uTracer (Micro Curve Tracer)
Moderators: pompeiisneaks, Colossal
- martin manning
- Posts: 13355
- Joined: Sun Jul 06, 2008 12:43 am
- Location: 39°06' N 84°30' W
3 others liked this
Re: uTracer (Micro Curve Tracer)
Another update for the uTracer is in the works. Here is a small addition to test for shorts, H-K leakage, and heater continuity. It consists of an indicator circuit that transitions a bi-color LED from Green to Red when it sees resistance going from ~3M to 1.5M. It's a simple circuit that uses a BJT to change the LED color when a small leakage current flows through a voltage divider connected to the transistor's base. It’s powered from the raw negative voltage supply (-40V), and the cathode terminal (+20V). These voltages are present when power is on, so nearly 60V is available for the H to K leakage test.
My first thought was this circuit would be used exclusively to test for H-K leakage, where a mechanical switch would disconnect heater power from the DUT, and connect H and K to the indicator circuit. This would be done after the DUT is connected to the uTracer panel terminals, and prior to tracing/testing. It could also be used with the heater and cathode at temperature.
There is another way that it can be used, however, and that is to insert the DUT, switch the indicator circuit in, and use the patch cords to probe the pins of the DUT before it is connected to the uTracer terminals. For example, the cathode terminal is connected to the DUT pin 1 jack, and then a patch cord connected to either u-Tracer heater terminal is touched to the DUT pin 2-9 jacks. The cathode terminal connection is then moved to pin 2, and the heater terminal cord is touched to pin 3-9 jacks, etc., until all possible pin-pin connections have been checked. This will reveal shorts, H-K leakage, and verify heater continuity.
The footprint is kept very small, about one square inch, by mounting the PCB directly to the pins of a mini-toggle switch. I've been breadboard testing it, and should have it installed within a few days.
My first thought was this circuit would be used exclusively to test for H-K leakage, where a mechanical switch would disconnect heater power from the DUT, and connect H and K to the indicator circuit. This would be done after the DUT is connected to the uTracer panel terminals, and prior to tracing/testing. It could also be used with the heater and cathode at temperature.
There is another way that it can be used, however, and that is to insert the DUT, switch the indicator circuit in, and use the patch cords to probe the pins of the DUT before it is connected to the uTracer terminals. For example, the cathode terminal is connected to the DUT pin 1 jack, and then a patch cord connected to either u-Tracer heater terminal is touched to the DUT pin 2-9 jacks. The cathode terminal connection is then moved to pin 2, and the heater terminal cord is touched to pin 3-9 jacks, etc., until all possible pin-pin connections have been checked. This will reveal shorts, H-K leakage, and verify heater continuity.
The footprint is kept very small, about one square inch, by mounting the PCB directly to the pins of a mini-toggle switch. I've been breadboard testing it, and should have it installed within a few days.
You do not have the required permissions to view the files attached to this post.
Last edited by martin manning on Fri Jan 26, 2024 11:03 am, edited 2 times in total.
- martin manning
- Posts: 13355
- Joined: Sun Jul 06, 2008 12:43 am
- Location: 39°06' N 84°30' W
1 others liked this
Re: uTracer (Micro Curve Tracer)
Shorts and leakage test installed and working. Running out of room in the box ;^)
In the bottom photo, the 6L6GC has its heater and cathode connected. The leak test is switched in and showing green on the lower LED, indicating no H-K leakage (more than 3Meg resistance H to K). Touching H with one hand and K with the other will switch the LED to red due to the ~1Meg parallel resistance. Even though there is almost 60V potential I can't feel a thing with only 20-30 uA flowing. This circuit can be made to work with uT6, but note uT6 negative voltage supply is -85V, which would result in over 100V from H to K. The -15V supply could be used instead by changing two resistor values to adjust the LED current and the sensitivity. I have extra boards available in case anyone is interested.
In the bottom photo, the 6L6GC has its heater and cathode connected. The leak test is switched in and showing green on the lower LED, indicating no H-K leakage (more than 3Meg resistance H to K). Touching H with one hand and K with the other will switch the LED to red due to the ~1Meg parallel resistance. Even though there is almost 60V potential I can't feel a thing with only 20-30 uA flowing. This circuit can be made to work with uT6, but note uT6 negative voltage supply is -85V, which would result in over 100V from H to K. The -15V supply could be used instead by changing two resistor values to adjust the LED current and the sensitivity. I have extra boards available in case anyone is interested.
You do not have the required permissions to view the files attached to this post.
- martin manning
- Posts: 13355
- Joined: Sun Jul 06, 2008 12:43 am
- Location: 39°06' N 84°30' W
1 others liked this
Re: uTracer (Micro Curve Tracer)
Trying out a Java script interface for uTracers created by Ihor Smal, as described here: http://boffin.nl/wp/utracerjs/ For Mac-centric people, this is an excellent way to run the uTracer directly from a browser tab. No additional hardware required, and no booting into Windows. The interface is functionally very similar to the Windows UI from the designer, and it works for all versions of uTracer and auxiliary boards (Low/+ Vg). It supports QuickTtest and curve tracing, and has most if not all of the features of the Windows GUI, plus it seems to run a bit faster. There are some nice features that are not available in the original R Dekker UI, such as automatic SPICE model generation, compensation for the sometimes annoying test voltage drop, and the capability to download characteristic plots in several graphical formats (.png shown below). So far, it seems to be working very well, with only one minor glitch encountered in the plotting. Ihor originally posted about this UI here on TAG over a year ago https://ampgarage.com/forum/viewtopic.p ... 02#p427402, and has been improving it continuously.
You do not have the required permissions to view the files attached to this post.
Re: uTracer (Micro Curve Tracer)
The interface looks much improved, Martin! Can you increase the number of points to characterize the curves, or are you limited to a certain sample size?
- martin manning
- Posts: 13355
- Joined: Sun Jul 06, 2008 12:43 am
- Location: 39°06' N 84°30' W
1 others liked this
Re: uTracer (Micro Curve Tracer)
Yes, you can specify the number of points you want to acquire. You get smoother curves but it takes longer to get the full set. As I mentioned, it seems to be a bit faster than the original UI, but much faster for me since I can stay in MacOS.
- dorrisant
- Posts: 2636
- Joined: Tue Sep 21, 2010 1:27 pm
- Location: Somewhere between a river and a cornfield
- Contact:
Re: uTracer (Micro Curve Tracer)
No I am thinking I need a Mac as well.
"Education is what you're left with after you have forgotten what you have learned" - Enzo
- martin manning
- Posts: 13355
- Joined: Sun Jul 06, 2008 12:43 am
- Location: 39°06' N 84°30' W
- dorrisant
- Posts: 2636
- Joined: Tue Sep 21, 2010 1:27 pm
- Location: Somewhere between a river and a cornfield
- Contact:
Re: uTracer (Micro Curve Tracer)
I meant to say "now" instead of "no". This comment was based on the fact that the universe seems to be sending me signals to buy into the Mac world... This was just another thing that removes a roadblock, so to speak.martin manning wrote: ↑Sat Jul 23, 2022 4:55 pmNot sure what you mean by that... This JS UI runs from a browser, so it doesn't matter if you are on a Mac or Windows. I have Windows running in Bootcamp on my Mac, and that is how I've been running uTracer.
Great reporting as usual.
I am seriously considering selling my daily driver Hickok 605A. When the time arrives (near future) where I can suspend the near daily need for it, I most likely will sell it.
"Education is what you're left with after you have forgotten what you have learned" - Enzo
- pompeiisneaks
- Site Admin
- Posts: 4222
- Joined: Sat Jan 14, 2017 4:36 pm
- Location: Washington State, USA
- Contact:
Re: uTracer (Micro Curve Tracer)
If you want another person's input stay away from apple like the plague. I'm a software engineer/IT guy and have had to use them a lot and back in the day they were, by far, the best hardware and software out there. Somewhere after Steve Jobs died, the went to a company built on profit over quality instead. They horribly suck now IMHO and I'd avoid them as much as possible. Ask norburybrook he just moved off of apple hardware entirely for his recording studio due to his disgust with them as well.dorrisant wrote: ↑Sat Jul 23, 2022 7:06 pmI meant to say "now" instead of "no". This comment was based on the fact that the universe seems to be sending me signals to buy into the Mac world... This was just another thing that removes a roadblock, so to speak.martin manning wrote: ↑Sat Jul 23, 2022 4:55 pmNot sure what you mean by that... This JS UI runs from a browser, so it doesn't matter if you are on a Mac or Windows. I have Windows running in Bootcamp on my Mac, and that is how I've been running uTracer.
Great reporting as usual.
I am seriously considering selling my daily driver Hickok 605A. When the time arrives (near future) where I can suspend the near daily need for it, I most likely will sell it.
My 2c
~Phil
tUber Nerd!
- martin manning
- Posts: 13355
- Joined: Sun Jul 06, 2008 12:43 am
- Location: 39°06' N 84°30' W
Re: uTracer (Micro Curve Tracer)
It's neither here nor there, IMO. I deal with both platforms on a daily basis... Windows machine for work, and personal MacBook, and regularly in a hybrid way by running a Windows installation on my MacBook, and MS Office apps under MacOS.
- martin manning
- Posts: 13355
- Joined: Sun Jul 06, 2008 12:43 am
- Location: 39°06' N 84°30' W
Re: uTracer (Micro Curve Tracer)
The issue with the uTracerJS plotting I mentioned above seems to have been resolved, and a new beta test version is available. The problem turned out to be with proper resolution of conflicts between Pa max, the "compliance limit" (software max anode current), and max anode voltage for each grid curve.
Here's one last test of the JS interface in combination with the Low/+ Vg accessory, using the screen channel to measure grid current in a triode (Is is actually Ig). No problems observed, or expected really, since I’d already tried the Vg low mode, and the software doesn't know that grid current, as opposed to screen current, is being measured. It does need to be informed (via the "Vg low" tick box at top center) that the Low Vg accessory is being used to generate the grid voltage, and thankfully an annoying bug in the Windows UI (failure to load setup data when switching from a standard to Low/+ Vg) is gone. This tube is actually an ECC803S, so almost a 12AX7. As you can see, the grid current starts to take off when Vg reaches zero.
Here's one last test of the JS interface in combination with the Low/+ Vg accessory, using the screen channel to measure grid current in a triode (Is is actually Ig). No problems observed, or expected really, since I’d already tried the Vg low mode, and the software doesn't know that grid current, as opposed to screen current, is being measured. It does need to be informed (via the "Vg low" tick box at top center) that the Low Vg accessory is being used to generate the grid voltage, and thankfully an annoying bug in the Windows UI (failure to load setup data when switching from a standard to Low/+ Vg) is gone. This tube is actually an ECC803S, so almost a 12AX7. As you can see, the grid current starts to take off when Vg reaches zero.
You do not have the required permissions to view the files attached to this post.
Last edited by martin manning on Sun Jul 24, 2022 3:14 am, edited 1 time in total.
Re: uTracer (Micro Curve Tracer)
Hi Martin,martin manning wrote: ↑Sat Jul 23, 2022 4:55 pm This JS UI runs from a browser, so it doesn't matter if you are on a Mac or Windows. I have Windows running in Bootcamp on my Mac, and that is how I've been running uTracer.
So, to clarify, uTracer can now run native in MacOS, via a browser? I don't really want to deal with Bootcamp or other emulator.
Thanks,
Dave
- martin manning
- Posts: 13355
- Joined: Sun Jul 06, 2008 12:43 am
- Location: 39°06' N 84°30' W
Re: uTracer (Micro Curve Tracer)
Not exactly. Ihor has written an alternative (replacement) UI in Javascript/node.js that is run from a browser window, something that multiple operating systems can support. There are specific installation packages for macOS x64 (Intel), macOS arm64 (M1), windows x64 and linux x64. Running the app opens a terminal window, and then a browser window is pointed to local host. You control the uTracer hardware through the browser window, with the terminal session running in the background.
- pompeiisneaks
- Site Admin
- Posts: 4222
- Joined: Sat Jan 14, 2017 4:36 pm
- Location: Washington State, USA
- Contact:
Re: uTracer (Micro Curve Tracer)
Totally get it, I use windows as my main OS for gaming and the like and use linux for everything else. WIndows used to be more hated by me than macos, but that's swapped and windows 10 and 11 finally started not sucking enough to be usable againmartin manning wrote: ↑Sat Jul 23, 2022 10:32 pm It's neither here nor there, IMO. I deal with both platforms on a daily basis... Windows machine for work, and personal MacBook, and regularly in a hybrid way by running a Windows installation on my MacBook, and MS Office apps under MacOS.
~Phil
tUber Nerd!
- martin manning
- Posts: 13355
- Joined: Sun Jul 06, 2008 12:43 am
- Location: 39°06' N 84°30' W
Re: uTracer (Micro Curve Tracer)
A few recently built uTracer 3+ have experienced negative power supply failures as a result of a shorted BD138 transistor. It is thought that substandard part quality is the root cause, but an improvement to the circuit has been devised by a user in Germany. Two resistors are added to the negative voltage supply as shown below (in light brown), a 100 ohm to reduce reverse bias on the BD138's B-E junction when the transistor is turned off, and a 4k7 across the inductor to reduce ringing. There is an added bonus, as I am now seeing the boost converter operating a little more efficiently as evidenced by reduced idle current draw. My bench supply shows 0.25 A total current instead of 0.26 A, so something less than a 10 mA reduction occurred. My tracer has been operating for nine years without a failure of that transistor, but why not?
You do not have the required permissions to view the files attached to this post.