My Utracer is very inconsistent.

[randy mccuddy]

Hopefully this is the place to get this figured out. 

I built my Utracer to help me test and match up the couple hundred thousand tubes I've collected over the years. 

But so far, it's more than doubled the time it takes to simply match up a pair of tubes. 

Hard to explain. But I have a TV-7, an I-177, a Hickok 6000A, and a Hickok 539B, all restored and calibrated. Recently finished building my Utracer setup with an external heater supply, and was excited to put it to use. But, my readings are VERY inconsistent from tube to tube. 

The same tube will test consistent, but from tube to tube, it is not. I can test a pair of 12AX7's that show perfectly matched on my 4 other tube testers, and the Utracer will not follow. Not even close. 12AX7's in general almost all test as weak/bad on my Utracer. 

I have recalibrated and checked my calibration many times. Even monitored the voltages and currents with external test equipment. Grid voltages are spot on. Anode and screen voltages are spot on. Current measurements with 10K resistors are a perfect diagonal line following the procedure. 

Tonight I had an order of 12BH7's, and not a single pair of tubes were even close to being matched according to my Utracer. And that is out of 100+ NOS tubes and 200+ loose ones. Took the same tubes over to my TV-7, and the first pair read perfectly matched. Also tried them on the 6000A...same thing. Put them back in the Utracer, and showed horribly mismatched. I tried datasheet typical operation values and the values they will see in the amp they are going in. 

I'm using the typical operating parameters out of my RCA Receiving Tube manual and other random datasheets. And another users setups he sent me. 

Pretty sure it's not an oscillation problem, as my curves look good, just wrong. Even on EF86 and EL84 tubes. 

Help. I'm very frustrated with this thing. So far it's cost me hundreds of hours instead of saving me time. 

[Hoeberlin]

Hi, Randy,

i have my uTracer since many Years, and never hat a Problem like this. But during a Meeting of Vintage Radio Enthusiasts, a Colleauge showed me his Tracer, and he had Problems like You described.

Reason in his case is long Cabeling with many Ferrite Beads. 

I would recommend to have a single 9-pin Socket prepared with max 10cm. Cables, and attached directly to the corresponding Terminals on the uTracer Board. Please also use the Heater Supply of the uTracer, put the 2 Heaters in series to get a Heater voltage of 12.6 Volts.

Test a few Tubes with this setup. This will give You the Answer to "Problem on the uTracer Board" vs. "Problem outside the uTracer Board"

Getting good looking Curves is no Indication if You have Oscillation or not!

Hope this helps!

BR Henning 

-----Ursprüngliche Mitteilung-----
Von: randy mccuddy <lwr...@gmail.com>
An: uTracer <utr...@googlegroups.com>
Verschickt: Mi, 17. Aug. 2022 4:48
Betreff: My Utracer is very inconsistent.

--
You received this message because you are subscribed to the Google Groups "uTracer" group.
To unsubscribe from this group and stop receiving emails from it, send an email to utracer+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/bc737ecc-dca1-41ee-85f6-0df6181e0e19n%40googlegroups.com.

[randy mccuddy]

Well, too many ferrite beads is deff not my issue. 

But, cable length might be. I also have an 8 pole 5 position rotary switch. Total length of my wiring is 12-18 inches, including my connecting cables. 

I built mine a little unconventional. 

You received this message because you are subscribed to a topic in the Google Groups "uTracer" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/utracer/b36kVXcxwls/unsubscribe.
To unsubscribe from this group and all its topics, send an email to utracer+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/916902169.3030792.1660744852257%40mail.yahoo.com.

[theold...@gmail.com]

Hi Randy,

 

First of all, I am going to assume you have the uTracer 3+, not the 6.

 

As mentioned before, wiring of the tube sockets is very important, and to do a test, wire a single socket as close as you can to the board, eliminating all the other wiring in your current setup. Rotary switches are very nice, but can be a nightmare because of the stray capacitance they can add to the circuit. Ferrite beads are a good thing if used properly. Generally speaking more is not always better, but not enough of them isn’t either. How many? Well, that depends entirely on your specific build, no easy rule there. Oscillations *usually* show in ragged curves, stray capacitance can alter a otherwise nice looking curve. In Ronald’s write-up he recommends a “circular” wiring scheme, especially for multiple sockets. This is a very useful method since it help prevent signal reflection of an improperly terminated data link. Remember the terminator resistors in computer data cables?  Trail and error is the path.

 

Many of the old testers were very limited in what they actually test, i.e. most of them simply gave a measure of the tube emission. The uTracer shows a lot more of the tube, so be sure you are comparing apples to apples when comparing results to a vintage tube tester. For the small tubes you mention I like to use the quick test option; it compares the closest to many vintage testers, and is a quick test too.

 

Good luck,

Bill v

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/CADK59eoEzgvwCVCu2JzpVCm%3DhbfzRrR827ThZJMVjyt5CN9KzQ%40mail.gmail.com.

[randy mccuddy]

Bill, 

Yes, its a 3+. I can't wire my sockets up like a "loop", as only 1 socket is connected at a time. I did experiment with ferrite bead amount, and settled in with 1 bead with 1 wrap. Just the wire through a single bead caused bad oscillations. 2 wraps caused very low readings on everything. So, 1 wrap around 1 bead seemed to be the sweet spot. I also hooked up an oscope during these test and could clearly see when it did oscillate. 

All of my old analog testers are mutual conductance. I am aware it's very hard to compare apples to oranges, but I have compared directly knowing what configuration the standard "Hickok" GM circuit uses along with DVM measurements from my TV-7 while testing a tube. And I used those values in the Utracer so I could directly compare. Some tubes, like a 6L6, match right up and produce the same readings. It's smaller signal tubes like 12AX7's etc that seem to test very differently. And more specifically, tubes that use below -5v on the grid. That's when the Utracer gets very inconsistent. 

Thanks,

Randy

[theold...@gmail.com]

Randy,

 

As suggested before, your problem is either inside or outside of the uTracer. As a test, wire a socket close to the uTracer board eliminating all the extra wiring, and compare. All that extra wiring adds stray capacitance to the circuit, something that may affect the small tubes more than the larger ones.

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/2eef9169-06dd-4548-9e67-7d71ddadf931n%40googlegroups.com.

[randy mccuddy]

Bill,

Yessir, I'm working on that now. 

I got distracted with different readings between a direct RS232 connection and my Bluetooth adapter. Since my adapter is quite a bit more powerful than the standard HC-06 used, and since it's basically stuffed inside a metal box along with all the wiring and the Utracer board itself, it might be creating some interference and inducing some voltages. 

[randy mccuddy]

Tried a 9 pin socket wired up very close to the Utracer itself. Less than 4 inches of wire length, 1 ferrite bead per wire. 

No change. Each tube will test consistent test after test, but swap in another of the same type, and the readings vary drastically.  On 2 NOS 12AX7's that test perfectly matched GM wise on my other tube testers. 

One thing I noticed over the last few minutes is the grid voltage goes up with every anode pulse. For example, testing a 12AZ7, with 250v on the plate, -2v on the grid, the grid voltage will be -1.99 volts as soon as I press test, then go up to -2.64v with each pulse, but comes back down to -1.99v between the pulses. It does not go up with each pulse as the anode voltage increases. I know it's during the pulse because, well, I can hear it. My Utracer has a loud REEEEEEEE when it's turned on. According to the build manual, this is normal. But is it? 

Anyways, seeing my grid voltage change during the high voltage pulses is probably my issue. 

[Hoeberlin]

Hi, Randy, to nail  Your Idea of variing Grid Voltage, You may use one or two AA Cells, to generate a negative Voltage of 1.5 or 3 Volts, and this schould give You very consistant Results. 

If that is the case,you may want to check again, if there is something wrong in the Grid Circuit. Maybe you can get in direct Contact with Ronald Dekker. I assume, he will give You good hints, of what to check next.

Best Regards,

Henning

Von der neuen AOL-App für iOS gesendet

https://groups.google.com/d/msgid/utracer/ee34fe10-d0f4-4c5e-8456-668a26e6e461n%40googlegroups.com

.

[randy mccuddy]

Henning,

With a simple battery and -1.5v on the grid, everything seems to test like it should. My results are in line with my old analog testers now. 

[BHdeC]

I've had that exact problem too since the beginning.

With some tubes, the tracer just does not work. 

I've never solved this problem, but I think you guys nailed it with the grid voltage supply.

I'll follow closely the outcome.

[Hoeberlin]

Hi, Randy,

i searced on Ronalds Website (dos4erver.com) and found this Problem described. Ronald already did research on this, and the outcome is, to just add two Capacitors to the Board, to fix this Problem.

Please see Ronalds Website for more Information.

Inthe List of Content it's #32 (Tweaking the Grid Supply )

Please Check if You have done this already, of you missed that.

Hope this helps,

Best Regards,

Henning

-----Ursprüngliche Mitteilung-----
Von: randy mccuddy <lwr...@gmail.com>
An: uTracer <utr...@googlegroups.com>

Verschickt: Mi, 17. Aug. 2022 22:45
Betreff: Re: Re: My Utracer is very inconsistent.

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/1854b3e7-b559-4f2c-8b93-9366a4a746ccn%40googlegroups.com.

[Martin Manning]

The increased capacitor values in the grid bias supply appear in the uT3+ construction manual v 3.4, where v3.2 has the original values. Anyone who built from the earlier version (and did not catch the update) will get consistently low current readings.. I can recommend the Low Vg accessory for testing and tracing tubes with bias voltages approaching zero, and for getting a true Vg = 0 trace, which the on-board grid bias supply can not do.

[randy mccuddy]

Well, I read this:

https://www.dos4ever.com/uTracerlog/tubetester2.html#grid2

Since mine is a 3+, it should already have this. I'll check when I get home from work today. 

But I did find a very easy workaround last night.

Setup a 12AX7 to test with a socket extender, or a way to measure pins 2-3 during a trace. I used my Fluke 199c to measure. Use the standard 1.2ma 250v -2v setup.

Open your calibration window. 

Run a trace. Note the -2v grid bias dropping during each pulse. 

Adjust the -4v slider until the grid voltage is exactly -2v during the HV pulses. It will be a tad off between, but that does not matter. It's important it's correct during the HV measurement pulses. 

My 12AX7's now measure correctly. 

When we calibrated our grid voltages, it was done under no load. When the Utracer pulses the HV, it puts a pretty big load on the circuit. According to Ronald's write up, it's the octocuplers doing so. 

Repeat on a higher -V grid bias, but adjust the -40v slider. I used a KT88 at -40. 

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/76b42a80-161d-4780-b31b-026d7d01348cn%40googlegroups.com.