Building the Project

The construction manual included in the µTracer kit is simply superb. The process it takes you through includes testing along the way, so there is little chance you will be faced with a non-working board at the end. Parts are meticulously labeled and extras were included in some cases where one might be lost. There were additional parts supplied just for calibration and ad­just­ment as well. I found nothing to frown about and it went very smoothly for me. Hence, this section will focus on the Heater Regulator (HR) board and building the system.

However, there are some parts which we may want to remove from the µTracer board, so you might want to leave those off: The two LEDs, L1 and L2, are replaced by solder terminals going to front panel LED lamps. I replaced the screw terminal connectors with solder terminals for (4) tube ASCG connections, (2) Heater connections and (2) Power connections.

µTracer/HR Construction Gallery

Here is a gallery of captioned photos intended as a guide for construction. (Gallery keys: Home, Rarrow16x13, Larrow16x13, End.) You might want to right-click the gallery and open it in a new window, to have it up while reading the sections below (which reference specific slides). For casual browsing, I find starting at the end and going forward to be more interesting. (The first half is about physical construction.) The normal order is chro­no­logical though. You can download a zip file of all the pics by clicking here. (That file is also included in the full documentation download.)

290 uTracerHR gallery

About the Vinyl Wrap

With this article, I’m proposing the idea of using custom-printed vinyl car wrap as a way to produce nice front-panels for electronic projects. It’s tough, durable, waterproof, attractive, accurate, full-color, self-adhesive and not hard to apply. This is far superior to the paper and spray-adhesive technique used for the VTA and Audio Level Meter projects, though they were quite sat­is­fac­tory. Mine was produced from this PDF file by for just $28 plus shipping. (Includes 12 x 54" of printing.) Made for use on vehicles where it endures sun, dirt, rain and car washes for years, Comparing vinyl print to HP laser 400it should be very robust for this indoor, residential application.

Of course, a printer designed to cover cars can hardly be expected to be as sharp on close exam­i­na­tion as one which handles only letter-size paper. Yet I was pleasantly surprised at the sharpness of the vinyl image I received. At right, you can see a magnified photo comparing the vinyl image on the left with an HP color laser print on the right. The “+19V” text is only 6 points, as it’s just a subtle reminder that the µTracer holds the cathode at 19V relative to ground. (Only significant if you connect external text equipment.) The most common font size for classic (pica) typewriters was 12 points. I’m pleased and amazed that VVivid’s vinyl printing is roughly comparable to the laser.

I chose 3M IJ180CV3 Media with a matte finish. This is a 2-mil white vinyl (printed) substrate with a pressure-sensitive adhesive on the back. It also has a 2-mil, matte, vinyl overlaminate and a protective paper backing to cover the adhesive. Slides 2 to 8 of the Construction Gallery above illustrate applying it to the chassis but for further instruction on applying the wrap, I rec­om­mend watching some of the helpful Youtube videos showing how to apply it to vehicles and boats. Though the process I used was simpler than what you see in the videos, I still found them helpful to understand the graphic material and how to handle it. A few that I watched include:   (But there are many others you can find.)

Finally on the vinyl application techniques, in my notes I wrote a step-by-step process which I used as a guide when I did the application. You can read that here. By the way, printing from a vector-based graphics app like CorelDraw to a PDF file stores the image in a very compact and pixel-free format. My file was only about 400KB, yet the resolution of the final image is as good as the pixel density of the printer.

Heater reg PCB details for quoteGetting the Heater Regulator PCB

I have two extra American-made HR PCBs and will be happy to send one of those anywhere in the USA at my cost ($33) plus $8 for Priority Mail. I have to say though that you can get it much cheaper by ordering directly from Elecrow in Shenzhen city, China or a similar supplier. Through, my quote was just $16 for quantity-5, delivered. At that price, it takes 20-days (click to door) but they offer 14-days for $34. [I assume that these figures are working days.]

Elecrow’s rating at the website is 4.6/5 stars. I also saw a quote from Seeed in China (rated 3.2/5 stars) for 6-day delivery at $31 (also quantity-5). I would start at, get the quote and click the link to the vendor. The info you need for the quote is at right. Though I entered quantity-1, they changed it to 5. The PCB data files are here, as sent to my American vendor (Advanced Circuits) but please note that I haven’t used far-east vendors before, so they might prefer something a little different. You will want to amend the “Readme TPCB105.txt” file or delete it if not needed. (My personal details are removed.)

Notes on Getting the Parts

You can use the BOMs linked in Part II to order parts for the µTracer and heater regulator.

µTracer System Parts

  • I used the painted version of the enclosure to provide a primer on the notoriously hard-to-paint aluminum, even though the original finish won’t be seen.
  • Sorry that I can’t identify the fuse holder I used. The hole size shown on the template must be changed to 15.2mm/0.6” for the one specified.
  • I had few choices for the 2-pole DC power connectors used, due to the limited space and 400V signals. They aren’t rated for that much voltage but were the only pick that I thought would be okay with that voltage and up to 3A of current. My only issue with these is the plug is too hard to insert and remove. They’ve loosened a little over time but I still have to twist while inserting and removing.
  • The male DB-9 connector of the USB/serial adapter must mate with the female one on the µTracer/HR chassis but the adapter has female-threaded stubs which interfere with the screw heads holding the chassis connector. I cut off the stubs with a Dremel cutting disk. It works but I would prefer something more secure.3-terminal plus gnd terminal strip
  • As noted in the BOM, one of the (3T + Gnd) terminal strips needs to have the orientation seen at right. The other needs one terminal cut off to have ground in the middle.
  • See above for details on the vinyl car wrap.
  • I’ve specified a nice, flexible, silicone wire for the patch cords (“pin cables”) but some might find the $16 price tag daunting. To be honest, I forgot about the wire I bought for it and used a small, 2-conductor wire something like speaker cord. The relatively thick insulation gave me some confidence that it would be okay with the 400V signals. I recommend you use 600V wire since it may be handled when live.
  • The stranded, #18 hookup wire specified is needed because it may see 3A of heater current, has to support 400V and must fit through the 0.09” inside diameter of the sleeve beads. This particular wire is excellent because it has a thin PTFE coating, making it impervious to a brief encounter with a soldering iron. (The coating doesn’t impede wire stripping like thick PTFE insulation does.) It also won’t shrink back during soldering like some PVC hookup wire.
  • Tube socket details for the ones I bought:   (Yours might vary.)
  •             Hole size   Spacing   Screw   Shaft   Del   DrawHole
    Compactron  1.125”       1.5”      #6     1.038”  87mils 0.375”
    9pin mini   0.787 20mm   1.1       #4/6   0.715   72     0.375
    7pin mini   0.709 18mm   0.893     #4/6   0.623   86     0.375
    Octal       1.125        1.5       #6*    1.045   80     0.375

    * It would be better to use #4 for the octal because the pan heads on the screws I used prevent full tube insertion.
    Hole size - Actual diameter of the hole to be punched.
    Spacing - Distance between centers of the socket mounting screws.
    Screw - Nominal mounting screw size.
    Shaft - Actual diameter of the part of the socket poking through the hole.
    Del - “delta” Difference between Hole size and Shaft. When centered, clearance is Del/2.
    Drawhole - Size of pilot hole needed for my draw punches. Yours might vary.

Heater Regulator Parts

  • C203 and C205 must be film caps to preserve the very high gain of the precision opamp, which is critical to accurate regulation.
  • You can probably find cheaper header strips. I use them only as solder terminals. Nature abhors a connector :)
  • Okay, I splurged on the 1ppm voltage reference but every time I look at the heater voltage, it’s exactly the same, down to a couple millivolts or so. I need something like that to give me the confidence to use it without checking the voltage. Oh, what a feeling!
  • L203 and L204 were pulled from extra $2 switching regulator modules because it was a cheap and easy way to get chokes which are said to work at 5A.
  • Speaking of those XL4005 modules, those were a cheap way of getting that function, with current limiting, without any development or risk. If you have trouble finding them, I bought a cache of 20 to support this article and will be happy to mail one if you need one to build the project and they are no longer available. Due to the limited quantity, you’ll need to get the two chokes from a supplier.
  • Getting the PCB is discussed above.

Building, Testing and Adjusting the Heater Regulator Board (HRB)

Building the HRB
(Please see Appendix B for a detailed discussion of the HRB design.) Though there are mounting holes provided, headroom inside the enclosure is limited, so I suggest you mount the switching regulator module (SRM) directly on the HRB, in physical contact. There are no traces on the top side of the HRB under the SRM. The five wires which connect the SRM hold it very securely. To mount the SRM:

  • Remove the screw terminals from each end of the SRM. I used heavy duty wire cutters to cut the plastic apart between the terminals. Then unsolder each terminal and clear the hole.
  • Remove surface-mount resistor R2 from the SRM. It’s between the edge of the module board and the leads of the big IC. I cut the resistor in half with a discarded pair of wire cutters. Then the ends come off easily with a soldering iron.
  • Solder (about 2”) #22 bare wires through the four holes from which the screw terminals were removed. I left about 1/4” of wire above the SRM surface to attach test leads.
  • Orient the SRM according to the pictorial on the HRB and pass the wires through the matching holes in the HRB. Push the SRM down to the HRB and bend the four wires sharply under the HRB to hold the SRM in place.
    SRM to HRB wire 320
    XL4005 module with heatsink 320
  • Solder the four connections under the HRB.
  • (See photo at right.) Solder a 2” #22 bare wire through the hole in the HRB labeled, “Pin2”. It’s near the edge of the SRM where the big IC is. Bend it and trim the length to allow it to clear the top of the SRM by say, 1/8”. Straighten it back out, put PTFE insulation on it, bend and solder it to the big IC, second pin from the edge of the module.
  • As seen at right, affix the 14x14x7mm heatsink to the top of the chip with adhesive. The peel-and-stick pads sup­plied with my heatsink order were very weak, so I used contact cement instead. The heatsink extends beyond the chip and the inductor is nearby so the heatsink won’t be centered but it’s okay.
  • That completes the SRM mounting.

Continuing with stuffing the HRB:

  • Install 2W resistor, R214 using 3/16” lengths of PTFE insulation to hold it off the board by about 3/32”.
  • Solder the two IC sockets, being careful to observe orientation, but don’t install the chips.
  • Remove inductors from two SRMs and use those for L202 and L203.
  • Cut 5 and 13-pin segments from the header strip. Solder the short pins to the board. Clip off every other pin, leaving the ones on the ends, so that each P-number (e.g. P211) is at an unclipped pin.
  • Cut a 1-pin segment and solder it to P206, near the board’s title text. You can see it in the upper right corner of the photo above.
  • Install the rest of the HRB components, being careful to match polarities of C201, C202 and diodes. Slide the jumper on JP201. Center the three single-turn pots on the HRB.
  • Solder a 0.22uF cap, C207, on the bottom of the HRB as shown in slide 25 and the heater regulator schematic. Use PTFE insulation on the leads. Note that the two pads at the bottom connection are connected on top, so no worries about shorting.
  • Turn the voltage pot (closest to the big IC) on the SRM clockwise 20 turns or until you hear it faintly clicking. Also turn the current limit pot (farther from the big IC) clockwise 20 turns or until you hear it clicking. I would prefer to set that for a 5.1A limit but that requires a 0.98Ω, 25W resistor, so is beyond our scope. Without that, it’s protected by the chip’s internal 8A limit.
  • Install the two chips, being careful to observe orientation.

Testing and Adjusting the HRB
That completes assembly of the HRB. Normally, it would be sensible to test the HRB before installation into the chassis. However, IF (a) you are very careful with assembly, (b) usually don’t make mistakes, (c) install with wiring routed to one side so the board can be flipped-up AND (d) don’t have current-limited lab supplies, it may be better to install it, untested and do the tests after building the system.

Otherwise, you should test it first. You’ll need +19V and -15V supplies. I highly recommend using current-limited lab supplies but if nec­es­sary, you can use the AC adapter for the +19V. If there is no other choice, you could use -15V from the µTracer but I advise against it. (Info on connecting to the -15V is in Part IV of this article.) You will also need a Digital Multimeter (DMM) and an oscilloscope (scope) would be nice. The other terminal of the +19V and -15V supplies goes to ground. Connect the terminals of the HRB according to this table:

Heater Reg
Board Terminal

Full Name

Connect To


GND (P201)


Power Supply Gnd


+19V (P202)


+19V Power Supply


-15V (P203)


-15V Power Supply


SwWip (P208)

Switch Wiper

P209, P210 or P211

Selects 5.0, 6.3 or 12.6V

Sw5.0 (P209)

Switch 5.0V Position

P208 or Nothing

Selects 5.0V

Sw6.3 (P210)

Switch 6.3V Position

P208 or Nothing

Selects 6.3V

Sw12.6 (P211)

Switch 12.6V Position

P208 or Nothing

Selects 12.6V

HGDrv (P206)

Heater Ground Drive

Output “Gnd”

Not the same as Ground

HtrDrv (P204)

Heater Drive



HtrSen (P205)

Heater Sense



HGSen (P207)

Heater Ground Sense

Output “Gnd”

Not the same as Ground

  • This setup has the two power supplies going to the HRB and the output of the HRB available between Output and Output “Gnd”. You can put a load there but it’s not needed, initially. (There’s a 390Ω internal load.)
  • Begin with P208 connected to P210, to produce 6.3V output.
  • If you’re using lab supplies, I would start with the +19V limited to 500mA and -15V limited to 50mA. After it’s been powered up and currents are seen to be low, you can increase the +19V supply limit to say, 2A.
  • Connect the DMM and scope between Output and Output “Gnd”.
  • Prepare load resistors for the three voltages. The goal is to pull 2.0A at 12.6V and 3.0A at 5.0 and 6.3V but your choices of power resistors may be limited. Ideal values for 5, 6.3 and 12.6V are respectively: 1.67@10W, 2.1@13W and 6.3Ω@25W Use the lowest resistances you have which don’t go much below the ideal values. Of course, you can wire multiple resistors in series and/or parallel to get what you need. Don’t have power resistors?
          Don’t forget that an audio load for power amp testing works well here. An exemplary lab load unit will do 2Ω, close enough to cover the 6.3V case. Parallel an additional 10Ω, 3W resistor with it and you have the ideal load for 5.0V. You could get away with using an 8Ω instead of the 10, pulling 3.12 instead of the ideal 3A. For testing 12.6V with an audio load,  paralleling a 30Ω 5W with an 8Ω load works.
          One more trick: You could use actual tube heaters for the load. (Imagine that!) 5U4 pulls 3A at 5V. The 6550 draws 1.6A at 6.3V and two in parallel would be an acceptable overage. A pair of EL34s would be even better, at 1.5A each. For testing 12.6V with heater loads, all that comes to mind is combining four 6L6s in series/parallel, so each sees 6.3V. That provides a nominal 1.8A load, 10% below the goal.
  • Exactly how you connect the load resistor, meter, drive and sense lines is critical to accurate load testing. If done properly, you should be rewarded with seeing that the load has almost no effect on the voltage. See Appendix C for details on how to avoid the error. The main thing is the meter should not connect directly to the drive line or the socket; rather, it should connect to the sense line and the sense line connects to the drive line at the socket. This goes for the ground side as well as the voltage side of the heater connections.
  • You may notice that the voltage adjustments have a very narrow range (nominally ±1.2%). That’s necessary because pot settings may drift 1% of their range, over time. To get rock-stable results, they have to be caged into a narrow range. But to support that, the fixed resistors and voltage reference have to be very accurate and stable. Hence, the 0.1%, low-tempco resistors and 0.04%, 1ppm/C voltage reference. On my unit, the 6.3V setting is currently reading 6.3003V—within 0.005% of its initial setting! The pots are all close to center, in spite of the narrow range. 

WHY ALL THE PRECISION? Please see the sidebar here.

To do the test procedure:

  • With the HRB connected according to the table above, P208 should be connected to P210 to select 6.3V output.
  • Turn on the +19V and -15V supplies and verify that their currents are less than (say) 50mA and 20mA, respectively.
  • The HRB output voltage should be between 6.24 and 6.36V.
  • Adjust R205 (6.3V ADJ) for exactly 6.300V.
  • Connect a load resistor, ideally 2.1Ω, 20W, which will pull 3A. The voltage should change very little. My unit increases about 2mV with a 3A load.
  • The scope should not show low frequency (under 1kHz) oscillations. I looked for the 300kHz switcher noise of the SRM but couldn’t find it in the output of the HRB. There, about all I could find were some 10kHz pulses at about 30mVpp, most likely from the microcontroller on the µTracer board.
  • Repeat the adjustment and scope and meter checks for the other two voltages:
    • 5.0V, Rload ideally 1.67Ω, 15W. The P208 connection now goes to P209.
    • 12.6V, Rload ideally 6.3Ω, 25W. The P208 connection now goes to P211.
  • Congratulations, that completes the heater regulator board! If you have problems with your unit, I will be happy to help—just post your issue in the Reader Comments section on the first page of this article. All Reader Comments sections on are monitored and are sent to the author within 24 hours (usually much sooner).

Coming up next: Assembling, wiring and taking ‘er out for a spin...