Zoyi ZT-XT1 Transistor tester / multimeter

[uncle_sem]

Have you performed a calibration after the FW update?

hmmmm.... interesting idea, I'll try it later...

[kae64]

The service sent me the firmware - XT1_ZOYI_108_113, but my transistor tester doesn't initialize; it shows version 0,0,0. On XT1_ZOYI_110_116, everything seems to work except for the automatic DC voltage measurement mode.

[kae64]

Of the downloaded firmware, could you tell me if the XT1_ZOYI_108_113 firmware was fully installed? My multimeter installs version 1.0.8, but the transistor tester doesn't, showing version 0.0.0. The XT1_ZOYI_110_116 firmware is fully installed, but the DC voltage auto measurement mode behaves oddly: up to 10 volts, it shows erroneous values, but above that, it measures correctly. When changing registers, it measures correctly.

[jebem]

Did someone checked the video below?
Apparently the ZT-XT1 is not able to identify high power transistors like the PNP 2SA1302 or the NPN 2SC3281 identifying them as a pair of diodes.
The reviewer then used his trusted atlas DCA55 to confirm that the transistors are in good shape.

Most probably the ZT-XT1 needs to be further tested at the factory and bugs corrected.
The reviewer conjecture was that the ZT-XT1 is using very low test current when compared to the atlas  DCA55, and that combined with larger transistor internal capacitance may not be able to provide correct bias.

[jebem]

It was very hard for me to find a picture of the front side of the PCB assembly.
Checked one of the popular Chinese tech sites and found one.
I am attaching it here, I hope our Chinese friends will not mind.

[coromonadalix]

el cheapo transistors and component testers will do a better job,   and they have better support  loll

[@]

I have started a project for an alternative, open-source firmware. It's hosted on Github: https://github.com/sw/zt-xt1-dmm


At the moment it doesn't really offer an advantage over the official firmware, but hopefully that can change.

I have attached the binary if you want to try the current state.  zt-xt1.zip (143.44 kB - downloaded 20 times.) You can always flash Zotek's firmware from http://www.szzotek.com/h-col-159.html later. But you use this at your own risk.

Note that with their latest release (XT1_ZOYI_110_116.bin), Zotek have activated the flash protection on the N32G031 (component tester) MCU. That means that every time you upgrade *their* firmware, you ideally have to do a calibration of the component tester, because the calibration values get erased - don't blame that on me! It's unfortunate and silly that they have done that, as they're still publishing the firmware images in unencrypted form. It should however still work without the calibration procedure, it'll just use the default values.

My firmware is mostly working ok for the DMM part. That was actually not too difficult because the (proprietary?) DM1109CEN chip does most of the work. It outputs the measurement as 7-segment digit data over UART. I haven't found any obvious way to get more out of it, for example higher update rate or different ranges.

As for the component tester part, this is quite tricky as the ADC in the N32G031doesn't seem that well-behaved. It's rather poorly specified, there's an erratum, but even taking that into account, it takes a bit of fiddling with delays and sampling time.

So at the moment this is more of a learning project for me than anything useful for the world, but I'm publishing it nonetheless.

I used this as an opportunity to learn:

• reverse-engineering using Ghidra
• how the component tester actually works
• using the LVGL framework for building the UI. A big advantage is the ability to run the UI on the development machine:
• unit-testing the component detection functions using Ngspice to simulate various components.

The documentation is a bit sparse at the moment, I hope to improve on that. Also note that the source code is structured based on the reverse-engineering of the original firmware, so it's not the most elegant.

I probably will try to fix the broken hFE measurement for Darlingtons next.

[@]

Had a look at the mentioned power BJTs. While I don't have the devices at hand, I can run the spice model. It seems that these fail the measurement of the base-emitter voltage which is expected to be between 0.5V and 0.9V. Both the NPN and PNP are around 0.45V, because the collector current is so low (~1mA). So these are not classified as BJTs, but later in the procedure the two diode junctions are detected.

[kae64]

I have started a project for an alternative, open-source firmware. It's hosted on Github: https://github.com/sw/zt-xt1-dmm
(Attachment Link)

At the moment it doesn't really offer an advantage over the official firmware, but hopefully that can change.

Thank you very much.
Your firmware helped me fix a glitch: the firmware component responsible for the component tester wouldn't install. Also, your firmware doesn't have an option to access the settings menu, so maybe that's just the issue with my glitchy unit.

[SammysHP]

Thanks a lot for your work on trying to reverse engineer and improve the firmware! @ajar171 did something similar with the ZT-703 (until he disappeared from the forum), although his version was developed completely from the ground. I think Zoyi makes simple and relatively inexpensive, yet still capable hardware and just needs a little help with their firmware. Too bad they don't open-source it.

Anyway, I'm in contact with Zoyi regularly, and because Zoyi isn't actively monitoring this forum, I offered to collect bugs, complaints and feature requests for them. In addition to things already mentioned here, do you have any additional feedback?

PS: I won't tell them about the reverse engineering as I fear they might lock it down and required signed builds.

[@]

Also, your firmware doesn't have an option to access the settings menu, so maybe that's just the issue with my glitchy unit.

I haven't implemented the settings menu, as I didn't see a need for it, personally. The calibration procedure for the component tester is also missing, I might do that some time.

In addition to things already mentioned here, do you have any additional feedback?

Nothing else has come up so far. But with the already discovered bugs and the so-so performance as a multimeter (missing low current ranges), I wouldn't really recommend people buying it. Better get a ZT303 and an ATmega-based component tester.

Anyway, I had a go at improving Darlington measurements and also relax the checks for power BJTs. The latest version is on Github and attached as a binary.

[jebem]

Just received my ZT-XT1. Had a drief test of the component tester part.

While it can identify some power TRIACs that my old trusty "Multi-function Tester - TC1" can't, the new ZT-XT1 fails to identify some of the small P-FET and N-FET transistors I have in stock,  namely the 2SJ176 and the BF245, seeing them as resistors.

Well, let us hope the manufacturer is reading this forum and take advantage of our free testing service to debug the firmware.

[@]

Zotek seem to have broken the JFET detection in version 110_116. Unfortunately a direct downgrade to from 110_116 to 108_113 is not possible.

If you're feeling adventurous, I've attached the latest version of my alternative firmware  zt-xt1.zip (141.55 kB - downloaded 16 times.). That has the old JFET detection logic. That is also not perfect, but I need to read up on JFET theory before I can try to improve it. I don't really have any experience with them. Just using a J109 for testing, which gets identified correctly.

Once you install my firmware, it would then be possible to install 108_113, which I've also attached.  XT1_ZOYI_108_113.zip (86.68 kB - downloaded 23 times.)

[jebem]

Zotek seem to have broken the JFET detection in version 110_116. Unfortunately a direct downgrade to from 110_116 to 108_113 is not possible.

If you're feeling adventurous, I've attached the latest version of my alternative firmware (Attachment Link) . That has the old JFET detection logic. That is also not perfect, but I need to read up on JFET theory before I can try to improve it. I don't really have any experience with them. Just using a J109 for testing, which gets identified correctly.

Once you install my firmware, it would then be possible to install 108_113, which I've also attached. (Attachment Link)

Thank you for your work.
I may try your firmware as soon as I have the chance.

Concerning the jfet identification issues, the J109 and the BF245 are very different N-JFET devices.

They have different pinout on the same TO-92 package.

While the J109 is a switching device with 85pF high capacitance, 40mA high current, 12ohm low RDS(on), 1MHz cutoff, 16mSiemens high transconductance, Vgs(off) of -2V to -6V;

The BF245 is a RF device with 3pF low capacitance, 6.5mA low current, 500ohm high RDS(on), 700MHz cutoff, 6mS transconductance, Vgs(off) of -0.5V to -8V.

[SammysHP]

FYI: I forwarded a summary of all the issues we identified here to Zoyi. Let's see if they are at least trying to improve it.

[jebem]

When doing repairs, I always use a DMM for semiconductors or a LCR for passive components to check them.
That is my first choice, not a Component Tester.

A Component Tester is handy to have additional information on semiconductors, like pinout and device type, specially when the markings are obscured or are SMD coded.

Data sheets are absolutely essential, but sometimes it is very hard to find a datasheet for the markings we see on the component.

Additionally, many datasheets are incomplete, showing just static parms and no dynamic graphs.

Even when the datasheet is comprehensive, that does not tell where the component we are checking is behaving in that way.
Neither the Component Tester.

Unfortunately these Component Testers do not test the semiconductors using the proper static and dynamic bias settings.
How could they?
They identify the component type (if we are lucky), but they use a limited set of bias settings per type, not the real life adequate bias specific per semiconductor reference.

So, a Curve Tracer is the way to go if we need to see the real behaviour of the component.

Bottom line, I like the XT1 2 in 1 approach, as I need a DMM and then a component tester when doing repairs.
Having to use two separate equipments on the workbench when I can do the job with just one, I opt for the XT1.

[jebem]

FYI: I forwarded a summary of all the issues we identified here to Zoyi. Let's see if they are at least trying to improve it.

Thank you very much for your work on this subject.

I have contacted ZOYI support by mail reporting my two issues found with the 2SJ176 and BF245 JFET components that my TC1 can easily correctly identify.

Also I asked them for a firmware update and where exactly could I download it.
Their response was that all equipment leave factory with the latest version, not clarifying what was that latest version. But they provided the exact IP address to their repository.

Now, the strange thing here is that my unit came with version "Main:1.0.8 Tester:1.1.3", but for some strange reason, on their repository this version is not available.
Instead they offer another more recent version "XT1 Firmware Version 110-116" that have some fixes including some type of JFET components like the 2N5457.

From the experience from others members we know that the above 110-116 is even worse than the shipped factory version "Main:1.0.8 Tester:1.1.3".
So, I asked ZOYI support once again for clarification and got no answer so far.

[SammysHP]

Version 108-113 was shared earlier here in the thread:

https://www.eevblog.com/forum/testgear/ziyi-zt-xt1-dual-transistor-tester-multimeter/msg6270760/#msg6270760

[jebem]

Here we go for another random testing on small TO-92 or alike encapsulated semiconductors.

Equipment under test:
E0: "ZOYI ZT-XT1" Firmware M:1.0.8 T:1.1.3

Equipment for comparison:
E1: "FNIRSI LCR-P1" and
E2: "Multi-function Tester - TC1"

Identification errors in this test:

1. BS208 - P-channel enhancement mode vertical D-MOS transistor, https://bg-electronics.de/datenblaetter/Transistoren/BS208.pdf
E0 see a Transistor(NPN).
E1 and E2 agrees.
(My DIY curve tracer also agrees)

2. 2N5457 - General Purpose N−Channel − Depletion JFET, https://www.onsemi.com/pdf/datasheet/2n5457-d.pdf
E0 see a Resistor.
E1 and E2 see a N-JFET but D/S pinout is incorrect on E2.

3. BF245 - JFET VHF/UHF Amplifiers N–Channel — Depletion, https://www.onsemi.com/download/data-sheet/pdf/bf245a-d.pdf (previouly reported here)
E0 see a Resistor.
E1 and E2 see a N-JFET but D/S pinout is incorrect on both.

4. J176 - P-channel silicon symmetrical field-effect transistor (Drain and source are interchangeable), https://www.nxp.com/docs/en/data-sheet/J174_175_176_177_CNV.pdf
E0 see a Resistor.
E1 and E2 see a P-JFET but E1 gets confused on D/S and assigns pin 1 to both.



But not is all bad in this XT1.
It can identify most of my transistors, and not only the plain NPN/PNP bjt ones. For instance:

5. 2N5064 Thyristor
E0: Perfect identification.
E1: See a NPN with hfe 636
E2: See a NPN with hfe 16

6. SK152 - N-Channel JFET (S/D pins are interchangeable)
E0, E1, E2: Perfect identification, but E0 S/D assignment differ from E1 and E2.

7. BS170 - Field Effect Transistor N-Channel, Enhancement Mode
E0, E1, E2: Perfect identification as a MOSFET(N-MOS)

8. MPF102 - JFET VHF Amplifier N−Channel − Depletion
E0, E1, E2: Perfect identification.


Just an observation:
I'm starting to see a pattern here. The components that are failing identification, show their transfer curves with very low current on my DIY curve tracer, meaning the curves appear overlapped at the bottom side of the graph, as the curve tracer is programmed for 0 to 50mA in the vertical axis. I would need to modify the programm to use a scale of 0 to 5mA to be able to see the curves.

[indman]

Here we go for another random testing on small TO-92 or alike encapsulated semiconductors.

You need to have at least one tester with an original author's firmware instead of a Chinese one in order to correctly assess the capabilities of these toys in a beautiful case.

[uncle_sem]

original authors firmware also had a bugs, most of them was fixed by community, and I think author do not supporting this device anymore

[indman]

original authors firmware also had a bugs, most of them was fixed by community, and I think author do not supporting this device anymore

You are mistaken, one of the authors madires and today supports his version of m-firmware. The other author k-firmware is not so active lately, but his firmware version is still relevant today and does not contain such stupid and childish errors as the Chinese one.

[liutas]

And where can I find their software?
Thanks

[indman]

liutas
https://github.com/madires/Transistortester-Warehouse

[liutas]

liutas
https://github.com/madires/Transistortester-Warehouse

When I opened your link, I didn't find anything related to this topic. „Zoyi ZT-XT1