8. Tunnel diode Cross Reference (Russian / Tektronix)
Several of the Tektronix instruments I bought for the project use tunnel diodes (TDs), an obscure component that in the 80s due to its versatility, was viewed as a panacea for all manner of applications, and they were relatively cheap and plentiful then. Not so these days, when most people have forgotten about them; the odd few that come up on eBay are usually very expensive. Having experienced failure of some instruments with these the likely culprits, I set about finding a way to test them and sought alternative, less expensive replacements - the fall of the USSR a particular benefit in this respect. A precaution to bear in mind when de-soldering and replacing TDs is their sensitivity to heat; a wise precaution with any soldered semiconductor replacement is to clamp the leads first using a pair of pliers to wick away the heat.
One of the easiest ways to check a TD is with a curve tracer and I have two: a relatively modern Tek 571 with limited ranges but useful features like cursors and retrievable non-volatile setups, and an older Tek 577 with analogue trace storage, more expansive ranges and higher current and voltage capability. This section categorises the relevant TDs and describes the issues I encountered with these instruments.
First, a small list of tunnel diodes used in Tektronix instruments:
http://w140.com/tekwiki/wiki/152-0177-02
Second, a useful large table of standard JEDEC numbered tunnel diodes:
http://w140.com/tunnel_diodes_table_cs.html
The Tektronix 152-0177-02 is a 10mA, 2pF [according to Tek's parts list for 7S14 it's 4pF] tunnel diode in a DO-17 'top hat' package. It is used in the Tek 067-0681-01.
The 152-0177-02 is very similar to the 152-0177-01, but the 52-0177-02 has a tighter tolerance on Ip, +/- 0.5mA for the 152-0177-02 compared to ±1mA for the 152-0177-01. (Is the 152-0177-02 simply a selected 152-0177-01? The 152-0177-02 is listed in the 067-0681-01 as being manufactured by Tektronix. Was it fabricated by Tektronix or just binned and/or selected by Tektronix?)
The 152-0177-01 is also unobtanium.
The Tek 067-0681-01 tunnel diode pulser calibration fixture uses 152-0177-02 (see above).
Ip ±0.5mA vs ±1mA for 152-0177-01. No idea what the 7S14 152-0177-00 tolerance is.
The 7S14 delayed ramp gen has the following tunnel diode (SM = Service Manual schematic page):
SM cct P.5A CR459 Tek 152-0169-00 1.0mA 10pF aka GE 1N3712
The 7S14 trigger board HF sync oscillator has the following 2 tunnel diodes:
SM cct P.2 CR220 Tek 152-0177-00 10 mA 4pF is a Tek special
SM cct P.2 CR221 Tek 152-0070-00 0.1mA 3pF aka GE 4JFBD4 (GP tunnel diode) American µSemi also made it
I considered Russian tunnel diodes as potential alternatives: following the fall of the USSR, many cheap devices can be found on eBay, including high quality mil-spec. After creating my own large TD specs table from multiple eBay auctions (see bottom of this page), a subset of which is below, I later discovered someone else had done the same thing and put it up here:
http://w140.com/tekwiki/wiki/Russian_tunnel_diodes
The following Russian tunnel diodes are the closest I could match from the lists at the bottom of this page:
NOTE: I am unfamiliar with TD specs, therefore these are my best guesses and since Tek only lists the capacitance and current in their instrument BOMs (items lists), my guesses
are based largely on these however I am sure there are other parameters that matter!
Values at 25°C: Tek NEAREST Tek NEAREST Tek NEAREST
Tek part 152-xxxx-00 0169 GUESS 0070 GUESS 0177 GUESS
Alternative part number 1N3712 AI301A 4JFBD4 3I101A - AI201A No.
Manf GE/GPDC Russia GE/AuS Russia Tek Russia Manf
Technology Ge GaAs Ge GaAs Ge? GaAs Semi
Spec - Com Mil Mil Mil? Ext Spec
Fmax, GHz 2.3 - - - - - Fmax
Peak point current, Ip in mA 1.0 2.0 0.1 0.75-1.25 10 9-11 Ip
Peak point voltage, Vp 0.350 0.18 0.17 0.16 - 0.2 Vp
Projected peak-point voltage, Vpp 0.065 0.65 - - - - Vpp
Projected peak-point current, mA 0.18 - - - - - PPI
Peak to valley current ratio, Ip/Iv - 8 1 5 - 10 p/v
Forward voltage, CW, Vf 0.500 - - 0.35 - 0.4 Vf
Reverse voltage, CW, Vr 0.040 - - 0.035 - - Vr
Valley point terminal capacitance 10 <12 3 3 4 3.5 pF
Inductance, nGn 0.5nH <1.5 1.5nH 1.3nH - 1.3nH L
pulse forward current Ifp, mA - 1.2 - - - - Ifp
dc cw forward current Ifc, mA 5 1.2 - - - - IFc
dc cw reverse current Irc, mA 10 - - - - - Irc
Impedance, Ohms 4 - - 24 - 8 Z
Pulse power, mW - - - - - - P
Max temp, °C - +70 +100 - - +85 °C
Sold from: - Ukraine - Ukraine - Ukraine From
Price June 2016: - $8.99 - $3.79 - $3.99 $
I bought these tunnel diodes, including 8-off AI301A (2mA 12pF) to experiment with:
Above, 571 setup Below, AI301A in 571
Above left, what the 577 produced. Above right, theoretical TD curve, above that, what the 571 produced.
577 setup:
Max peak power 0.15W, max peak volts 6.5V, 120Ω series resistance, step/offset amplitude 100µA, horiz 200mV/div, vertical 500µA/div, 'collector' ~32%.
571 setup:
Max power 0.1W, max 2V, max2mA, 100Ω series, horiz 200mV/div, vertical 200µA/div.
The trace on the 577 (top right) for Russian AI301A TD#4 looks close to a tunnel diode characteristic (superimposed on the Tek 571 setup menu above right) but is noticeably jagged where I would expect the trace to smoothly curve downwards after the initial peak. At first I suspected the 577 but then I found it did the same thing on the 571. A little digging on the 'net revealed TDs are prone to oscillation and it's very difficult to display their theoretical curve.
The following website depicts a tunnel diode trace taken with a 571 that exhibits a lot of noise:
http://electronics.stackexchange.com/questions/165969/tunnel-diode-i-v-curve-noise
Here is a Youtube video showing very similar results to mine, for Russian TD 3I306G:
http://www.youtube.com/watch?v=LPExywj5pJM
Commenter Donald Pauly explained: 'Your diode is oscillating. Contact me thru www.gonascent.com and I will tell you how to stop the oscillation and see the real curve.'
The following article seemed to have clinched it [E47], Bias Circuit Instabilities and Their Effect on the dc Current-Voltage Characteristics of Double Barrier Resonant Tunnelling Diodes, citing inductance as a major contributor to tunnel diode instability, and investigates R,C biasing to remove it.
Curious, I tried shunting a Russian AI301A TD with short-lead 10pF, 100pf, 1000pF, 10nF and 100nF
capacitors but there was no improvement. Adding 20Ω in series was similarly ineffective.
Frustrated, I contacted Donald Pauly at the above link and he replied: 'The requirement is to force a resistor in series with the diode whose value is greater than its least negative resistance. This will force the oscillation to stop'. He suggested I add a 50nH inductor in parallel with a 47Ω resister in the TD anode back to the curve tracer using very short leads, tacking it very close to the case of the diode anode. 'The bodies of the diode, inductor and resistor must be as close together as possible'. This forms a resonant circuit at 205MHz with a Q of 0.7 in conjunction with the 12pF of the AI301A TD capacitance':
I planned to use the cheap Chinese LC meter and attempted to calibrate it first, but it was always problematic in calibration mode and this time I think it overwrote its code, because now it only displays 0 all the time. Instead, bought a UK-based Peak Semiconductor LCR45 to replace it. Although I can't calibrate it myself any longer, the yearly cal price of £20 is acceptable, and once that's done I'll immediately use it to calibrate my other equipment.
The smallest 47Ω resistor I could find was a 30 year old Allen Bradley C3 carbon film. After all of those years, out of my pack of 50 the lowest value was 50.4Ω.
I had some old air core inductors too, although measuring 50nH was really pushing the LCR45.
Adding the 50Ω resistor and 50nH inductor to the tunnel diode anode certainly improved things, but there was still a spike halfway down, followed by a vertical drop.
Donald said it was still oscillating:
I replaced the resistor with a non-inductive Allen Bradley 81E 500Ω
single turn Cermet trimpot set to 47Ω and re-ran the tests.
I found I could vary the pot between 6Ω and 84Ω beyond which fierce oscillation resumed.
At 84Ω the curve looked smoother, but I could not remove the small peak at the top:
Donald said the coil might be physically too big and suggested I manually wind a 500nH coil around the resistor. Given the C3 resistor measures 1.83mm dia x 3.86mm long and my smallest magnet wire is 0.5mm thick, I didn't think I could get sufficient turns on it, and sought an alternative solution.
I hunted through my inductors and found a 500nH on a plastic former with a ferrite tuning slug that I removed. Taking the coil off the former, I noted the inductance dropped to 400nH. I knew I couldn't change the resistance more, but the lower inductor value was worth experimenting with.
I soldered the C3 resistor inside the coil to minimise size.
Unfortunately it was not to be.
The new layout resulted in a sharp peak on the downward slope:
Tacking on decoupling capacitors of various sizes made it worse.
We were hoping to tune out the TD native oscillation frequenc(ies). Donald's initial best guess helped, but didn't completely nail the oscillations, so I looked for 'smoking gun' oscillations using the SignalHound spectrum analyser (SA). I was surprised to find nothing of significance.
First of all, to get a general picture without affecting the source, I plugged just the TD into the 571 and held a WTA-HA777 28-1800MHz general purpose aerial tip half an inch from it. The image below left is what I found with the SH scanning from 100MHz to 1GHz. The peak at 163MHz (Marker 1) looked interesting, but when I physically turned off the 571, it and all of the other peaks remained - I surmised they were coming from my PC, the only other active circuitry in the lab.
Next I directly sampled the TD anode on the 571 using an ac-coupled Auburn Technology P20A,
100kHz - 3GHz probe (http://www.auburntec.com)
The image below right shows what I got with the SA scanning from 100MHz to 4.4GHz - a fairly benign scan with about 5dB of noise. This time Marker 1 is not on a peak. Later, closer inspection of the 571 waveform on an oscilloscope revealed it to be an incredibly electrically noisy signal. For now I could discern no peaks of interest from the TD that exceeded the noise from the PC and the 571, so it was back to traditional investigation with an oscilloscope.
I used the Tek 7834 to view the 571 anode waveform, the only instrument I had that was capable of seeing the ~300MHz sinewave I found superimposed on its pulsed output, viewable only when utilising its variable persistence storage mode:
Donald said 'The 2mA diode has a peak current at 0.1V and a valley current at 0.4V. This is ~2mA change in 0.3V or -0.3V / 2 mA = -150 ohm negative resistance. You must place more than this in series with it to get it to stop. This is the average negative resistance from 0.1V to 0.4V. The smallest negative resistance will be several times less around 0.2V.'
He suggested I try 220Ω, adding: 'A 500nH coil has a reactance of 942Ω at 300MHz which should force the RF current into the 220Ω resistor.'
I checked the rest of
the coils on formers and the closest I could get to 500nH was 660nH minus its tuning slug.
I didn't have a C3 220Ω
resistor, so I used the 500Ω cermet pot again, the closest achievable value being 224Ω.
This experiment produced a curve (below right) almost identical to the one with 47Ω and 50nH, the only difference being the peak at the top is slightly reduced. For now this will have to do.
It is impractical photographing the dynamic waveform from the 571 so I took several video recordings of it and the 571 simultaneously displaying the trace in the setup below, using a Tek P6201 900MHz FET probe on the 571 anode into a 7A29 input amplifier with a 7B15 timebase on the 7834. The probe is conveniently held in place with a rubber band hanging off the trigger input BNC of the Rigol DSO above it, out of picture. Links to the videos are provided below.
Above, the best AI301A curve with 660nH and 224Ω.
Below, the P6201 Fet probe on the TD anode.
My experimentation has improved the TD plot but I still think it could be better. I plan to return to this experiment at a later date when hopefully I will have a range of SMD resistors and inductors.
Videos:
Note - These videos appear thin because they were filmed horizontally (as that was the only way I could get it all in view), and then digitally rotated.
308v01 P1140715 7834 scope 100ns 100mV ac-coupled.MP4 - 1814MB
308v02 P1140716 7834 scope 20ns 100mV ac-coupled.MP4 - 1758MB
308v03 P1140718 7834 scope 100ns 100mV dc-coupled.MP4 - 1828MB
308v04 P1140719 7834 var persist storage 2ms 100mV dc-coupled.MP4 - 2342MB
308v05 P1140720 7834 scope 50ns 100mV dc-coupled.MP4 - 1914MB
Russian AI301A tunnel diode on Tek 571, anode voltage on 7834 scope
P1140719_7834_var_persist_storage_2ms_100mV_dc-coupled VHD
P1140718_7834_scope_100ns_100mV_dc-coupled VHD
P1140716_7834_scope_20ns_100mV_ac-coupled VHD
P1140720_7834_scope_50ns_100mV_dc-coupled VHD
Below is a list of all Russian Tunnel diodes on eBay June 2016
NOTE:
This list is provided here as a resource, however:
All of these values were taken from SEPARATE eBay auctions, and different auctions list different parameters, sometimes under different names, so there will be errors and omissions.
Furthermore, I do not speak Russian so I cannot check the values in their original
datasheets.
If you notice errors, please feel free to suggest corrections.
Feedback from Nikolay Pavlov:
'GI 1I (military variant of GI) are Germanium not GaAs diodes [CORRECTED]. They usually have the lowest valley voltage so are best for detectors, etc (you can look up long range RFID tags based on tunnel diodes).'
Part no. also listed: GI103A GI103V GI104A GI304A GI304B GI305A GI305B GI307A
PART 1I103A 1I103V 1I104A 1I304A 1I304B 1I305A 1I305B 1I307A
Manf Russia Russia Russia Russia Russia Russia Russia Russia
Technology Ge Ge Ge Ge Ge Ge Ge Ge
Spec Mil Mil Mil Mil Mil Mil Mil Mil
Fmax 10GHz 20GHz 3.15GHz - - - - -
Peak point current, Ip in mA 1.3-1.7 1.3-1.7 1.3-1.7 4.6-5.2 4.8-5.4 9.2-10.49.6-10.81.8-2.2
Peak point voltage, Vp 0.09 0.09 0.09 0.065 0.065 0.07 0.07 0.65
Projected peak-point voltage, Vpp - - - 0.4 0.4 0.4 0.4 0.4
Projected peak-point current, mA - - - - - - - -
Peak to valley current ratio, Ip/Iv 4 4 4 8 8 8 8 7
Forward voltage, CW, Vf 0.4 0.4 0.4 - - - - -
Reverse voltage, CW, Vr 0.02 0.02 0.02 - - - - -
Valley point terminal capacitance, pF 1-2 0.7-1.3 0.8-1.9 <20 <20 <30 <30 <20
Inductance, nGn 0.2nH 0.2nH 0.1nH - - - - -
pulse forward current If, mA - - - - - - - -
dc cw forward current If, mA 1.5 1.5 0.51 10 10 20 20 4
dc cw reverse current Ir, mA 1.5 1.5 1.5 - - - - 10
Impedance, Ω 6 7 - 10 10 20 20 4
Pulse power, mW 200mW 100mW - - - - - -
Temperature range, °C - - - -60+100 -60+100 -60+100 -60+100 -60+100
eBay sold from: Russia Russia Russia Ukraine Ukraine Ukraine Ukraine Ukraine
Price June 2016: $3.80 $7.25 $3.79 $3.99 $3.99 $3.99 $3.99 $3.99
PART 3I101A* 3I101B 3I101D 3I101E 3I101G 3I101I 3I101J 3I101V
Manf Russia Russia Russia Russia Russia Russia Russia Russia
Technology GaAs GaAs GaAs GaAs GaAs GaAs GaAs GaAs
Spec Mil Mil Mil Mil Mil Mil Mil Ext
Fmax, GHz - - - - - - - -
Peak point current, Ip in mA .75-1.25.75-1.251.7-2.3 4.5-5.5 1.0-3.7 4.5-2.5 4-6 1.7-2.3
Peak point voltage, Vp 0.16 0.16 0.16 0.18 0.16 0.18 0.18 0.16
Projected peak-point voltage, Vpp - - - - - - - -
Projected peak-point current, mA - - - - - - - -
Peak to valley current ratio, Ip/Iv 5 5 5 6 6 6 - -
Forward voltage, CW, Vf 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.35
Reverse voltage, CW, Vr 0.035 0.035 0.035 0.035 0.035 0.035 0.035 0.035
Valley point terminal capacitance pF 3 2-6 2.5-10 2-6 2 4.5-10 2-6 2
Inductance, nGn 1.3nH 1.3nH 1.3nH 1.3nH 1.3nH 1.3nH 1.3nH 1.3nH
pulse forward current IF, mA - - - - - - - -
dc cw forward current IF, mA - - - - - - - -
dc cw reverse current IR, mA - - - - - - - -
Impedance, Ω 24 22 14 10 16 7 8 18
Pulse power, mW - - - - - - - -
Temperature range, °C - - - - - - - -60+85
Sold from: Ukraine Ukraine Russia Russia Russia Ukraine Russia Russia
Price June 2016: $3.79 $3.79 $3.60 $3.60 $3.60 $2.66 $3.60 $3.60
PART AI101E AI201A* I301A* AI301B AI301G AI301V
Manf Russia Russia Russia Russia Russia Russia
Technology GaAs GaAs GaAs GaAs GaAs GaAs
Spec Ext Ext Com Com Mil Mil
Fmax 2.3GHz - - - - -
Peak point current, Ip in mA - 9-11 1.6-2.4 5 - 4.5-5.5
Peak point voltage, Vp 4.5-5.5 0.2 0.18 0.18 - 0.18
Projected peak-point voltage, Vpp 0.16 - 0.65 - 0.6 -
Projected peak-point current, mA - - - .85-1.1 50.05 -
Peak to valley current ratio, Ip/Iv - 10 8 8 - -
Forward voltage, CW, Vf - 0.4 - - -
Reverse voltage, CW, Vr 6 - - - 0.27 -
Valley point terminal capacitance, pF 0.5-0.6 3.5 <12 <25 0.4 -
Inductance, nGn 2-6 1.3 <1.5 1.5 - -
pulse forward current IFp, mA - - 1.2 - - -
dc cw forward current IFc, mA - - 1.2 1.2 - -
dc cw reverse current IRc, mA - - - - 1 -
Impedance, Ω - 8 - - 16 -
Pulse power, mW - - - - - -
Temperature range, °C +85 +85 +70 +70 - -
eBay sold from: Ukraine Ukraine Ukraine BulgariaBulgariaRussia
Price June 2016: $3.99 $8.99 $1.51 $1.30 $1.54 $6.98
Part no. also listed:
PART 3I306E 3I306G 3I306K 3I306J 3I306L 3I306M 3I306N 3I306R
Manf Russia Russia Russia Russia Russia Russia Russia Russia
Technology GaAs GaAs GaAs GaAs GaAs GaAs GaAs GaAs
Spec Mil Mil Mil Mil Mil Mil Mil Mil
Fmax - - - 10MHz - 10MHz - 10MHz
Peak point current, Ip in mA 1.8-2.2 1.8-2.2 5.5 4.5-5.5 11 11 9-11 4.5-5.5
Peak point voltage, Vp 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17
Projected peak-point voltage, Vpp 0.85 0.85 - - - - 0.85 -
Projected peak-point current, mA - - - - - - - -
Peak to valley current ratio, Ip/Iv 8 8 8 8 8 8 8 8
Forward voltage, CW, Vf - - - - - - - -
Reverse voltage, CW, Vr - - - - - - - -
Valley point terminal capacitance, pF 4-12 <8 8-25 15 12 30 15-50 4-25
Inductance, nGn - - - - - - - -
pulse forward current Ifp, mA 2.4 0.8 1.2 0.9 0.4 1.2 12 1.2
dc cw forward current Ifc, mA 1.8 0.8 0.9 0.4 0.4 0.9 9 0.9
dc cw reverse current Irc, mA 4 4 10 10 20 20 20 10
Impedance, Ω - - - - - - - -
Pulse power, mW - - - - - - - -
Temperature range, °C -60+100 -60+100 -60+100 -60+100 -60+100 -60+100 -60+100 -60+100
eBay sold from: Ukraine Ukraine Russia Russia Russia Russia Ukraine Russia
Price June 2016: $1.36 $14.16 $12.26 $1.48 $12.25 $1.48 $14.61 $1.48
Note - One auction says 3I402B / 3I402G = backward tunnel diodes - are the rest too?
Part no. also listed: 3I306S - 3I402B 3I402D 3I402I 3I402V
PART - AI402B AI402D AI402I AI402V
Manf Russia - Russia Russia Russia Russia
Technology GaAs - GaAs GaAs GaAs GaAs
Spec Mil - Mil Mil Mil Mil
Fmax - - - - -
Peak point current, Ip in mA 9-11 - - - - - -
Peak point voltage, Vp 0.17 - - - - -
Projected peak-point voltage, Vpp 0.85 - 0.6 0.6 0.6 0.6
Projected peak-point current, mA - - 0.05 0.05 0.01 0.01
Peak to valley current ratio, Ip/Iv 8 - - - - -
Forward voltage, CW, Vf - - - - - -
Reverse voltage, CW, Vr - - 0.27 0.27 0.27 0.27
Valley point terminal capacitance, pF 10-50 - 0.4 3.5 6.0 2.7-5.0
Inductance, nGn - - - - - -
pulse forward current IFp, mA - - - - - -
dc cw forward current IFc, mA 9 - - - - -
dc cw reverse current IRc, mA 20 - 1 1 8 1
Impedance, Ω - - 16 14 14 14
Pulse power, mW - - - - - -
Temperature range, °C -60+100 - - - - -
eBay sold from: Romania - Russia Russia Russia Russia
Price June 2016: $1.50 - $4.00 $6.98 $6.98 $6.98