eTracer -- a next gen curve tracer for tubes

Attachments

  • etracer_kit.jpg
    etracer_kit.jpg
    747.7 KB · Views: 495
Kit arrived!

Ok, about 1 h after taking that photo most of the housing was assembled, and the eTracer board and the PSU were installed. I hooked it up to my computer and it connected to the software fine (eTracer software 2.12 running under Wine 3 on Ubuntu Linux). Looking good so far.

I have not yet built the top plate with the tube sockets, so I have not yet connected any tubes to the eTracer, and I can't comment on the performance yet. That will have to wait until after the holidays...

Questions and comments so far:
- The manual says that the GND of the USB connector might be in contact with the enclosure. It's a bit unclear if that is a good thing or not. I feel it would make sense to have the eTracer GND connected (or disonnected) from the enclosure in a more defined way.

- The manual also says that it's ok to connect the safety earth from the IEC block to the chassis. I don't agree with this. As far as I can tell, this is MANDATORY, not just ok.

- The chassis has a rather large hole for a cooling fan (which is not part of the standard eTracer configuration). It's easy to stick in a finger or other things that do not belong in there while the eTracer is connected to mains power. This is not safe, so I'll mount some kind of a grille or similar to avoid bad things from happening. It would be better to revise the chassis to use smaller holes to keep fingers etc. away from the eTracer internals.
 
Last edited:
I spent some more time in the workshop fussing around with the Molex connectors, the LEDs+wires, the safety bananas, and installation+wiring of the tube sockets. This took some time and patience, and I have not not finished all tube sockets, but my eTracer is working fine so far!

The configuration file I received from essues seems to have a slight zero offset. I guess I'll have to recalibrate my unit to get everything spot on.

Cool!
 

Attachments

  • Screenshot from 2018-05-01 19-18-56.png
    Screenshot from 2018-05-01 19-18-56.png
    190.1 KB · Views: 461
  • DSC07724.JPG
    DSC07724.JPG
    424 KB · Views: 435
  • KT66_GT_triode.png
    KT66_GT_triode.png
    35.7 KB · Views: 449
  • 45_Sylvania.png
    45_Sylvania.png
    36.6 KB · Views: 427
  • ECC81.png
    ECC81.png
    33.5 KB · Views: 162
Thanks for sharing your experience with us. Here are my opinions on the grounding issues.

As stated in the manual etracer PCB does not need any grounding arrangement to work properly. If the USB GND contacts the chassis it means the chassis will be at the same potential as the USB GND at the other end (a PC or a laptop). If the USB GND doesn't contact the chassis then the potential of the chassis is either equal to E if the E terminal of the AC mains is connected to the chassis or floating. People might not like like the idea of a floating chassis but that's how most of modern devices like iphone or Macbook works.

As to the E terminal on the AC mains. Connecting the E terminal to the chassis is only effective when the AC mains is properly grounded. In many regions or countries where two-prone AC plugs are common connecting the E terminal to the chassis does nothing. Even if the AC mains is properly grounded if the cable is only two-prone (eg. AC adapters commonly used for laptops) or three-prone with the earth pin remove then the grounding also fails. To sum up, grounding (tying E to chassis) is recommended but it might not be as useful as one might expect.
 
The configuration file I received from essues seems to have a slight zero offset. I guess I'll have to recalibrate my unit to get everything spot on.

It seems I messed up the calibration file when I copied it to my workshop computer. Once I replaced it with the original file from the essues email everything was fine.

The next step is to learn how to make my own tube calibration files (I have some 6H30 on the bench).

EDIT: ooops, I just realised there is a 6N30 file, which is 6H30.
 
As stated in the manual etracer PCB does not need any grounding arrangement to work properly. If the USB GND contacts the chassis it means the chassis will be at the same potential as the USB GND at the other end (a PC or a laptop). If the USB GND doesn't contact the chassis then the potential of the chassis is either equal to E if the E terminal of the AC mains is connected to the chassis or floating. People might not like like the idea of a floating chassis but that's how most of modern devices like iphone or Macbook works.

I am no expert on safety measures, but I guess if there is a three prong mains connector, the protective earth (PE) pin must be connected to the chassis, no matter if the PE is implemented in the house wiring. All conductive parts of a unit must be connected to the PE pin (there may be exceptions to this depending of the safety regulations in place, such as with "Class II" devices with double insulation).

If I wanted to make a firm and well defined connection from the eTracer PCB GND to the chassis (not via the whacky USB GND), where would be the right spot on the PCB to make this connection?
 
For the PCB ground you can use the mounting pin of J1 (The shield of the USB connector) and bring a wire with a solder tag to the mounting hole as indicated by the attached photo.

I just looked into this. Before soldering a wire to the USB shield I disconnected the USB cable from the PCB (J1), expecting that the GND-to-chassis connection would be broken now. It's not! It seems the PCB GND is connected to the chassis via the mount screws / holes. Is this intended? If so, that's a much better path to connect the PCB GND to the chassis than via the USB shield.

Some other questions:
  • I observed that the HV LEDs do not always work as expected. It seems they remain off if the HV is less than 75 V or so. That's a bit confusing.
  • I tried the new "Combo" feature of the v2.13 software. I like it very much, and it makes batch analyses much more convenient. However, the software forgets the directory setting after exiting the program. Next time the program is started, the setting is back to the default setting. It would be nice if the software would remember the setting (in the config file?).
  • Also, in view of the easy batch analysis, it would be great if users could set the data path to automatically use a dedicated directory for every tube type and analysis date. For example something like this: directory = path/to/my/etracer/measurements/#TUBETYPE/#YYYY#MM#DD (where #TUBETYPE = name of valve settings file).
 
Oh, and yet another question: can the eTracer deal with tubes that need fixed current rather than fixed voltage at their heaters? I am thinking of PCC88, but there may be others. I guess a "current regulator" could be implemented in the software by tracking the HTR current and adjusting the voltage accordingly, but I don't know.
 
I have no plan to do it as it might cause liability issue.

The mounting holes are surrounded by the ground plane and hence the screws might short the ground to the chasis if any of the ground copper is exposed (scratched paint, etc.).

The HV LEDs indicate HV hazard. It only light up when the HV voltages on the capacitors are higher than about 50V. I remember it is described in the PCB manual.

I am collecting users feedback and will improve the software gradually.

Chris
 
I have no plan to do it as it might cause liability issue.

I don't quite understand the liability thing, but I guess you have good reasons not to implement the heater current control in the software. So what's my best option? Use an external heater supply with current control?

The HV LEDs indicate HV hazard. It only light up when the HV voltages on the capacitors are higher than about 50V. I remember it is described in the PCB manual.

Sorry, I should have read the manual more carefully. I guess that if an LED is supposed to indicate a "hazard", it would be better to put a clear "DANGER!" label on the chassis next to the LED.

All the other LEDs are "on" if the corresponding voltage/function is active (HTR, NEGV, POWER). They are not related to hazards. Its confusing to have the HV LEDs following a different logic. Following the function of the other LEDs, the HV LEDs should be on if the HV voltage is active, no matter of the voltage level. Is it possible to change this somehow?
 
$400 vs $0.00

The etracer PCB was originally priced at USD$600. Many European buyers were hit by the high VAT and it was the reason I decided to charge for the software to bring the claimed value for the hardware down. The answer to your question is yes.
Chris

Hi Chris,

Well, in that case, if you don't mind, I will stick to this tracer
Thanks for response anyway.

Cheers, E.
 

Attachments

  • uTracer-front-S.JPG
    uTracer-front-S.JPG
    685 KB · Views: 274
The heater supply is based on a DC-DC converter which regulates the output voltage. I am not sure what will happen if the voltage setting changes rapidly based on output current. If you want you can change the heater voltage manually in the software until you get a current reading you desire.

the two HV LEDs are controlled by two comparators in the hardware. They are not controlled by the software.

I don't quite understand the liability thing, but I guess you have good reasons not to implement the heater current control in the software. So what's my best option? Use an external heater supply with current control?

Sorry, I should have read the manual more carefully. I guess that if an LED is supposed to indicate a "hazard", it would be better to put a clear "DANGER!" label on the chassis next to the LED.

All the other LEDs are "on" if the corresponding voltage/function is active (HTR, NEGV, POWER). They are not related to hazards. Its confusing to have the HV LEDs following a different logic. Following the function of the other LEDs, the HV LEDs should be on if the HV voltage is active, no matter of the voltage level. Is it possible to change this somehow?
 
current vs voltage control

I don't quite understand the liability thing, but I guess you have good reasons not to implement the heater current control in the software. So what's my best option? Use an external heater supply with current control?
[..]

Tubes are designed with a particular cathode or filament temperature in mind. So, eventually, it's the temperature that matters. Opposed to current control, this temperature is more precisely defined by means of a voltage controlled heater supply. This also applies to tubes designed for series connected filaments, of course.
 
Yes, most tubes are constructed such that the correct temperature is attained with the heater voltage as given in the data sheet. However, there are a few exceptions (such as the PCC88) that are designed to work at a certain heater current. This current is given in the data sheets of these tubes, and I guess it's usually best to use these tubes with the rated current.

I found that tubes from different batches or different manufacturers may require different heater voltages to establish the rated heater current. For occasional testing of current rated tubes it's easy enough to adjust the heater voltage until the correct current is established.