UNI-T UTG932/UTG962 200MSa/s Function Arbitrary Waveform Generator

[Dakkahun]

I checked the documentation provided for the older and bigger brothers, in hope of seeing if they have described a protocol or something, but no such luck. 
I also sent an e-mail to Uni-T support asking about the protocol they are using. Let's see if they answer.

Regards.

[jbaribeault]

Yeah I did the same when I received mine - complete radio silence from UNI-T support so far...

[Dakkahun]

In the meantime, I was looking at the photos of the internals kindly provided by @ataradov and @Mechatrommer, did anybody investigate the pin headers present on both boards?

[mnementh]

THIS is new...

mnem
 

[Anthocyanina]

Hi!, I've been following this thread for a while before registering, i'm setting up my home lab since access to the uni lab is pretty much indefinitely closed, and when lookign for a function generator i came across this one, and the koolertron and feeltech ones. This one seems to be better built than the feeltech and koolertron ones, and its features also eem pretty good, but now i wonder what is the actual difference between something like this and other awgs that cost 3 or more times as much as this one? regarding performance. Thank you!

[ataradov]

The specs are better. You can just take the manuals for all of them and compare the stated numbers.

But really, the only thing you can note on this device is relatively slow rise/fall times on the square waves.

Others may have better remote control and may be better (or any at all) support, if that matters to you.

For most practical uses this generator is fine.

[Anthocyanina]

Oh, i see, i'll look up the datasheets of those and see what they do different, i think those rise and fall times should be ok for the low speed things i do. thank you!

[TurboTom]

My current go-to budget AWG is Rigol's DG811 after it's "liberated". In the class below that, this Uni-T appears to be the most decent choice but I haven't got first-hand experience with it so I haven't got a real comparison.

What I like about it is that it seems to be solidly built (I simply don't like gear that's got a flimsy, cheapish appearance), uses "real" DACs (none of the directly FPGA-driven, discrete R-2R networks of questionable accuracy), its design with a single, low voltage supply (so it can easily be powered by a USB power box), the presence of a numeric keapad (thoug it may be tiny), and it's general small, compact design that makes it a good choice as an "on-the-road" instrument.

I don't like too much the absence of dedicated sample memory, the apparent lack of decent PC software support (other instruments, even much more expensive ones, may not look much better here...), the way the transconductance amplifier is located directly after the DAC and not with the reconstruction filter in between (this opened Uni-T the chance to go with a single-ended reconstruction filter design and safe a few components at the cost of higher distortion), that it only has one AUX port for all "advanced functions" and it lacks a reference frequency input - that's probably the price one has to pay for the form factor and may be acceptable considering its main applications (there are "almost-high-end" instruments available that aren't equipped that much better...yes, Siglent, I'm talking about you  ), and maybe that Uni-T lost the opportunity to include a battery option within the instrument's casing.
Edit: During a quick fly-over of the manual, I noticed that the multi-function AUX input/output of the UTG9x2 has just digital capabilities. This means, no external analog modulation is available. This is really disappointing news. It excludes the use of the AWG as a wide-range VCO!

I'ld love to do a comparison of one of these Uni-T AWG vs. an entry level instrument a class above it (for axample the Rigol device that I mentioned), we may actually be up to a surprise (in either direction...) but I just can't be bothered to spend on another AWG that I won't really need   .

Anyway, if one is in the market for hobby grade AWG, and the available funds are hard-limited to round about 100 US, EU or UK bucks, this Uni-T device probably won't be the worst choice.

[JamesLynton]

Ah yes, the HP 54600 series scopes, one of the businesses I was involved with a little while back had loads of them and was rather fond of them.. decent machines which are oddly satisfying to use.

[rernexy]

Got my utg962 yesterday (Weekend delivery!). Problem is, Uni-t did not include an Australian socket adapter for their plug pack. So I'm powering it with a LiPo power bank I had lying around (comsol branded from Officeworks). Also don't have a 50R BNC terminator lying around, so 2 100R resistors in parallel are doing that job.

EDIT: This energizer USB power adapter could power it, it's current rating (2.4A) is higher than the uni-t supplied one. But then it wouldn't be portable.

[JamesLynton]

Took nearly a month, though had a personal reply from Uni-T on the much needed matter of saving configuration presets on the UTG962 - it is being passed onto the development team.
Hopefully sometime in the near future that feature will find itself onto these devices; shall see.

I may also poke them to suggest better implementing the USB bus commands for remote controlling the unit as well, in order to more cleanly automate things like bode plotting.
If anyone has suggestions on a preferred control instruction implementation, please let me know and I will pass it along.
Thanks.

[Dakkahun]

This is good news, my e-mail exchanges with them regarding the communication protocol and firmware upgrade capabilities have been less than satisfactory so far.

[mnementh]

      

https://www.eevblog.com/forum/testgear/replacement-knobs-feet-and-fittings-for-test-equipment/msg3393818/#msg3393818

I like mine well enough that I bothered to print up a wedge base for my 54645A so the UTG962 fits right there.



   

That said... I feel obligated to pass on a warning to my fellow enthusiasts aboot the horrible cheap-arse BNCs on these things (well, at least which came on mine); as you can see, the center contact is not the usual machined bit, but rather just rolled sheet metal, and not very carefully rolled at that.

I noticed this when I was tinkering with it during the photo session for my wedge base above; it had a finicky contact on one of the BNCs that would drop out if you wiggled the cable at all laterally. At first, I thought there was something wrong with the RG58/U cables I'd made up from leftovers. But then I noticed that it did the same thing with the brand-new factory-made BNC jumper which came with the thing, and always on the same BNC jack.

Further inspection revealed the horror seen above; but worse, inspection of the cable showed this connector was actually shaving metal off the center pin of those BNCs. I took many pics, but this was the best of them... unfortunately it doesn't well show the actual shaving of the metal which is visible to the eye under magnification. This is a brand-new BNC plug with maybe 6 mating cycles on it.

Yes, these things are built to a price; but I feel that having connectors which damage your expensive interconnect cables is definitely on the wrong side of the too fukkin' cheap line; be forewarned.

So... while this generator is a great value function-wise, I advise you not to use any cables or probes with it that you care aboot... and that you probably should consider replacing these little horrors with a set of decent BNC jacks as part of the cost of ownership on this little beast; I'm planning to add them to my next BOM.

It might be okay with cables using a stainless steel BNC or maybe with a decent stainless BNC right angle as a intermediary connector; but I've only ever seen those in 75Ω.

EDIT:   

Disassembly and measurement shows these are clones of Amphenol 031-5486; the version with solid pins in white Valox. These are ~$7 each at Mouser or DigiKey. Any of the above Low-Profile receptacles should be a suitable replacement, however.

Datasheet attached below.


mnem

[mawyatt]

Another option would be to place a quality BNC male to female adapter on the AWG outputs. Something like this.
https://www.l-com.com/Coaxial-Adapter-BNC-Male-Female?gclid=CjwKCAiAi_D_BRApEiwASslbJzHAXZeio_rmvaq7TTGiT2ggStb4EMVJnWhKYriNI8lE6rD2-e8oehoChz4QAvD_BwE

Cheap BNC and other adapters/cables cause endless headaches. Here's one I just found The end just pulled off when I removed the attached cable. I've got my fair share of these cheap items and need to send most to the trash can

Best,

[mnementh]

Yeah; that's what I was talking aboot when I suggested a decent stainless BNC right angle as a intermediary connector; but like I said, I've only ever seen those in 75Ω.

The usual right angle adaptor with gold-plated bronze pins in 50Ω version is of course a possible solution; but I can't count on it to fix the intermittent connection problem I'm having, nor can one be sure how long they will work with that hinky contact damaging the pin in the adaptor.

I get wanting to not void warranty; but at this price I for one am not going to be sending the UTG962 back to China, and those connectors are only $7 each in Amphenol brand, which is just aboot the same as the right angle adapter. I think it's better to bite the bullet and fix the problem rather than pay for fixing it two or more times with band-aid solutions.

Just my 2 pesos worth.

mnem
 

[Labrat101]

@  mnementh
Hi , nice base . I have the same scope HP54645A also I have added the FFT option pack on the back
These scopes will read very happily into the Ghz .. ( Don't tell every one or they will all want One ) 
You just had Vectors turned OFF . Nothing weird 
If you have the FFT option on the back of the scope will read will read even higher GHZ in FFT   

PS set the load to 50ohm 

[mnementh]

Oh, I figured it was just imperfect isolation between the two channels on a $99 Arb and the signals were beating together; not exactly unexpected. If you look at my first pic with the UTG962 above, you can see it displaying 5GHz, just letting it Autoscale. 

I haven't even scratched the surface of what this scope can do; it knows what I'm looking for before I do half the time. Honestly, I quite like using it better than the DS1054Zed I had for 2 years before I moved to the Great White North.

If you want to stop using your old meters as cribbing blocks , the STLs for my wedge base are over in the 3DP Feet & Knobs thread:

https://www.eevblog.com/forum/testgear/replacement-knobs-feet-and-fittings-for-test-equipment/msg3393818/#msg3393818

Cheers,

mnem
 

[Labrat101]

Oh, I figured it was just imperfect isolation between the two channels on a $99 Arb and the signals were beating together; not exactly unexpected. If you look at my first pic with the UTG962 above, you can see it displaying 5GHz, just letting it Autoscale. 

I saw the 5Ghz . my one was also flicking between 1 & 5ghz in auto . . If you use direct cable set to x1 not
auto . Yes these scope are great they just work and show what you want when you want .
 Without all the bells and BS of the Digital stuff . (Old school Rules OK)
When set with Vectors OFF it will except odd frequencies  on the channels
 its quite normal . your scope may need recalibrating . there is a switch on the back to unlock the memory
 then you need a RG link cable to Channel 1 and follow the on screen instructions .
 And you can use the UTG962 to check the square pulse when ask for.
Make sure the 962 is set for 50   and not High Z ( utility .. System menu ) 
The Bandwidth of this scope is way higher than the 200Msa/s on the front panel .
 I have pulse this scope from my Neobodnar GPSDO that goes from 400Hz to 810Mhz Cmos square .
 Without blowing any thing up 

O did you find the Easter Egg in the Utility/print menu
 Also a Game . enjoy

[mnementh]

Yup. Got mine from AllTest on US fleaBay: https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg3180550/#msg3180550

There's also a TETRIS game you can play on these; it's stored in the RS-232/Printer add-on board if you have it.

If you don't have it, I recommend snagging a copy of this 54645A resource as well: https://archive.org/details/HP54645NCLIPImgs



Now I have an excuse to get back into my 4070A; I received my order of AD811ANs from China last week, but I've been putting it off as other things higher priority. But I just noticed while looking through some other pics that it has those Amphenol 031-5486s as well; 6 of 'em in fact. If I can't resurrect the 4070A, it will become a BNC socket donor and I'll get half my investment in the thing back just from not buying them.

mnem
 

[Labrat101]

Thats only one Easter Egg . There are 2 more . One is Bugs press Utility/Print press the 2nd & 3rd buttons
and hold for a few seconds (Hardcopy & IO Buttons ) and the game will start.
Its Called BUGs   enjoy
The third is Little unknown and not Published as far as I know .
Its called the Jackal menu This give you access to the main op system and can change memory and lots of other stuff . etc  ""System Level ""
That is entered by holding the service menu button & the button to the left of it together (soft Buttons)
this will give a new menu .
Jackal Menu , Acquire Menu , Anolog Menu ,  DAC ON .
You can turn hidden options ON/Off as well ie Zooming enable . which is Off .
Make notes what you change . also you can reset the Zero line if its off . And about
anything else .
Be careful what you change .  I have not played with any of the setting as my scope is
fine . I did a Full recalibration about a year ago and its spot on .
Enjoy the Bug game 

Thanks for the link I will look at it. 

[mnementh]

Yeah, I knew aboot the System Level menu;   I think I read aboot it in that service documentation I linked to. I just don't volunteer that tidbit with people I don't know; at least until I have an idea whether they know enuf to not nuke their scope in 30 seconds flat. 

It is my understanding that BUGS/TETRIS is an either/or thing; which you get depends on which version of the RS-232/printer board you get. EDIT: Or maybe it's the FFT board...? As I don't have anything that can emulate a LaserJet4, hasn't really been a huge motivation to get the board. I keep my eyes open, but it has to be "for a steal" prices, if you know what I mean.

mnem
 

[ab5eu]

Per your musings on comparisons this document still seems to be current:

https://www.element14.com/community/groups/roadtest/blog/2020/05/17/tektronix-afg31052-verification-tests

Enjoy!
AB5EU

[redjr]

This is the third time I have had trouble with DHL they are the worst as they use other carriers.

This is location dependent, so use whatever carrier works best in your area. In my case DHL is the most reliable carrier.

Not really in my case it never left china and never got to the main distributor . UNI-T sent it out DHL collected and after 10 days
never got processed .
I had the same thing with DHL UK they lost a package at the airport .    It turned up 3 months later . I had already  got a refunded
and a resend via another Currier within 7 days .
 DHL has the worst track history and the best delay system .

That's not been my experience from Asia.  DHL to the US is the most reliable and cost effective.  JLCPCB pretty much uses it exclusively and orders get to my address in 7 days!  If more Chinese companies had logistics operation like JLCPCB, they'd be big winners in getting stuff to the US, and other worldwide locations.  Sadly, we have to deal with the likes of Ali-Express and Banggood, using their sorry a$$ China Speed Post.  More like the Pony Express!    I just ordered this FG from Ali-Express and requested DHL shipping.  We'll see.  Covid is still affecting everything - especially here in the USA with our dismal USPS.  Priority Mail used to be promised in 2-3 days.  Now it takes 2-3 weeks!  Hopefully, our new president will fix that issue in time with our sorry a$$ PostMaster General getting the boot.

[Adam Baum]

Is there a menu setting to select language?  If so, the "E" suffix claimed to signify "the English version" by China-based sellers who I trust as far as I can throw the nation of China doesn't make sense, especially since letter suffixes on UNI-T multimeters have significant meanings.  If there isn't a setting, maybe their claim is true.

[Labrat101]

Is there a menu setting to select language?  If so, the "E" suffix claimed to signify "the English version" by China-based sellers who I trust as far as I can throw the nation of China doesn't make sense, especially since letter suffixes on UNI-T multimeters have significant meanings.  If there isn't a setting, maybe their claim is true.

Yes.
Press Utility ..
Then press the 4rd button on the bottom row  ( System)
Then press 3rd button  (Language)
Then press the first button . (That's English  ).

Counting from Left is one ..

PS  "E" suffix  was the first series and its does mean English .

My one also came in Chinese used the Google translator on Phone