS-2204.

This was about three years ago and I canvassed the offerings of all major manufacturers thoroughly. Instek won, by a mile. It does have some useful window functions for pass/fail but no statistical analysis. However, it offers streaming and download into a PC and there you can analyze the data ad nauseam.

Another downside with the Instek is the paltry screen resolution. But since I run it tethered to a laptop most of the time it doesn't matter to me.

AFAICT it just keeps on revolving the sample memory at full tilt, of course with a 1GSPS limit. IOW if you select a sample memory length that is too slow for the occasion then the number of waveforms per second naturally goes does down unnecessarily.

My scope has a lot of fancy error catching modes that I rarely use but nice to know they are there. Overshoot, "pre-shoot", underwidth, out-of-spec delay, and so on. Then there is the qualifier mode where you can store numerous waveforms and have it catch the event when something deviates beyond pre-set limits. I think it stores 20 of those which I think is plenty. I guess one can always have another big repository on the PC.

In the beginning it was like driving a new luxury sedan. Every day I'd discover a new cool function. They also laid the whole command set bare so if anyone is inclined to do so (I am not ...) they can write their own remote operation software. I wrote to them asking if I could have the keys to the firmware ... nope ... I couldn't have those.

--
Regards, Joerg

http://www.analogconsultants.com/
Reply to
Joerg

Huh? Do you mean setting an otherwise constant sample rate to arbitary level, or changing the sample rate during the arb waveform? If it's the first, my $350 arb gen can do that. But it can't switch gears to another sample rate in the middle of the waveform. And I never had the urge to do so.

--
Regards, Joerg

http://www.analogconsultants.com/
Reply to
Joerg

While reading this thread, I wondered, why is such equipment so hugely expensive ?

Is it simply due to low production volumes or something else ?

A sample rate generator using NCO (Numerically Controlled Oscillator) should be able to generate frequencies at any sampling rate.

A SIMM/DIMM package (64-128 bits) can deliver quite rapidly 5-10 samples with 12 bits, especially executed in fast page mode or similar. Using sufficiently word lengths (64-1024), any data rates can be generated.

Thus the only expensive component would be the DAC, but at only 200 MHz, that would not be hugely expensive.

Reply to
upsidedown

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.