Replacing the R6 resistor

I could now get at the circuit board and, with the meter disconnected, the 560-ohm R6 resistor was no longer in circuit. I measured its resistance and found around 1150 ohms! Next I unsoldered the resistor from the board and tested it again. Out on the bench it read 840 ohms. I can't account for the difference in the two readings but both are completely out of the ballpark.

I soldered in a new 560 ohm resistor, temporarily connected the meter, and turned on the dipper. I turned the sensitivity control and... wham! - the meter needle hit full deflection with gusto. The problem was fixed.

I made temporary connections to the meter and tried finding a resonance. I found that it worked well for lower frequencies, say below 20 MHz but needed pretty close coupling to get a dip at higher frequencies. At the higher frequencies there is a good dip but only if the tunnel dipper's coil is very close to the trap being tested.

Fixing the Heathkit HM-10A Tunnel Dipper

In October 2012 I received an email from Nigel, VA2NM, who had bought a Heathkit HM-10A Tunnel Dipper for $2 (!!). He found an excellent description and theory of operation of the device published in the December 1964 issue of 73 magazine (click here to dowload a .pdf version).

Nigel couldn’t get his tunnel dipper to dip and was considering replacing the circuit in the dipper, possible with one using JFETs. He wrote to me to compare notes and this prompted me to take another look at the problem with my tunnel dipper.

Here is my experience with the Heathkit HM-10A Tunnel Dipper:

I purchased it off the Internet and suspected that it had a bad tunnel diode. What I found, though, astounded me: the carbon composition resistors in the unit went up in value and not just a little bit. I found, for example, that resistor R8 (removed from the circuit) went from 560 ohms (marked value) to 620 ohms (measured). I understand from people on the Internet that this is caused by moisture intrusion into the resistors. I could not get the voltage values that were on the schematic - they were off a small amount but this indicated to me that all nine of the resistors needed to be changed.

I originally suspected that the tunnel diode was bad (it is part number GE STD633 which I believe was a custom part number made for Heathkit by General Electric). Instead, what I found was that the tunnel diode bias voltage was changed so dramatically by the faulty resistors that the tunnel diode was no longer biased properly. I decided to measure the tunnel diode but this is not easy because the measuring instruments add a great deal of error at these low voltages and currents. I found that the VP was 45 mV at 145 µA and the VV was 166 mV at 49.6 µA. These two values may make it much easier to find a replacement tunnel diode, possibly a 1N2941, should it be necessary. Tunnel diodes, especially germanium, are getting very hard to find and are expensive. I suspect the F0 was greater than 1 GHz and the capacitance less than or equal to 50 pF.

My next step was coil connector cleaning. The coils use RCA jacks and for some reason they have corroded badly. I found the best way to clean them was with a soft eraser which removes the crud without removing the finish. I had to cut an eraser to fit on the inside of the RCA jack. I found one of the coils had a broken wire in it and I suspect that is not going to be easy to fix. The coils appear to have been pre-made and not part of the kit, as each coil has been dipped in a different-colored hard material with the RCA connectors epoxied on.

I made an RF snifter probe out of some insulated 10-gauge house wire: I wrapped four turns of wire around a three-quarter-inch cylinder (close wound) and soldered it to the end of a piece of RG58 coax. This allows the sniffer to fit over the top of each coil. The sniffer, when placed around a Tunnel Dipper coil, outputs a signal level of around -50 dBm on my spectrum analyzer.

By the way, I found that velcro dots work very well for sticking the coils to the inside of the lid.

Thanks for that, Allen. Your findings echo mine in that replacing all nine composition resistors is likely all that is needed to rejuvenate a Heathkit HM-10A Tunnel Dipper.

Feedback

On 12th March, 2013 I received an email from Allen Ripingill who writes:

Checking the voltage on the meter

My main problem with the tunnel dipper was that the meter needle never reached more than about 80% deflection. I had found a copy of the Heathkit Tunnel Dipper manual on the Web and it included a circuit diagram with indicated voltages for a 50% deflection. For instance the voltage on the meter at 50% deflection should be 0.3 volts. That made sense since the current through the meter went to ground via a 560Ω resistor (R6) so that 0.3 volts would give a current of 1000 * 0.3 / 560 mA = approx 0.5 mA. Since it’s a 1 mA meter, 0.5 mA gives 50% deflection.

When I set the sensitivity control on my tunnel dipper to give a 50% deflection, I measured 0.5 volts on the meter. That meant that the 0.5 volts was being grounded through 1000 ohms, not 560 ohms. I did a Google search and found that composition resisters, which is the kind used in the tunnel dipper, can significantly increase their resistance if they are exposed to dampness or high humidity. I decide to replace the resistor.

Getting the tunnel dipper apart

The copy of the Heathkit Tunnel Dipper manual that I found on the Web unfortunately did not include the pages dealing with construction. Therefore, I had to determine my own way of taking the tunnel dipper apart to get at the circuit board.

The first thing that seemed necessary was to disconnect the meter. I started to unsolder the meter connections but found it difficult to do and ended up cutting the wires. I reasoned that, once I had the thing apart, it would be easy to solder in new wires.

Next I removed the two control knobs. The knob on the tuning capacitor was easy but the one on the sensitivity control was not. Its set-screw seemed to need a very small hex wrench but the one I tried seemed to fit O.K. but slipped. I then found that I could undo the screw using a very small flat screwdriver. The tuning capacitor shaft also had a nut and locknut that I removed.

After that, I removed the two bolts at either end of the switch and was then able to pull on the grey metal casing and ease it over the switch. The casing could then be slipped off the tuning capacitor and sensitivity potentiometer shafts.

Replacing the other resistors

All nine of the resistors on the circuit board are composition resistors and if one can fail, they all can fail. Therefore, a proper job might be to replace, or at least check, the other eight. However, that would probably mean removing the circuit board which involves unsoldering the variable capacitor. Instead, I decided to leave that as a future project.

The bias on the tunnel diode

The tunnel diode is biased to make it operate in its “negative resistance” range. The bias is supplied by the potential divider formed by R7 (47 ohms) and R8 (560 ohms) giving a calculated bias of 1.5 * 47 / (47 + 560) = 116 mV.

When I measured R7 and R8 in-circuit a couple of days ago I got 50 and 870 ohms giving a calculated bias of 1.5 * 50 / (50 + 870) = 82 mV. Once again a 560 ohm resistor was reading much too high!

When I re-measured the two resistors in-circuit today I got 50 and 640 ohms giving a calculated bias of 1.5 * 50 / (50 + 640) = 109 mV. This is the second time an in-circuit reading has come down considerably over a day or so, so maybe running the tunnel dipper is bringing the resistors back into spec! Today I also measured the actual bias and got 110 mV with the “Diode” switch setting and 108 mV with the “Oscillator” setting so the calculated and measured voltages are in agreement. The bias is a little lower than design but not significantly so.

Update

I found a site from which you can download a .pdf version of the complete manual, including all the construction pages.