Agree these bench meters are on the very low end $ side, which obviously means compromises. If we were using these, getting a set of known good stable valued reference capacitors (maybe resistors also) covering a broad range would be in order. Having these "reference caps" measured with a quality LCR meter would give more confidence when using theses ET meters, and other lower cost meters.
As shown earlier we've developed a direct connect BNC means (no cables or other fixtures) for this very purpose. These have shown excellent long term (>1yr) stability and repeatability with Mica and PP types caps from 2pF to 1uF ranges. We just finished a build of special a Soviet 10nF Mica as shown here. This Soviet mica cap is from batches others have been involved with, and are very old, but have demonstrated excellent stability and temperature TC (within over office/home lab). Apparently the Soviet Mica is some of the best mica available, these caps certainly show that!! Here you can see the enclosed Soviet 10nF Mica directly attached to LCR meter, at 100kHz, 2Vrms, and DUT V at 1.692V and DUT current at 10.64ma.
Best
I'm quite interested in the results. Since the references appear to be in the US, it would be nice if you could borrow them to a trusted ET user in your area, let him measure them, measure them with a high-confidence meter and publish both results here. That's not a request, but I think this could be an interesting distribution for several readers in this forum.
Actually we have a Tonghui TH2830 and Hioki IM3536 LCR Meters that we use often, both acquired new and both agree very well over a broad range of components and frequencies. If you search our related threads we've developed and modified a few fixtures to enable better measurements with SMD low and high Z components, as well as creating a DC isolation fixture for adding external DC bias to various measurements (beyond the normal LCR meter capability).
If someone want's to send us a reference cap(s) for measurement, or wants to acquire a reference cap(s) like we've shown, we'll be happy to measure such, just PM if interested.
BTW with these special fixtures and internal ref caps, the repeatability we are discussing is out to 6 places. Not metrology grade, but good enough for our meger needs
Best
Since the Open and Short Calibrations are created right at the Male BNC connectors in the fixtures panel we've shown, almost anything that moves the test measurement away from the front panel BNC will suffer more uncertainty due to the extension. Same goes for the reference caps as shown.
A given reference cap should perform better overall when directly connected to the 4 LCR Meter BNCs, of course the amount of degradation for remote connection will depend upon the extension length, cables, fixtures, additional connectors, bends and so on. Quality cables and connectors can achieve similar results, however these cables/connectors become expensive relative to the simple direct BNC method shown.
Recall someone in Japan was also doing this direct BNC connect reference capacitor fixture, and remember their results were also quite good and repeatable.
Common engineering sense hints that eliminating additional cables/connectors should produce better repeatability, it certainly has in our case with the meager reference capacitors we have.
Best
I'm quite interested in the results. Since the references appear to be in the US, it would be nice if you could borrow them to a trusted ET user in your area, let him measure them, measure them with a high-confidence meter and publish both results here. That's not a request, but I think this could be an interesting distribution for several readers in this forum.
IMHO, that would not be a reliable comparison. The result would be valid only for that particular ET meter. There were very suspicious results earlier in the thread about the ET meter manufacturer calibration accuracy, (if at all, at least for some particular devices, see also post #817). Guess why I have two of them
. This meter is maybe the best for home/hobby use in the price range. The firmware flash difficulties and 30ohm impedance problems are very minor issues. Compared to if you buy a super expensive LCR meter only to realize that it also has it's issues and limitations.
I'm quite interested in the results. Since the references appear to be in the US, it would be nice if you could borrow them to a trusted ET user in your area, let him measure them, measure them with a high-confidence meter and publish both results here. That's not a request, but I think this could be an interesting distribution for several readers in this forum.
IMHO, that would not be a reliable comparison. The result would be valid only for that particular ET meter. There were very suspicious results earlier in the thread about the ET meter manufacturer calibration accuracy, (if at all, at least for some particular devices, see also post #817). Guess why I have two of them . This meter is maybe the best for home/hobby use in the price range. The firmware flash difficulties and 30ohm impedance problems are very minor issues. Compared to if you buy a super expensive LCR meter only to realize that it also has it's issues and limitations.
A good approach would be similar to what folks are doing with the precision voltage reference, pass this around from one to another by means of postal service/UPS/FedX/DHL or whatever, while recording and posting measurements. Having a set of reference capacitors to pass around is easier as they don't need external power like the voltage reference.
Best
Yes, I read about your script, but since I don't have any idea of python or programming, hoped to avoid it.
When I decided to write my own firmware updater, doing so in Python felt like the natural choice to me. After all, the repository I added it to was a Python class for controlling the device. I underestimated the hurdle of getting Python and any dependencies up and running for people who are not into programming. After some feedback along these lines, I decided to re-write the firmware update tool in a compiled language so I can provide binaries for LINUX and Windows without hassle. As I had been planning on learning Rust for a while, I took this as an opportunity to get going. But I have to admit, that I did start from an AI translation of my original Python script and spent some time tweaking that until it worked. So it may not be fully ready, but I have successfully flashed my meter on LINUX.
Binaries for Windows (untested) and LINUX can be found in the latest release of the repo
https://github.com/philpagel/ET44.py/releases/latest/.
Source code and instructions are here
https://github.com/philpagel/ET44.py/tree/main/tools/fwupdaterI would be grateful for feedback – especially for the Windows version.
Update: my firmware update tool now does some basic sanity checks on hex files before uploading them. See repo for binaries.
Just updated mine with your tool (for Windows).. Bootloader is indeed stubborn, succeeded at the 5th try, but when it starts, it goes all the way... Mine is updated now.
Thank for your feedback! that helps a lot!
Bootloader is indeed stubborn, succeeded at the 5th try, but when it starts, it goes all the way... Mine is updated now.
I spent some more time playing with my program after your comment and I believe I have solved the temperamental bootloader problem. So far, each time I had sent the magic number, I checked for a response and went on to read and parse the menu as soon as the interface had non-zero bytes to receive. After some experimentation, I found that adding a short sleep once the instrument responded to the magic number seems to make the trigger reliable. In my hands I have a 100% success rate now. Let's see if others observe the same now.
I spent some more time playing with my program after your comment and I believe I have solved the temperamental bootloader problem. So far, each time I had sent the magic number, I checked for a response and went on to read and parse the menu as soon as the interface had non-zero bytes to receive. After some experimentation, I found that adding a short sleep once the instrument responded to the magic number seems to make the trigger reliable. In my hands I have a 100% success rate now. Let's see if others observe the same now.
If there ever will be a new firmware for this unit, I'll try with the new version and give you the results!
Any at all. As soon as the chip detects that you are trying to read flash it will block the read, but if you try any "tricks" it will erase flash. Don't ask how I found out...
So without a hack/glitching/.... , there is no way to extract the image.
I have an old version of the VICTOR 4090B. after flashing the panel firmware back, the boot screen changed to the East Tester logo instead of the original VICTOR logo,
That is really cool! Can I have these dumps, please? And do you know the firmware version?
Indeed, I would be interested, too!
That is really cool! Can I have these dumps, please? And do you know the firmware version?
This firmware is for personal learning purposes only and is not meant for general updates. Other users, Due to major hardware differences, please do not update
Thank you! As far as i remember, you wrote earlier that you were able to read the front-panel as well as the main-pcb firmware. What you've uploaded seems to be the front-panel-firmware, only and the description seems to be reduced. Has the readout of the main-pcb-µC also been successful?
Thank you! As far as i remember, you wrote earlier that you were able to read the front-panel as well as the main-pcb firmware. What you've uploaded seems to be the front-panel-firmware, only and the description seems to be reduced. Has the readout of the main-pcb-µC also been successful?
MFW is the main-pfb firmware for st32f103rbt6
FFW for st32f103vct6
I see, MFW was not visible, when I first saw the message. Thank you!
After some issues with the communication (
the serial chip was broken), I've updated to latest firmware with Phil's tools (on Linux). Thanks!