This post is being written after completion of the project so it is more of a “design retrospective” than a piece meal walk though of a current design. Thanks for reading.
All of the design files for this project can be found at my github
The old uTracer
In 2012, I ran into the uTracer project, a compact vacuum tube test circuit that was available as a DIY kit. This tester kit was a combination of a PCB containing all of the hardware to generate all of the necessary signals for energizing the tubes and a software package that talks to the hardware and presents the data nicely on screen. The tester is very powerful and flexible, giving the user tons of options for gathering useful data.
At the time that the kit came out I was heavily involved in designing and building tube guitar amplifiers and I was gearing up to start my business Kraig Amplification so I jumped at the opportunity to have a tube tester on my equipment bench. Since the tester came as just a PCB kit with no enclosure, I had to throw something together to house the board along with all of the extra connections needed to wire everything up.
I was not particularly interested in spending a lot of time making the enclosure look pretty as this was just intended to be a piece of test equipment that only I was going to use. The resulting enclosure was drilled a bit rough and I just used a sharpie to create some “custom” artwork. Here is an image of the enclosure that I came up with:
The case was a plastic project box that I picked up at a local electronics shop. I installed both 9 and 8 pin tube sockets in the box as practically all tubes used in guitar amps fit in those and I did not really have a need for any other types.
The uTracer was originally intended to be run off of a 19V laptop power supply. Since my tester was going to be next to a bench power supply at all times I opted to skip buying a laptop supply just run flying leads that I could connect to my bench supply. Communication with my computer was done via a USB to UART cable that emulated a COM port on my PC. This was the recommended communication method from the uTracer maual so I just ran with that instead of doing something more fancy.
To connect the tube sockets to the tester inside of the enclosure I used a method that many other people who have built uTracers have employed. In the enclosure I installed two rows of banana jacks. One of these rows was connected to the pins on the tube sockets and the other row connected to the signal outputs on the uTracer. The signal connections on the uTracer could be patched from their output to whatever pins on the tube socket the user wanted by physically connecting the banana jacks with small patch cables that I built. This allowed for easy re-configuring of the tester to connect to a variety of tube pinouts.
A PLAN FOR A better TESTER
I used the tube tester for years in the state above and was certainly happy with the how it performed, however I always wished that I spent more time to build the unit better and make it look nicer. Cosmetics certainly do have a role to play in engineering and my old enclosure left a lot to desire. In addition, the method of patching connections from the uTracer signals to the tube sockets never sat well with me. Having wires dangle out of the box and crisscross around each other just to terminate back at the box does not seem to be very conducive to making accurate and low noise measurements. On these wires there are high current pulsing waveforms, high voltage anode and screen signals, and sensitive grid signals all in close proximity to each other. This just does not seem like the best solution and can be prone to issues.
One of the most discussed topics on vacuum tube electronics is lead dress and wire management for reducing noise and cross talk between adjacent wires. The method that I used for connection seems to break a lot of the commonly accepted rule for proper wiring of valve equipment. I certainly have plenty of experimental data to prove that this method is acceptable for the application, however I always felt that with careful design I could make a circuit that employs better electrical design practices and allows for tighter signal control (also I was interested in making the up to 400V signals not be easily accessible from the outside of the enclosure).
In June of 2018 I decided to kick off the project to completely gut and rethink my uTracer. To begin, I compiled a “wish list” of what I wanted to upgrade from the old uTracer. At the heart of the project was the uTracer kit which had certainly proven that it was capable of working well. The things that I wanted to change fit into either a category of cosmetic or ease of use.
Here is the wish list of things I wanted to implement in the new design:
- A more universal, safer, and controlled method for connecting signals from the uTracer hardware to the tube sockets
- Better signal routing
- A more professional looking enclosure with artwork
- integrated usb support – no need for an external USB-UART dongle if it could be included in the enclosure
- Single carrier board PCB for all of the i/o and signal routing electronics
Here’s what came of those requirements: a more compact, more cosmetic, easier-to-use tester. I have yet to select knobs for the device – as we will see when we get into the design, the spacing of the encoders was pretty much fixed and didn’t allow for a wide choice of knobs. I am still deciding how to fix this, and am even considering milling my own knobs. Until then, I’ll be using the device knob-less
A Quick Look
Here is a quick glance at the new design and its features.
- 9 seven segment displays one for each pin on the tube sockets
- 9 rotary encoders to select pin functions
- Relay Matrix for switching signals to each tube socket
- 7,8, and 9 pin tube sockets
- STM32 microcontroller for control
- Integrated usb-uart bridge IC
- Grid loupe circuitry for measuring grid currents and positive grid voltages
With this new enclosure, the user can select how the uTracer signals are sent to the tube sockets by using the rotary encoders. After turning on the unit, all 9 displays will show a line indicating that no connections to the tube sockets have been made.
To create a connection, the user can press the rotary encoder corresponding to the pin which they wants to make a connection. The display will begin to blink and display one of 7 characters. The user can scroll through these characters by rotating the encoder until they reach the signal that they want to connect to. The list of connections are:
- (A) Anode
- (S) Screen
- (C) Cathode
- (H) Heater 1
- (H.) Heater 2
- (g) Grid
- (–) Nothing
When the user has rotated to the selection they would like, they can press the encoder a second time to confirm the select. The uTracer carrier board will now automatically turn on the correct relays to route whatever signal was selected to the appropriate pin on the tube sockets. All three tube sockets are wired in parallel so the configuration on the display applies to all three sockets. After all the connections have been established, the uTracer software can be used to gather data just like it had with the old design.
In the image above the tester is configured for a standard 12AX7 style vacuum tube with the following connections
- Heater 1
- Heater 1 (Having two pins with the same connections just shorts them – in a 12AX7 pins 4 and 5 are connected for a 6.3V heater configuration.
- Heater 2
At this point the tester is ready to have a tube plugged in and to begin testing! I am not going to go into detail on the software side of the testing as there is plenty of information already written on the uTracer website.
Below are a handful of curves taken from 4 6v6 tubes. Both the anode and screen currents were gathered across 9 grid voltages ranging from 0 to -32 volts. The screen voltage was set to 400V for this test. Values like these are what you might expect to find in a 20 to 30W amp.
In the next post we will begin looking at the circuit and how the design evolved.