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

[Yuriy V.]

In my opinion, it looks more like some kind of problem with the built-in DC-DC power converter (overheating & microcracks).

Solved as I hope:

1. Resoldered IC DC-DC & his choke
2. Installed optional heatsink
3. Removed metal shilding grid (bekause it useless on my opinion, only going unit to overheating)
4. Made some holes on bottom side of case back

Consumption current with both channels 'ON' has down to ~~ 965 mA.

[electr_peter]

BNC connectors were suspect on UTG932 unit. Inner plastic shafts with holes had uneven lengths (two longer, one shorter). Expected higher quality, at least to have similar dimensions. Changed to proper BNCs as suggested by mnementh

To increase BW to 60 MHz, EEPROM does not have to resoldered, only 2 wires are needed for SCL/SDA as suggested by Dandymon

[Yuriy V.]

Solved as I hope:

A few days later the problem reappeared. Apparently, somewhere there are microcracks that have arisen due to overheating of the device or as a result of factory assembly defects. But there are some clarifications. Before failure, the current consumption begins to increase.

I can say one thing: I am very upset by this situation. Although this device is very good in its functionality, it has shown itself to be extremely unreliable in operational terms. I am thinking about what to do next: buy 932(e) again or buy something else. A critical requirement is the availability of synchronization of the oscilloscope when sweeping. But I have not found other generators with such synchronization in this price range.

[bffargo]

Then I bit the bullet and wrote the EEPROM...... and then I smiled!

Thanks to fremen67  for doing the initial discovery.

Thanks. I just did the same hack using an UNO; though I used a chip clip to avoid soldering anything. I dumped out all non-FF pages of data to the debugger and captured that just in case. I will compare it against the previously posted ones and see if there are any differences out of curiosity.

For some reason even though I only changed the one byte, it re-booted into Chinese. Finally found the menu option to flip it back.

For those that want a bit better formatted output from the dump() API, here is a quick hack to clean it up a bit:

Code: [Select]

void dump(int base_addr){
    for (long i=0;i<(rows*line_length);i+=line_length){
    if (i+base_addr<16)
      Serial.print("0");
    if (i+base_addr<256)
      Serial.print("0");
    if (i+base_addr<4096)
      Serial.print("0");
    Serial.print(i+base_addr, HEX);
    Serial.print(":\t");
    for (long j=0;j<line_length;j++){
      if (readEEPROM(j+base_addr+i)<16)
        Serial.print("0");
      Serial.print(readEEPROM(j+base_addr+i),HEX);
      Serial.print(" ");
      }
    Serial.print("\t");
    for (long j=0;j<line_length;j++){
      Serial.print((char)readEEPROM(j+base_addr+i));
      Serial.print("_");
      }
    Serial.println();
  }
}

It reports things without floating spacing now. Some examples (forum is eating some of the text now too)

Code: [Select]

0000: 55 54 47 39 4D 61 72 20 31 39 20 32 30 32 31 20 U_T_G_9_M_a_r_ _1_9_ _2_0_2_1_ _
0010: 31 37 3A 31 30 3A 30 31 00 00 00 00 00 00 00 00 1_7_:_1_0_:_0_1_
...
0100: 00 00 00 00 00 00 00 00 00 00 00 00 00 88 C3 40
...
1F00: 55 54 47 39 30 30 45 3B 2A 2A 3B 36 30 3B 31 2E U_T_G_9_0_0_E_;_*_*_;_6_0_;_1_._
1F10: 30 31 3B 31 2E 30 30 3B 0A 00 00 00 00 00 00 00 0_1_;_1_._0_0_;_
...
1F40: 77 77 77 2E 75 6E 69 2D 74 72 65 6E 64 2E 63 6F w_w_w_._u_n_i_-_t_r_e_n_d_._c_o_
1F50: 6D 0A 0A 0A 00 00 00 00 00 00 00 00 00 00 00 00 m_


[Mechatrommer]

just did a battery pack mod. not really stellar appearance and voltage ripple (1Vpp) well... better late than never, better something than nothing... since my old powerbank cant supply ~2A (it tripped), so i have to strip away the original board and old batteries, and total redo the circuit using available modules, i only reuse the enclosure. cheers.

[balnazzar]

Guys, I bought an UTG932 and I'm more or less happy with it, but I'd need to control (& program) it from a PC, so I'm trying to install the software.

Unsuccessfully, until now. Basically, the installer install a "Device Manager", which on startup probes for the device (that is connected to the computer with the bundled USB-B cable).
However, it hangs up and doesn't seem to be able to find it:

Any opinion? Thanks!

[Mechatrommer]

anyone knows where the internal arbitrary wave data is stored? i'm thinking if i can overwrite the internal wave with more usefull arbitrary signal.

[Mechatrommer]

3D Printed Handle for UNI-T UTG FG/AWG
if someone with 3d printer interested in printing the handle for it, but dont know how to make a model, attached is the zipped stl file.. it needs 3mm diameter (steel preferably) rod where you can slip the handle strip around. the rest, use imagination.

[dr_when]

I just bought a UTG962E from Amazon and returned it a day later after not being able to resolve the same problem. So sick of hardware with crappy software with drivers of a suspicious origin that just plain don't work. I tried to get it to work on Windows 11 with no luck. The manufacturer offers no useful support other than saying trying it on older Windows versions. I don't think so. I refuse to support crap like this. Too bad because the hardware looked good but I needed to create and load my own waveforms. Turns out my Analog Discovery 2 does that all just beautifully.

[electr_peter]

Noticed few issues with UTG962E. Setup - channel 1 terminated in 50 ohm, SYNC unterminated, triggering on SYNC.

Max SYNC output frequency is only 2MHz (could not reach higher with my setup). If channel frequency is >2MHz or modulated, then SYNC frequency drops drastically.

Channel 1 phase of sine wave is not stable w.r.t. SYNC output depending on frequency. Phase is stable only for "nice" frequencies 2MHz, 5MHz, 8MHz, 16MHz, 20Mhz, 25Mhz, 32Mhz, 40MHz, 50Mhz (plus some others at 2/5MHz multiples). Even 1Hz deviation from these frequencies results in either smeared phase or several quantized phases. For example, 24/30/48/60MHz shows three different phases, 55 MHz show 11 phases. Small frequency change from offset results in smeared view. This effect is for all frequency range.

FM modulation results in a drift w.r.t. SYNC output depending on FM frequency and deviation settings.

[nctnico]

I just bought a UTG962E from Amazon and returned it a day later after not being able to resolve the same problem. So sick of hardware with crappy software with drivers of a suspicious origin that just plain don't work. I tried to get it to work on Windows 11 with no luck. The manufacturer offers no useful support other than saying trying it on older Windows versions. I don't think so. I refuse to support crap like this. Too bad because the hardware looked good but I needed to create and load my own waveforms. Turns out my Analog Discovery 2 does that all just beautifully.

Even for high-end waveform generators you'll find the waveform editing software is mediocre at best. In the end you are supposed to use commercial (third party) software or create waveforms from your own programs / scripts. Waveform editing software is just a checkbox item.

[Markus2801A]

Has anybody figured out how its possible that arbitrary waveforms transferred via USB to the device stay saved even when powering off the device?

It is possible to create a waveform and upload it to the UTG962. But after powering off the device, its gone as already other user have mentioned this a few posts before:

Another surprise, the data uploaded is not persistent. No waveforms survives over power off.
...

So we need a solution to make those transferred waveforms to stay permanent within the memory of the device!

[ataradov]

There are no non-volatile memories in the device that are big enough to store waweforms. There is no solution to that.

[Mechatrommer]

thats why i asked? how the internal arbitrary waveforms are saved (possibly overwriting those?), is it embedded in firmware? or came from factory and stored "only they know" where?

[ataradov]

I would assume they are part of the main firmware. The MCU looks big enough to store that. Which, I guess can be reused, but there are no firmware files and I don't think we ever determined the MCU type. I tried to match a few common MCUs that have this package, but nothing matched. But I have not tried too hard.

Although there is a 1MByte FPGA configuration flash. This might actually have some part dedicated to the storage.

The interface between the MCU and FPGA board is slow, so eventually waveforms have to end up in the FPGA for playback. But there is also 32 MByte of SDRAM, and I see no reason for that other than storage of the user data.

[dophuc]

Hello everyone, has anyone tried to hack the UTG962 (or UTG932) to use the Bode plot function with the Siglent SDS1104X-E oscilloscope? I like the compactness of this UTG962, and it would be great to be able to use it with the 1104X-E

[Markus2801A]

There are no non-volatile memories in the device that are big enough to store waweforms. There is no solution to that.

  Oh no! Why isn't Uni-T able to provide a fixed Firmware and Software which is able to safe the Data in non volatile section of the device?

PS.: I throw in an other F-Gen from OWON newly released last year: DGE2000 Series Arbitrary Waveform Generator
http://www.owon.com.hk/products_dge2000_series_arbitrary_waveform_generator

[ataradov]

It is a very simple and cheap device, so some functionality is limited.

And it looks like those custom waveforms are very short, so they can easily be stored in the MCU internal flash. So, they potentially could release a new firmware that does this.

There is no "hack" that would enable this though, so I would work under the assumption that it will not happen and if saving the waveforms is critical, then consider other devices.

At the same time, what are the actual use cases for custom waveforms on a device this limited?

Also, I'm wrong on CPU being unknown. It was identified earlier in this thread as  Gigadevice GD32F207ZCT6 due to poor laser etching. And the firmware is apparently not protected. This still does not help a lot, since significant reverse engineering would be required. And it does not look like there is a lot of interest, as otherwise the device is fine.

[Markus2801A]

...
At the same time, what are the actual use cases for custom waveforms on a device this limited?

Tried to upload a "stairStep up", it has only save StairsUP and Down (Like Triangle) for a primitive curve tracer.
Also the hight of the steps adjustable would be nice :-)
In the attached fotos: I have used an FY6800DDS, which has saved a StairUp Waveform, to achieve the blue traced curves.
The other attached waveforms are from the Uni-T just for information.

[ataradov]

Actually, given that firmware is not locked, it it may be possible to at least look if standard waveforms are stored there. If so, it should be possible to permanently replace the internal ones with a new firmware.

There is no dump attached to the post that said it was unlocked though, so someone has to read it out.

[Mechatrommer]

...
At the same time, what are the actual use cases for custom waveforms on a device this limited?

Tried to upload a "stairStep up", it has only save StairsUP and Down (Like Triangle) for a primitive curve tracer.
Also the hight of the steps adjustable would be nice :-)
In the attached fotos: I have used an FY6800DDS, which has saved a StairUp Waveform, to achieve the blue traced curves.
The other attached waveforms are from the Uni-T just for information.

yes this is my thought when curve tracer thread using FG proposed by mawyatt, but it can still be worked with UTG900 using StairUD and Tri Up, just a redundant trace on stair Down part of signal and possibly introduces noises in the traces. i havent tried though. other applications using arbitrary function i think we can find workaround/hack from existing built in (internal) AWF in this little/cheap device. but it would be better if... possibility to permanent save in internal AW data using by just adding an addition SDK/API call in FW/PC SW without the need for HW change...

[Markus2801A]

... but it can still be worked with UTG900 using StairUD and Tri Up, just a redundant trace on stair Down part of signal and possibly introduces noises in the traces...

Exactly it worked with only one AWG (UTG962) see image below. The schematic has to be changed, it's flipped, but it works!

Hopefully someone who is more well educated than me, can make a new firmware ;-)

[kavehm]

Hello,
How many waveforms can be designed with the PC software and uploaded to the device?
Is there any limitation?

[wasedadoc]

Hello,
How many waveforms can be designed with the PC software and uploaded to the device?
Is there any limitation?

AIUI you can design as many as you like but the generator takes only one and puts it in RAM.  Power off and it is gone until you upload it or a different one again.

[kavehm]

Is it possible to save the waveforms designed in the device itself for later use?