uTracer Tube Tester Build

[Dave451]

I've been building again. I had just re-built and calibrated my B&K 707 Dyna Jet tester, which I've been very happy with. I had been looking around, though, for a reasonably priced instrument to do more in-depth tube analysis and matching and found Ronald Decker's uTracer and decided to give it a try (google uTracer and you'll find his excellent website). Well, I bought the full kit and it was delivered quickly from the uTracer HQ in Holland through the support of Ronald and his wife Marie-Jose. I haven't seen much about it on AK, so I thought I'd share a few photos and notes about my initial experience with it and see if others have tried it.

The build took a few hours of careful work (a dense board with small parts--only a couple SMD parts, pre-installed--thanks Ronald!) and all worked well right off the bat, including the RS232 interface (via USB adapter) and an outstanding graphical user interface on my budget Windows 8.1 lap top. The construction manual is outstanding and basically has you build and test the unit a section at a time. Reminded me fondly of my Heathkit building days!

I bread boarded a 9 pin noval and 8 pin octal socket for some initial testing, but have a case and parts on order for a permanent housing (see the website for fully realized examples). The board and the parts are all high-quality and, to me, the circuit is remarkable for its ingenuity and effectiveness, consisting of a series of pulsed power supplies controlled by an on-board PIC and the PC GUI. You have to provide your own tube socket/switching scheme and builders have done some clever set ups (on the website under "testimonials").

The latest version 3 will generate anode and screen voltages up to 400 VDC. The whole thing runs on a surplus 18V 1.7A PC power supply I had on hand. Filament supply (pulsed) is provided up to the power supply voltage limit and about 1.5A. An outboard supply would be needed for higher-current, low-voltage tubes or voltages above the power supply max voltage.

I tested a few tube types, including a bunch of dual 12A(X,T,U,V)7 types, some 7591's, a couple of 5881's and a whole lot of 6V6GT's on hand from a tube cache I recently acquired. After a bit of study and practice, the instrument is pretty intuitive and straight forward to use and produces all sorts of excellent output curve traces for just any tube type you can hook up to it. Also very handy is a 'quick test' feature that tests the tube against the bogey values you put in from the tube data sheet and generates a printed quality report with mutual conductance, plate resistance, amplification factor, and plate/screen current at a given bias set point against tube sheet values, expressed as a number and % of the bogey values. Another neat feature is the ability to test both sides of dual triodes at the same time.

I'll show some examples of the output in the next couple of posts. I also decided to compare the uTracer mutual conductance values against the ones from my B&K 707 and I'll show that, too (short answer is they compared very well). The curves and quick test report below are for a (sadly) weak 5881 power pentode.

Interested in any questions and observations, particularly experiences from anyone with uTracer experience.

Dave

 

[whoaru99]

I bought a v3 kit several months back...it's still in the package.

Probably will have some questions if/when I ever get on it.

 

[Dave451]

Hey, heat up the iron and we'll compare notes!

 

[Dave451]

Here are some curves and an example of a 'quick test' report.

The 6V6GT output curves are for a tube that I designated my 'bogey' tube for comparing the others. It's pretty much right down the middle on the quality checks and the curves match the published curves well.

The 7591A output curves are for a new Tung-Sol Russian re-issue that checks very well on quality tests (as you would hope) and the curves also match or slightly exceed the published curves (Sylvania tube data sheet).

The report is for the tired 5881 output tube I mentioned above.

 

[peterh]

My version of uTracer has a small box that contains one tube socket and a "personality plug" where
the connections are done. This gives me ready made plugs for each type of tube - no mistakes
possible.

http://n.manet.nu/uTracer/

 

[arts]

A while back one of our members purchased one of these,and I was waiting for him (kind of a guinea pig thing,haha) to construct and evaluate his before going ahead with a purchase of my own.Never heard anything further though.I'll be following your experiences with interest Dave

 

[Dave451]

You bet, Art! I'll also document the final build on this thread when I get the case and the connectors and do some more actual tube work.

Peterh, I did actually see your build on the uTracer website--very creative! I see one of your interests is selling audio tubes, so this unit seems to make a lot of sense for identifying, matching and selling quality tubes. I would appreciate any advice, lessons, or learnings you have from your experience! I decided to use the 'pin to pin' jumper method for aligning the uTracer with the sockets after considering rotary switches. Thinking about how to wire them made my head hurt....

 

[Dave451]

As mentioned, I just gave my B&K 707 Dyna Jet tube tester a good going over recently. I replaced all the electrolytic capacitors, checked all the resistors and gave it a new tube, then re-calibrated it using the manual procedures/specs. All works well and I find this tester to be very convenient to use, particularly when testing multiple tubes. In the "Dyna Jet" section, the 707 measures mutual conductance (Gm), but reads out in % instead of directly in micro Mho's.

Since the uTester is a much more precise instrument in design and execution, I decided to compare the Gm results from the "Quick Test" feature with the 707 results. The uTester uses computer power to determine Gm, mu, plate resistance etc at a given bias point by calculating the corresponding derivatives of the curves for combinations of anode current and screen current versus anode voltage, grid voltage, and screen voltage.

I had 16 6V6GT tubes on hand (12 were JAN Sylvania NIB tubes from the 1940's), so I measured GM with both the uTracer and the 707 and plotted the results on the graph you see below. The data was pretty well behaved and the correlation was good. For the 6V6, the bogey value for Gm is 4.1 k uMho (arrow on the y axis) at the conditions tested. The plot shows that the 100% value for the 707 corresponds to a little over 4.0 on the uTracer. If you take out the outlier point on the bottom left, then the linear fit is very close to 4.1 at 100% on the 707--pretty good for the 6V6 tube type.

I don't know if this would hold up for all tube types. For example, the 707 is not a good tester of 12AX7's (one reason I got the uTracer) and tests them well out of the 'sweet spot' on the operating curve. Parameters for the Dyna Jet socket for 12AX7 is apparent much better set up for testing the 12AU7, but I haven't done comparison data for triodes at this point.

Just thought I would pass this along as a matter of curiosity satisfied. In reading about fixed tube testers, one becomes aware of their limitations for matching or even really quantifying the relative 'goodness' of a tube, even though they're pretty good at identifying bad ones. This lack was another reason to try the uTracer.

Dave

 

[trobbins]

Hi Dave, given the uTracer uses stepped and pulsed measurements, a few queries come to mind including accuracy of the V and I measurements.

I can see that precision resistors, and some amount of inherent averaging of measurement samples, were used in http://dos4ever.com/uTracerlog/tubetester2.html#examples to give a known reference to gauge accuracy of V/I (not sure if V and I are calibrated from a quick perusal). Do you go through that form precision resistor set up for calibration in a kit setup? The website is a mountain of information - a project in itself to read through!

Your B&K doesn't allow raw heater voltage setting it seems - were you testing with heater voltage at nominal, or do you need to use an external variac to allow that setting?

 

[x3workshop]

AKer @Brice has recently built one and we've compared the same tubes. His readings and the readings I get on my maxipreamp/maximatcher are consistent. I'm sure he'll chime in soon.

 

[gadget73]

Very cool that the Tung-Sol 7591 compares well to old datasheets. There has been a lot of concern about how well they actually match up to old production. I'm curious how one of the EH fat bottle 7591's compares. Unfortunately I don't have any otherwise I'd send one to find out.

 

[x3workshop]

There is an excellent examination of new issue/reissue 7591a tubes over on the Tronola.com site. Highly recommend it. http://www.tronola.com/html/7591a_tubes.html

 

[Dandy]

I went through my tube stock recently, and wondered about getting a kit myself. Especially for testing my 7591s and 6BQ5s, which mostly come from unknown sources, or old equipment I've bought.

 

[peterh]

A little bit of warning here : uTracer _may_ be damaged by shorted tubes. Do test unknown tubes elsewhere first!

 

[CZ4A]

I have a uTracer 3+ that I've built and experimented a little with. I bought a small Pelican hard case off of Craigslist to mount it into to make something but haven't made the time or gotten all of the materials I need. I have the sockets but not all of the wiring or a mounting board. This completed kit is what inspired me to get a hard case. http://www.stagecue.com/utracer.html

 

[Dave451]

trobbins: The calibration procedure is fairly extensive and is in the last section of the construction manual. There is a test with 1% supplied 10K resistors to calibrate the gain of the anode and screen current amplifiers. The check is for linearity of Ia vs. Va and Vs and calibration of the anode and screen current amplifier gain by using known (previously calibrated) Va and Vs levels. The B&K 707 does have line voltage correction (of a sort). I just set the heater voltage on "6" and ran the 707 tests alongside the uTracer tests. I did not otherwise measure 707 filament voltage. So, it's not a highly controlled experiment, but I just wanted to check what, if any, correlation.

gadget and x3workshop: I have seen the (excellent) 7591A testing and comparison on Tronola--great stuff! I have not used the EH tubes, but have used multiple quads of the Tung-Sols and have been satisfied with them (other than some quads are pretty hot and require a high bias level that may or may not be in the range of the stock unit). I plan to get a matched quad of EH and check them alongside my Tung Sol's already on hand.

Dandy: I did some preliminary testing of my 6BQ5/EL84 pulls and 7591A pulls against NOS and new production. If the 707 showed them weak or bad, they were weak or bad as tested on the uTracer. Haven't done the 7189's yet, but I will. With all this testing capability and data being generated, I need to ultimately get organized a bit to keep track of it all and use it best. Unlike the 6V6's, though, the 7591's can't be tested for Gm on a stock B&K 707--the socket is in the "switched" section where only emission is checked. One thing I think I'm seeing, though, is that emission levels from a tester don't correlate much with Gm, but I've got more work to do on that. That seems to match the conventional wisdom on the usefulness of emission test results on power tubes, I think. The EL84 series is checked for Gm, though, on the 707.

Peterh: Good advice. I think Ronald may have improved the units ability to withstand shorts through the "compliance" circuit, which shuts off the unit above a certain current level (which can be adjusted), but I haven't studied it enough. I am only testing tubes with the uTracer, though, that have already been screened for shorts and grid emission on the 707. Better safe than sorry!

CZ4A: I loved the stagecue build in the yellow travel case. I thought about doing it this way as well, but settled on a more bench top configuration with a fixed Budd metal case. I'm adding some complexity to mine by installing a 12 pin socket to check compactrons. I became more curious about using these tubes based on Pete Millet's work with them (I built the "Engineer's Amplifier" using the 6JN6 and have quite a few in my tube cache to check).

Thanks for the feedback, guys!
Dave

 

[whoaru99]

I envision my eventual build to be in a sloped front case, some version along this line.

A few tube sockets on the top flat, with the controls and pin jacks/jumpers on the sloped front.

 

[arts]

I envision my eventual build to be in a sloped front case, some version along this line.

A few tube sockets on the top flat, with the controls and pin jacks/jumpers on the sloped front.

Oooh,I've got a couple of those brand new stored away somewhere (?) from my studio gadget-building days.....I had totally forgotten about those....damn....

So tempting; I can feel my wallet getting a µTracer 3+ ''lightening'' already....gee,thanks Who,if the toy spending keeps up I'll be back on the cat food shortly

 

[Dave451]

I like that case. I decided to go with this one for a little more room for the 5 sockets and 12 pin connections I need to accommodate the compactrons, as well as a large 9 pin (eg. for 7868's), an octal socket, a small 9 pinner, and a 7 pin.

 

[Brice]

Yes I build one recently and I am very pleased with the result. This is very accurate and consistent with x3workshop's maxipreamp/maximatcher testers.

It's really nice to trace an tube to visually see it's condition. The quick test is really nice and matching sections is a visual exercise as you trace both side at the same time.

As far as I know, the Utracer supports all kind of tube shorts with it's compliance feature shutting down a supply is i > 200 ma max in a few micro-second.
For the heaters, I made mistakes blowing the internal dedicated fuse but no damaged.
For a grid shorted, you may want to add a 470R in the grid circuit to activate the protection.

For simplicity's sake I build tube enclosures per tube family I use, typically 12aXXX, KTXX, EL/6L6XX, and some dedicated like EF86, 6BH6, 6CG7.
That way I don't mess with jumpers. I grab the right tube enclosure, plug it to the tester, plug the tube and hit trace on the UI and that's it. Simple.

I think it is a affordable must have for any serious tube hobbyist.
It is just not that simple to assemble, but with time and method, anyone can build one.

NOTE: do not forget to install ferrite beads in your sockets, to prevent oscillation.

Here is mine