uTracer JS issue in uT3 running 13.4.1 Ventura OS

52 views
Skip to first unread message

Martin Manning

unread,
Aug 1, 2023, 1:33:57 PM8/1/23
to uTracer
I've observed some strange behavior where the "ping" test in the debug window fails to show the correct results:

Screenshot 2023-08-01 at 9.28.29 AM.png
All of the hardware tests, grid, anode and screen voltages, as well as current tests using 10k and 100k resistors A and S to k, look good, but when I attempt to trace a tube I don't get a current reading. Original R Dekker interface works fine, so I suspect the uTJS software has some issue with the recent update to the OS. Ihor?

Ihor

unread,
Aug 1, 2023, 2:04:51 PM8/1/23
to uTracer
Hi Martin, 

I am not sure about the Ping results, to me they look ok. The only useful numbers that Ping typically shows are Vsupply and Vnegative. Those should be correct and they are. The currents there are just some random numbers at the ADC because with the ping no actual voltages are applied. Ping only asks PIC to respond with values from the memory, without making any actual measurements. All the zeros are also fine, because in that Debug windows those are updated only during the measurements. 

Now just to clarify, so when you put resistors instead of a tube, then you can trace the curves (lines in this case), and when you put the actual tube, the whole acquisition does not work?  "No current readings" - does it mean that you get zeros or some noise, so flat horizontal lines for sweeping Va in the plots? 

Ihor

Martin Manning

unread,
Aug 1, 2023, 4:18:30 PM8/1/23
to uTracer
Hi Ihor, thanks for the reply. 

Here is the ping using the uT3 GUI. Vsup and Vneg are the same, only some small differences in other locations.

Yes, I can get straight lines using resistors, but when I trace a tube I get what you see below, essentially no current reading, 10-20 uA. I made that trace, and then booted over into Windows, and without touching anything, got the ping result below and made a successful trace of a 12AX7 using the uT3 GUI.

I was tracing curves using uTJS a couple of weeks ago with no problems, so the thing that comes to mind is the OS upgrade, which was recent.
uT3_GUI_Ping.PNG


Screenshot 2023-08-01 at 3.41.17 PM.png

Ihor

unread,
Aug 1, 2023, 4:37:44 PM8/1/23
to uTracer
Couple of things here:

- I think the ping is actually the same, in a way it looks good to me, but one thing I noticed is that the same values from uTracer are reported differently in uTracerJS and windows software. It means that the calibration values in both programs are not the same. Probably there is something with the setup of calibration values and sensing resistors(?)

- I am puzzled with the fact that it works with he resistors and does not work with a tube. I will try to test it on Thursday-Friday with my mac. 

Couple of possible tests meanwhile:
* is it possible then to test the uTracerJS in windows? There, there were no updates and things should work, so just to know if it is a mac/windows specific problem or the latest version of uTracerJS. 
* I see you are using "Vg_low"/grid loupe. Would it be possible to test that 12AX7 without the grid loupe, so in a standard configuration and see what happens. 

Ihor 


Ihor

unread,
Aug 1, 2023, 4:44:11 PM8/1/23
to uTracer
By the way, I am also using Ventura 13.4.1 but I just check for updates and there is already 13.5. I am not sure if it changes anything but I will upgrade to that one and do the testes later this week. 

Martin Manning

unread,
Aug 1, 2023, 4:51:08 PM8/1/23
to uTracer
Ihor wrote: "I see you are using "Vg_low"/grid loupe..."

<insert red-faced emoji here>

That's my problem. The Vg Low tick box escaped my notice. Somehow he setup was saved with that box sticked.

Ihor

unread,
Aug 2, 2023, 5:46:06 AM8/2/23
to uTracer
Good to hear! With that option this 0 or -1V of Vg were translated to actual -20 -30V on the grid, leading to deep cut off and those micro/milliamps:) 

I am still not sure why you get different IA and Is in the debug window for uTracerJS and Windows software, because Vglow should not influence that computation and the only cause should be different calibration values. 



Martin Manning

unread,
Aug 2, 2023, 7:25:40 AM8/2/23
to uTracer
I would not be at all surprised if the calibration values are different. I have been using uTJS exclusively for some time, and no doubt they have been tweaked. I made no  attempt to synchronize them on the OS and Windows platforms. This exercise did provide an opportunity to let MS update the Windows side of the house.

Thanks again for asking the right questions!

Reply all
Reply to author
Forward
0 new messages