• FT-736R internal SPSU repair
• TX key-up output power slope delay
• keying modification
• keying waveforms
• spectrum analysis comparing AF keying to CW keying
• 432 MHz RF RX filter
• FT-736R 432 MHz RF-front-end modification
• FT-757GX maintenance
• RF VOX (COR)
• Motorola AN762-based Broad Band HF Linear Amplifier with pair of 2SC2290
• Tektronix 492P repair tips
• Icom IC-703 S-meter calibration and receiver sensitivity

Unplug AC cord, remove radio's bottom cover, open 3 long M3 screws from the back securing the PSU heat sink bridge plate to rear chassis plate and those 2 screws from the bottom side in the middle holding the PSU base plate on the chassis. Iift out the PSU unit, open the PSU cage crews and the 4 PCB securing screws holding the base metal plate. I suggest you replace at least C8, C9, C12 or preferably all electrolytic capacitors on the PSU PCB with +105 C versions, but not necesarily the recitifier filter's 470 uF/200 V big caps, if they seem OK and you can measure their value or even better, the ESR too. Re-solder with new solder all the eroded solder connections on the PCB also, which means everything around the transformer. Remove tindrops (and possible solder bridges) with brush carefully. Reinstall the base plate, cage and secure the unit on it's place in the chassis and temporarily connect the AC cable and see if the unit now works. If it does, unplug AC cord and now reassemble the rest. Roll and shake the unit to see for any loose parts inside: the ventilation grill on top of the FT-736R cosist of wide slots which easly allow washers up to M8 size can easily fall accidentally inside the radio and cause severe damage! I have seen this on two occations. Remove any objects that do not belog there. The SMPSU should now work for for another 10...15 years.
 

Most of the VHF and UHF stations dedicated to weak-signal propagation modes such as MS, EME, FAI, etc, run high output power and use big coaxial relays for T/R switching. Those relays are slowish and take some time (up to 1/4 s) before they have completed switching from receive to transmit, while modern solid state transceivers and power amplifiers are capable to putting out full RF power within tens of milliseconds from keying. This has to be disabled, because hot-switching (with RF power applied) causes arcing of the contacts, causing interference, ruining the coaxial relay and being a potential cause for preamplifier failure. During relay contact transit from RX to TX position, the isolation given in relay's specifications will not be valid and this may yield a surge of RF, that ruins the preamp.

There are numerous ways to solve this problem, some of them involve use of auxiliary coaxial relay contact wired from the tower to the shack to drive a dedicated T/R controller unit. Another more easy way is to use a "blind" (no feedback) fixed delay in the transmitter, which in it's most simplest from, consists of a method of feeding a DC pulse to TX driver stage blocking the RF power output, until all the relays can be assumed to be settled to transmit positions.

This delay circuit was used successfully with Icom IC-211E since the 1980's and now has been applied to Yaesu FT-736R. ALC connection, whether the transceiver has such a connector or not, can not be used, because of it's slow time constants, so some modification inside the transceiver is required.
 

FT-736R

This gives transient free rise of the RF output and much lower initial power level (-50 to -60 dB from full output power) and is not sensitive by any means to DRIVE setting.
 

TX power-up profile of a modified FT-736R. The lower image is a 10-x zoom of the rising slope of RF output. The 12 dBm reference level is equal to 16 W through a 30 dB attenuator.

Between short dropouts to RX mode, the delay will be there, if the pause is longer than about 100 ms. Hopefully during shorter T-R-T trips the delay won't be needed. So the circuit is not perfect, but a lot better than it was before. Would need more work, and/or more components (a capasitor dischage switch across 22 uF driven from +RX)) .

This circuit is of general type and can be also used on most other transceivers. You just have to find +TX DC voltage and understand the TX driver stages and the RF output power adjusting circuit or the keying circuit, to know where to connect the delay pulse.
 
 

NOTE:

The TX Delay pulse generation circuit. p5 is the output for the TX blocking pulse. The 150 ohm series resistor is there to limit the current to prevent damage to other components. With other than FT-736 transceiver, you could consider using even slightly higher value. The value of the 22 uF capasitor can be increased up to 100 uF if longer delay time is desired. (1 nF= 1000 pF, ceramic or SMD)

FT-736R's TX unit: the TX Delay's PCB is installed vertically, the Q22 is in the middle right side.  MS-keying speed modification: replace C82 (4.7 uF) with 1 uF electrolytic capacitor - behind the white coil core just next to where the TX Delay unit's wire (inductor) is hooked.

Install the components of the delay pulse generator on a small piece of PCB, 30 by 30 mm and mount it vertically to the rear between TX and 144 MHz main unit (near connector J10), with a terminal lug soldered to this delay circuit PCB's ground and fasten the lug with the mounting screw at the corner of TX unit near AC input connector. Make sure the unit is firmly attached and can never get in contact with live AC connector terminals (will cause much damage! - and is a safety issue also), or other live parts on the AC input PCB, like the AC fuse!

Connect a short piece of wire from the TX Delay unit's p5 output of the delay circuit to the source of Q22 in the FT-736R's TX unit. This 4-lead transistor (FET) is on the right side of the unit between the shielded RF transformers. If you wish, you can use a small 10 uH inductor as jumper wire, to make sure no stray-RF can get to the Q22's Source and if you remove the TX unit to change the C82, you can use the (correct) vacant hole of TH01, that isn't there, to solder the other end of the inductor.

Connect the TX Delay unit's terminals p1, p2, etc. (only two shown) from the TX9V lines entering each power amplifier modules from their respective RF UNITs. The only way to make this and make it also work on SAT mode, is to wire the TX9V from the small PCBs that are soldered to the feed-through capacitors on the PA's shields. These small boards are not shown in the Technical Supplement, so I can't give you any pin numbers, but they are the leftmost (AC conn. side) of those 5 feed-through capacitors. Inside the cage I can see an orange wire connected to it. This is the correct spot to take the TX9V out, since the other wires leading into this unmarked PCB on 432 MHz PA, are at +9V all the time, when the SAT mode is selected. You may verify the correctness of the pin with a DVM. Thanks to Mike, N1JEZ for the SAT modification and pre-testing.

If you do not need this delay on one of the bands, do not connect the TX9V line from that band unit to this delay circuit. Inputs from only two bands (p1, p2) have been drawn in the circuit diagram above. By adding parallel diodes, all four bands can be included.
 

Adjust the 10 kilo-ohm trimmer to give you an adequate delay. For a HF-400 coaxial relay, a delay of 250 ms (0.25 seconds) seems proper. Raise the value of the 22 uF capacitor or the trimmer on the TX delay unit to make the delay longer, if necessary. An oscilloscope (preferably a digital storage scope), or a time domain sweeping spectrum analyzer connected to the 13 MHz RF-output connector at the rear of TX unit can be used to verify the delay and rise envelope of RF power.

One way to set the delay on 2 m is to verify if keying the TX causes any noise lines to appear on a VHF band 3, (180 MHz) TV screen. Too short delay with hot-switching will cause some noise pulses on the TV screen even with 20 W output. Increase the delay setting and re-test. When the picture is unaffected, add a little more delay and it is set correctly.

If you operate Packet Radio, set the TXDELAY in your TNC, or software long enough to wait, until RF power is up at full power and data may be sent. This reminds me some operators use brick RF amplifiers with RF-VOX relay switching on packet radio and this sometimes causes terrible noise pulses across the whole band. A delay circuit like this, would not be a bad idea to have in those systems too for many reasons.

Naturally, don't expect the Digital Squelch system in FT-736R to work with the delay in use, because the code is sent out, before the TX power rises up enough.

If you did not understand fully what was explained here, it might be better not to open the cover of you radio, but let someone else do the modification for you, just to be on the safe side.
 

HIGH SPEED KEYING AFTER C82 is ghanged to 1 uF

FT-736R high speed CW keying waveform with time sweeping spectrum analyzer, "R" at 1500, 2000 and 2800 LPM after C82 was changed to 1 uF. The upper practical limit depends on what is acceptable and readable by the receiving operator, so no speed claims are in order here, but it might be well over 2000 LPM.

Time sweep 2 ms/div (ignore the false label on the picture). Keying waveform at 3100 LPM, FT-736R on SSB mode, MS-Soft keying via DTR's 2300 Hz AF keying.

Frequency spectrum. FT-736R w. external DC power suppy, carrier only for comparison.

Frequency spectrum, average power on high speed CW via CW connector.

Frequency spectrum, average power on high speed CW via DRT's AF keying.

Frequency spectrum, peak power on high speed CW via CW connector.

Frequency spectrum, peak power on high speed CW via DRT's AF keying.

The AF keying causes less power to be spread to nearby frequencies, but it is a tradeoff with keying waveform. Using a monitoring receiver 20 kHz up at 144.080, the noise was at S3 on both cases, when noise blanker was on. Without NB the noise was at S6 using the CW keying, but there was no rise with AF keying, so the keying click impulses are extremely short in duration.

The first filter I used was part of a UHF duplexer (3-cavities) that was tuned to notch out 425 MHz. The filter has about 2..3 dB loss, which was insignificant since the filter is connected after the pre-amplifier(s). Since by nature a notch filter eliminates only a fairly narrow bandwidth, the unwanted signals some MHz below, still tended to cause IMD. To cure this, a commercial 4-stage UHF Helical cavity surplus filter was retuned to 433 MHz. The loss of this filter was about 6 dB, which still was tolerable, specially since 95% of the IMD stray signals disappeared from the receiver.

The frequency response of both filters at wide (left) and at narrow (right) spectrum sweep. The Helical filter covers the mostly used portion of the 70 cm band and attenuates just a little on the upper edge.

This modification met the target of getting rid of overloading out-of-band IMD signals received all over the 432 MHz band (see previous chapter) - the little there is left, is not generated inside the FT-736R. How much the dynamic range of the 430 MHz receiver improved, has not been measured, but I expect the maximum input signal level at 1 dB compression to be about -20 dBm (22 mV PD). QST measurement stated 3rd order intercept point to be -44 dBm on an off-the-shelf FT-736R on 432 MHz.

Basically, what should be done, are:
1. not using the RX PIN-diode T/R switching at front of the 432 MHz RX input when ext. pre-amp & T/R switching are used (the 2 cable (RX and TX) system)
2. replacing the FET-mixer with doubly-balanced diode mixer (DBM)

What you loose, is a little bit of sensitivity when you operate FT-736R on 432 MHz without external pre-amplifier, since the proposed passive mixer stage has a higher noise figure (Nf) (7 dB + 4 dB pad  and 3.4 dB of MAV-11 or ERA-5). Due to 1'st RF stage's estimated gain of 20 dB of the Front End unit, the resulting overall cascaded Nf  will result 2.35 dB, if it was 2.0 dB before (theoretical - not measured). On the other hand, if you do something for the PIN diode problem and succeed well on the modification, you may well win than back or make it better. With external pre-amplifier the system's noise figure depends in practice totally on the external pre-amp's noise figure.

I will not describe the modifications in detail, but as a hint: I have added a BNC connector on the back panel for direct RX input, which requires the back panel of FT-736R to be at least partially disassembled and PA unit's cover removed. I have installed a tiny RF relay to disconnect the RX signal path from connector to Front End unit from the PIN diode located inside the PA unit. The relay can be toggled by the pre-amp switch at the front panel to enable the normal use (single T/R cable) without ext. pre-amp input (the PREAMP switch wiring needs modification to do this, but you may never need it, so perhaps skip the switch thing). Accessing the 430 MHz P A unit (where the PIN-diode is and where the BNC is to be installed) at rear panel, is not easy due to excessive wiring to Power Amplifier modules, etc., but can be done, if you have patience and skills to do it. Caution: Unplug the AC-cord before opening radio's top cover for safety!

Modification 2. involves constructing a new separate 430 MHz band mixer unit and re-directing the RF signal from 1'st RF stage's output Helical filter to new mixer unit and feeding the IF from mixer unit's output to FT-736R's IF-unit and bringing the LO signal in to the new mixer unit.

In practice the work includes removing the 430 MHz Front End unit, opening the covers, removing C07, a 5 pF chip, soldering a thin (UT141) PTFE coaxial to ex. C07's Helical coil side terminal and braid to GND, preferably removing T01, which is the FET mixer output transformer to kill DC power to the FET mixer (this may not be necessary since both gates are at 0 V), adding a feed-through capacitor for feeding DC power in to the Front End unit, since the IF-cable at J03 will no longer be doing that job. The IF output connector J03 and LO input connector J02 are unplugged and are to be connected to the new mixer unit.

There are two ways to construct the mixer unit - cheap and little more expensive, but easier way. The easier way requires basically no access to RF measuring equipment (spectrum analyzer), if all goes well. I have not done it that way, but I see no reason why it would not work.

The cheaper way involves using discrete components, of which, a tunable IF coil, may need some tinkering.

Mini-Circuit's SBL-1, the DBM work-horse, costs only some bucks, and offers maximum of +1 dBm RF input level. It needs +7 dBm injection (LO), but works better with +10 dBm. Fortunately, the injection level at J02 connector is -2 dBm when terminated to 50 ohms and Mini-Circuit has a MMIC amplifier MAV-11, which gives 12 dB gain, so the we can get the +10 dBm with it, to the SBL-1. If you use ERA-5 type MMIC, you should get 7 dB more injection with it and may consider using some +13 dBm-level  mixer type such as RMS-1MH.

The mixer's IF matching network consists of an attenuator that fixes the mixer's output's poor match on frequencies other than 48 MHz. The broadband termination is necessary to help SBL-1 to work optimally giving low loss and good isolation from port-to-port, which it does when it sees 50 ohms at every port over the full bandwidth. The series-resonance LC circuit is to pass only the signals at IF of about 47 MHz to be further amplified, and the pi low-pass filter section further attenuates frequencies above 60 MHz,  which consists of numerous mixing products and the LO signal leak to IF port at about 385 MHz. Without these the MAV-11's would get a lot of undesired VHF and UHF signals from the mixer to their input and since they work up to 1 GHz, these unwanted singals would otherwise be also amplified increasing the possibility of a new IMD source to be generated as a result of poor design. The 10 pF capacitor at mixer's IF port was added as a result of experiment, since it raised the mixer output by a few dB.

To overcome the loss of signal from use of passive mixer and the output matching circuit vs. the original active FET mixer with presumably some 10 dB of gain, we use MAV-11s (or ERA-5s), one or two, as IF post-amplifier. The 50-ohm output (and input) impedance of MAV-11 suffers from slight mismatch when connected to FT-736R's 430 MHz band's RX IF input (T19), so the gain we get from the IF post-amplifier is only about 10 dB (2 dB below specs to 50 ohm match). The cause for using two-stage mixer post-amplifier when you are not using an external RF pre-amplifier, is the lag of some 4...5 dB of total system gain with just one MAV-11. If you have an external pre-amp, or two, you likely do not need the extra IF gain from the second cascaded MAV-11. Note: MAV-11 can be replaced with ERA-5 (or ERA-5SM) with higher output power and with 8 dB of more gain. Note: One ERA-5 IF amplifier stage gives almost the same gain as two MAV-11 amplifier stages!

Many of the discrete components used here could be replaced with integrated ones: the attenuator pad, the band-pass filter (Mini-Circuits PIF-50), the low-pass filter (Mini-Circuits PLP-50), but it raises the costs.

The diagram is close to what I have constructed, except for the additional second MAV-11 stage at IF. Mechanically, the components are mounted on two-sided glass fiber PCB, with 2.5 mm wide (strip)lines serving as signal paths and where the SMD and other components are mounted. The SBL-1 mixer and the adjustable 0.7 uH coil are mounted through holes on the opposite side of PCB which serves as ground-plane. If you can find miniature female coaxial connectors alike used by Yaesu in the Front End unit, installing them on new mixer (LO and IF ports) unit would make it easier to plug the LO and IF cables to it.
 

The PCB layout should loosely follow the schematic diagram. You find the pin numbers of SBL-1 and MAV-11 (and ERA-5) from the web at Mini-Circuit's site. Drill four ground-feed-through holes for wires around each MAV-11 MMIC's ground leads for good stability. Solder a brass wall around the PCB connecting both side's ground-planes and use a metal lid on the SMD component side. You may use a small pre-tinned metal enclosure made for small RF projects.

Tune-up: re-peak the last Helical coil of CV01 at Front End unit with signal at RX input before installing it's cover. Peak the 47.43 MHz IF series resonance circuit's coil (0.7 uH) located between the attenuator pad and the IF amp. MMIC. Now all should be in working condition. If you do not get noise or signals, check the DC voltages on MAV-11's outputs and inputs and look for shorts or cuts in the wiring and PCB. Check you have connected SBL-1's pins right (1=RF, 3 and 4=IF, 8=LO, all others GND) and MAV-11's with it's "index" at output side, but with ERA-5, the "dot" at input side.

When tested, seal the new mixer unit RF tight and secure it firmly on the right back corner by the 430 MHz RF unit.

Note: the MAV-11's draw about 50 mA DC current each from 8.5 V RX+ line and if you use 3 of them, Q28, the switching transistor on the 430 MHz RF unit has to take the strain to pass about 140 mA of extra current. If it fails, don't blame me.... specially if you caused a short circuit on the +8.5 V line. If it fails, a transistor with higher Icmax could help or add another PNP type switching transistor with its base and emitter in parallel to the existing transistor for delivering current from it's collector the new 432 MHz mixer unit.

Repairing the older version of Yaesu FT-757GX (so called mk 1, with rotary mode-switch) involves "some" work but, since the 757 is a good match  with 736R, I  bought a 20-year old  FT-757GX. During the first day I noted 3 serious faults, which were eventually resolved.

The usual things to do first, are cleaning the radio, replacing the blown dial lamps, treating all the switches, potentiometers and connectors with contact cleaner (CRC 2-26) - twisting&clicking them a lot to remove oxidation and then applying protective contact grease (Electrolube 2GX).  After this you can test how the receiver and transmitter works.

I had no problems with TX -  >100 W output on all bands (or so I thought), modulation and keying was OK, but the widely know fault of frequency drift and wobble by bad trimmer capacitor(s) in series with reference (and other) crystals on the LOCAL PCB board (upper) called for the trimmer caps to be replaced (TC06, TC04, TC05, TC01). In spite they shown no obvious contact problems, the frequency at times began to drift and wobble suddenly for no reason! Listen also for worn bearing noise from the PA blower - try to lubricate or replace the motor if noisy or if the fan doesn't do the short spin burst at all when the rig is switched on. If you need to open the PA's inside cover, check the antenna connector's center pin's jumper's soldering joints are OK, not cracked or loose. If the center contact on the Amphenol socket is loose, it may cause the jumper to break when antenna connector is plugged and un-plugged repeatedly, so use some glue on the inner side to secure it (or replace the socket...).

General inspection with a magnifying glass or a microscope of all PCB's solder joints is useful for future reliability, when a unit/PCB is removed for service. Cracked and bad-looking solder joints needs to be resoldered carefully - pay special attention to connectors and do not make solder bridges...

During troubleshooting and generally working on the receiver, to limit damages, I recommend you use a current-limited DC power supply set to about 1 A limiting, just in case an accidental short-circuit occurs.

You don't find the advice elsewhere on the web to fix these not-so-uncommon faults in FT-757GX!:

Everything else, but the FM receiver, worked - it was silent. Listening carefully at full volume I heard something - the squelch potentiometer worked OK - the problem was not actually a squelch problem, but the FM audio was just extremely weak. The reason was there was about +5 volts on the +8 V TX-line when receiving, which was weird. The RF UNIT has plenty of switching diodes which, when shorted, could have done this, since the switching transistor supplying the +8V TX line, was not shorted or leaky. With DMM you can try to find where the leak may originate by measuring exact DC voltages from the transformer coils, but this helped me none since the leaking diode was in the factory-made add-on modification wrapped in tape and located below the RF UNIT's back corner - close to the red +13.8V output RCA connector.

Those added components; two resistors, two capasitors and twodiodes, had one diode partially shorted, and it leaked DC to +TX line, which caused FM detector IC (MC3359 pin 15 had 5V!), to be muted - as if you were transmitting. The RF UNIT PCB can be flipped over for accessing the underside "modification", or doing whatever soldering work you need, by removing the keyer jack's nut, top PA-assy's front screws (2), loosening face-panel's side screws and sliding it out some, unplugging the RF UNIT's coax plugs at rear of the PCB and four PCB connectors and those 5 screws fastening the PCB. The glassfiber tape Yaesu had used for isolating the modification's components, had become very brittle which may have caused a short, or it maybe was just a bad diode (1N4148-type)? Anyway, all the original insulation work done with the tape needs to be re-done better with something, that is more durable over time! There are a few other places you need to do the same, like the cage over the LOCAL UNIT, which may short the PA fan wire terminals. When reassembling the PCB, make sure you have inserted all unpplugged cables back on their sockets and none are left under the PCB to cause short-circuit.

Measure the receiver's sensitivity on all bands; should be in the 0.2....0.5 uV (12 dB S/N, FM mode) region with RF AMP ON above 1.5 MHz.

If the FT-757GX receiver's sensitivity has worsened somewhat, like I had (about 3 uV for 12 dB S/N (FM)), there is something broken. If it is almost deaf, check first the light bulb (F01) by shorting it out. I discovered the RF-UNIT's BFP filter bank's input-side switching diodes 1SS97 Shottky's (D19, D07, D09, D11, D13, D15, + possibly others) had been damaged and leaked various amounts of RX signal to every BPF filter section. Real nasty to troubleshoot - the receiver was not suppose to have suffered such damage typical for contester's radios. Only trick to find out was to tune to 28 MHz and note the sensitivity, or S-meter reading of a carrier (from sig. generator) and then just cut D19's other lead and note the sensitivity & S-meter reading to improve (D20 is the respective diode on the filter bank's other side). I reconnected D19 and just for curiosity cut all the BPF's antenna-side Shottky diode's legs one-by-one and found all more or less RF leaky. If the 28 MHz BPF also has damaged diodes (MA190), the method might not work as assumed. Diodes are cheap and it's better replace many enough, so you get no second thoughts later. I used BA244 and some 1N4148 if no switching diodes are at hand. The damage was most likely a result of static surge from the antenna, but too much RF-power into the receiver from another transmitter (any band) can also cause such damage to any modern HF tranceiver. In fact this fault actually also reduced the transmitter's drive, since TX drive power is piped via the same BPF- bank and diodes - the Drive- knob had to be turned fully clockwise for 100 W output, but after repair about half-way setting was enough. The SWR metering seems not to be quite OK or is just designed wrong?: it indicates no reflected power (SWR reading = 1/1) when SWR is below about 1/1.7 or so. I do not know if this the case with all of this type of rigs.

Measured sensitivity on FM for 12 dB S/N ratio, RF amp ON, RF amp OFF:

The receiver LO1 level is +13 dBm, LO2 is -3 dBm from the coax plugs (unplugged, to 50 ohm load).

Last, calibrate the frequencies; the 15 MHz reference, then LSB carrier (TC05), CW carrier (TC04), USB carrier (VR05) and FM carries (VR04).  See service manual downloadable from the Internet for a full description of the procedure. The radio needs to be warmed-up, since it (the 15 MHz reference XO) seems to drift about 200 Hz during the first hour (no TCXO). I did not get the carrier XO frequencies just spot-on, but they are within given tolerances. The receiver's IF transformers may need a bit of tweaking, which gives you higher readings on the S-meter, which then should be calibrated (VR1 for 6 dBuV=S1 and VR11 for 50 dBuV for S9+60dB), verify  S9 = 50 uV = -73 dBm.

This comes from Ken, ZL1PKS: There were some problems in the PA board that were causing no tx/rx switching. The fault was eventually found to be the transistor Q(60)08 a
2SA1012Y. It was not showing the correct voltages as per the voltage chart - changed it for a TIP42C. On PA board 10 uF 25 V tantalum caps
C24, C27 and C32 were destroyed and showed some signs of burning.

Lower TX power on 28-30 MHz: 100 W on all bands 1.8-24 MHz but on 10 m about 60 W. Adjust VR(10)05 for power output on 10 m, VR(10)06 is output power for all other bands.

When everything works, tape, wrap and place all wirings carefully, specially the ones between PA and LOCAL unit, so they will not get pinched when you reassemble the PA and bottom cover. Negligence in this issue will cause damage to components when radio is powered and some wire(s) get shorted to ground (may happen even weeks later)!
 

If you are too lazy to hook up the PTT from the rig to a (homemade) brick RF amplifier and don't want to make a T/R sequencer, the above schematic diagram of  RF VOX or actually a COR (Carrier Operated Relay) may be useful, though it uses "obsolete" components such as mechanical relays and bipolar transistors. The 6 pF RF-pickup capacitor's value may be reduced if used on higher bands such as 144 or 432 MHz.

As you know, the relays, amps and neighbourhood don't like hot-switching. Some of the high gain amplifiers may not be unconditionally stable (break in to oscillation when unterminated) and start to oscillate fiercely the millisecond the input or output is left with no load. Such termination during T/R-switching is difficult to arrange, but if one switches-off the amplifier's DC power (or in some cases just the bias), it can't oscillate (and blow up). The second poles in the input and output relays are used to make sure both amplifier input and output have been connected before DC is switched ON by a separate relay (or a FET switch). When transmission ends the DC feed-relay is released immediately with no extra delay (except for the obvious delay the RF VOX needs for SSB/CW operation), while the 10 to 100 uF electrolytic capacitor holds the RF input and RF output relays still on transmit position releasing them maybe some 200 milliseconds (you can hear it) later to bypass the amp for receiving. The value of the capacitor depends on the relays, the amount of delay desired and other components.

Use common sense when selecting the relays (must have 2 poles or aux contact, at least one of them!) and make sure the 2N2905s can supply their coils without being damaged due to exceeding their maximum allowed collector current. The DC relay should not be excessively slow or you may have to add the delay to make sure things work in proper sequence. Do the first testst with amplifier DC power from the relay unconnected and  wil reduced RF drive, untill you know the VOX and the relays work as planned, or damage to the power amplifier may result! If you are afraid the relay wiring might pick up stray RF, use shielded cables and  add some small (1 to 10 nF) ceramic RF bypass capacitors where ever needed. Better be safe than sorry.

  The directivity achieved by trimmer resistor as terminator, was about 17 dB (at around 85 ohms). Insertion loss was negligible, 0.3 dB on 2 m (but closer to 1 dB on 70 cm). The circuit works through VHF and lower UHF, but it can not easily be calibrated for more than a single ham band. Power levels much above 50 W would call for shorter lines and wide gaps to reduce coupling and for higher bands the lines should be shorter. The RF connector's center pins are connected to the strip-line by short brass sheet jumpers. HP 5082-2800 Schottky diodes were used.
 

Mmatched pair 2SC2290MP was used instead of MRF454 due to price and availability. The amplifier worked when built with Amidon ferrites and 2SC2290 transistors per Granberg's application note, on all bands up to about 10 MHz, but higher up, the initial Pout was low - just 20 W on 28 MHz. To cure this the output toroidal transformer's (T3) primary parallel Mica capacitor bundle needed reduction, both 470 pF Micas were removed and replaced by one around 330 pF for best power on 28 MHz. Also slight changes in the T2 feedback loop's parallel caps increased the output power. As a last resort, the amp aimed to be driven by Icom IC-703 10 W radio, called for a 3 dB attenuator to be inserted at amp input. If you look carefully, you see the AN762 has holes and PCB traces for the pi-input attenuator, but as there would have been too little space for the resistors, I added a patch on the PCB, so the RF drive from the input T/R relay makes a longer route and the original attenuator space remained vacant. The attenuator was eventually bypassed with 820 pF capacitor to increase drive power on higher bands. The whole unit's bypass losses (also RX) were 0.5 dB on 14 MHz and 1 dB on 28 MHz.. The cabling and how you do it, affects, but also the ordinary PCB- type relays induce some losses.

IC-703 provides PTT output, which goes low during transmit. The necessary T/R relay drivers were built on separate PCB not shown here. That PCB also includes the SWR protection system's final end, from the sensitivity trimmer onwards. SCR triggers the T/R system when the SWR is too high and switches the T/R relays on bypass and +Vcc feed relay open. The reflected power sensor with a toroid, is better placed at the T3 secondary before the LPF switching relay RL3, if the LPF is selected manually and can be on wrong band by mistake. There is no extra space reserved for the SWR detector, but I used a small 2- sided PCB over the RL3 relay. The RF from T3 secondary gets looped via the detector and back to the amp PCB via short jumper. If the LPF switching is automatic, then the SWR detector can be installed on the coax that runs from T/R output relay to ANT connector. When the SWR protection (SCR) triggers, you need to switch off momentarily the amp from the ON/OFF switch to reset, or switch DC power off and on.

The LPFs uses 500 V Mica capacitors and Amidon toroid coils on both 5- element low pass pi-filters. The coils are made with typical toroid materials, the T50-0 may run little warm, so I would afterwards opt for the next larger toroid ring. The LPFs used here, were selected by the antenna in use; a 20-15-10 m tribander. This way and by using 2- pole relays, the design only has two LPF switching relays instead of 4 or 10. If a need arises to work on low HF bands, such as 7, 3.5 or 1.9 MHz, LPFs for those can be built on separate enclosures on antenna feeder cables. That way the LPFs are always "selected automatically" right, provided you switched on the right antenna and if you did not, the SWR protection will trigger. This scheme works, unless you have a 5- band trap-antenna, or something that is fed with just a single feeder on all bands.

Note: Ferrites I used in AN762:  T1: Amidon BN 61-202, T2: Amidon T 68-6 2pcs, T3: Amidon FT 50-61 14pcs, L3 and L4: Amidon T 68-6 one each, L1 and L2: VK-200 RFCs, Q3: MJE3055 (TO-127), D1: BD135. (C5 and C6 capacitance changed for best output on 28 MHz)
 

Use 1 mm copper wire or rivets on PCB feed-throughs and see that you get them all there. You may add some on the input and output ends of the PCB here and there, to make the ground planes work. Use adequate heat sink for proper cooling and preferably a copper bar plate as heat spreader under the transistors. Pay attention the PCB's stand-off's holes and those of the 2SC2290 flanges are positioned precisely. Test the bias circuit without the RF transistors first - you should get from about 0.7 volts up to little over 1 volt. When the unit is fully wired and PCB is secured on the heat sink, use some thermal conducting paste and bolt the 2SC2290s on, do not over tighten and use as thin layer of thermal paste as necessary. Last: solder the 2SC2290 leads on the PCB - avoid mechanical strain on the RF transistors. With current limited PSU, adjust the bias current to 200 mA and use reuced RF drive power when doing the initial tests and tweaking the amp. The unit I built, draws 21 A on low bands with 160 W output and little less on 14 MHz, but only about 11 A on 28 MHz where the achieved output in the end was about 90W.

Check the PCBs upper and lower side's patterns and holes matches - these images are edited from a hand-made PCB's photos - scale right and verify!
 

The AN762 application note can be downloaded elsewhere on the net with detailed description of the design and component values.

Please observe the ferrite material availability and the alternate Amidon ferrites used here.

1. about 10 dB loss of overall gain and thus calibration setting out-of-range when aiming for -20 dB full scale display on cal.
2. some (10 dB) lack of gain with 100 kHz BW filter setting
3. intermittent contact inside LO1 coupler
4. all numerical displays on screen are gibberish
5. severe lack of analyzer (Y) gain
6. noisy blower
7. modifying the input to have DC blocking (note: Mini Circuits BLK-18 limits lowest usable freq. to 10 MHz)

Useful tools: manuals,  a second spectrum analyzer with tracking generator (or at least signal generator and spectrum analyzer), a microwave bolometer power meter (HP-432), 8 mm wrench, extension card set (~$100)(3 cards and long SMB cables), DMM, ESR meter if the SMPSU has bad electrolytic capacitors. Depending on the faults to fix, you only need some of these tools. If you know the TEK 492 inside out, you need less tools, but still may need RF test gear. You may need spare parts; but buying a second analyzer is cheaper than buying a full set of spare cards and units for the 492P as they typically have price tags around 100€ each and there is a big stack of cards.

Loosen 4 big screws on casing's rear side's blue plastic stands and pull back and away the casing with the handle (the enclosure back side has very sharp edges, don't cut your hands!). If the screws are sealed, ask for a permit or any warranty may expire. If you need to replace the fan or repair SMPSU; at right upper back corner, FIRST: remove the PSU aluminum clamp securing inter-connector and carefully and gently pull open the blue flimsy rectangular multi-pin connector bolted inside the right side frame wall (see photo). Now, proceed by loosening the SMPSU unit securing screws (2+2) at rear side and pull PSU back and outwards - carefully!

Faults #1 and #3. The apparent loss of gain was actually accompanied with very bumpy frequency vs. level response, with dips down to -30 dB (see photo) - typical if having stubs along some RF path. Check the RF input N-connector to mixer signal path by opening mixer's RF- input SMA and sweep the attenuator array + other stuff on the RF path with another TG and analyzer -  got a fairly flat response? Click through the attenuator scale and see the response is still flat and level drop steps are as expected (-10 dB, -20 dB, -40 dB). If not OK, solve which part is damaged and repair it. Use a magnifying glass to see if the mechanical actuators on the RF attenuator have any cracks. If you see cracks, you may try to prevent them from breaking up by using very tiny amounts of Loctite Super Glue, but don't glue any of the moving joints solid.

If RF input feed to mixer was OK, check how power behaves at LO1 SMA output at front panel. If LO1 connector puts out flat power spectra across YTO tuning range, then use a spectrum analyzer (or power meter) covering 2...7 GHz and measure LO1 power level and it's flatness at mixer LO(1) input connector (from the SMA connected to mixer's LO port). Set 492P span at max. - slow or manual sweep. If you only have microwave power meter, such as HP432A, sweep manually through the whole YTO tuning range (whole band) and see if the LO1 injection power remains flat through the tuning range. If the level varies, and the unit has OPT. 3 installed, there are 2 couplers + the Bias-return slug on the LO1 feed from YTO. Find out which component is loosing LO1 signal and fix or replace it. Without OPT. 3 low serial # units; there is only LO1 coupler; check it does not have high loss. High serial # units all have the Bias-return slug.

Problems found: LO1 coupler has bad/intermittent contact inside; the SMA connector' center pins are not soldered to PCB traces. The PCB material does not allow it. Carefully remove the coupler, open it, but note on which side (top or bottom) of the PCB each of the flat SMA center pins go (mark them)  to and how PCB is oriented so you know how to reassemble the stack right. Use some slightly abrasive material and clean the PCB foils and SMA center pins for good contact, reassemble and test.

The faulty Bias-return slug caused a loss of displayed signal level and bumpy response curve; the input-side high-pass pi-filter's coil was cut at GND side and acted as microwave stub draining much of the LO1 power and causing loss of analyzer gain, etc. troubles.  Check with ohm meter the slug's input is close to zero ohms and output is 50 ohms, or feed some RF power of LO1 range (2...6.4 GHz) through it and measure how much it attenuates - it should not! Fix or replace or temporarily bypass it - the analyzer should work now if the LO1 to the mixer is now up at proper level and the mixer is not damaged.

The RF relays that make changeover switching of RF front-end's LPF vs.. preselector (YIG filter) between Band 1 and higher bands, may suffer from poor intermittent contact. If you find which relay is bad, and can open and clean it - fine, otherwise replace it.

#2. 100 kHz BW LC IF filters inside VR1 and VR2 units are de-tuned. You need the extension cards to tune them! Please be very careful when inserting any modules back on the back plane - the connector pins may bend if you force the card in wrong.  The Johanson variable trimmer capacitors might have contact problems. If you open the VR unit's covers, check the narrowest BW IF crystal filters do not have cracks on glass envelopes.

#4. A solder blob on PCB (manufacturing Q.C. issue) near card-edge connector inside the VR1 unit, caused one foil trace of the BCD data bus to be shorted to GND. Similar short may occur anywhere along the bus, so use DMM (or scope) and find it - it totally messes up the CRT numeric displays.

#4. this is from Glenn, K0BO, with thanks! Full Y-scale display can not be reached, but numerical displays are OK on screen; apparent lack of Y-gain. Detector&log amp. input level OK = 0 dBm, 1 kohm R4046 in log.amp unit is open, DC voltages wrong around the Q4035 stage (too positive) and this  amp. stage does not provide enough gain. Replace the 1 k ohm 1 W resistor.
Download the full desciption by Glenn!

#6, Noisy cooling fan. The original fan is run with AC from an internal low-voltage AC generator. Remove the old low-voltage 4-wire AC fan, unplug it and install new quiet PC-type 12 V DC fan of same size (note it blows air in to the analyzer, not out) and feed the new fan with DC from the -17 V DC line (fan's black wire goes to -17 V, red wire goes to GND).

#7. Modification: Add a DC block on the analyzer input. Get a Mini Circuits BLK-18 SMA adapter and install it inside, ahead of attenuator array. You may need SMA adapter and jumper cable to make it fit. This limits lowest frequecy response some, but it looks flat down to 700 kHz  (BLK-18's specs says 10 MHz), so if you need to use 942P way below 1 MHz, find some other DC block that is for lower frequencies and use it instead. The idea is to protect RF attenuator and 1st mixer from accidental DC feed to analyzer RF input.

You may find some de-tuning has occurred on 100 MHz Helical filters in the 100 MHz IF unit. Center the 100 MHz CAL signal carefully and use widest resolution filter (1MHz)  and peak the trimmer caps for maximum Y-gain on CRT screen center. Note the pass-band should remain symmetric and Gaussian in shape. Re-calibrate the analyzer at AMP CAL adj. screw at front panel to compensate for possible overall gain change. The full IF filter calibration includes setting Fc with 1 kHz filter at center, tuning every filter's response to symmetric Gaussian (LC) or flat-top (XTAL) in shape and then adjusting each filter-set's gain in VR units, so they all are all leveled. Don't touch the so-called "20 dB and 10 dB 1st and 10 dB 2nd" trimmers in the VR unit.

If you have microwave power reference to level calibrate band gains, they are adjusted from VR unit too. The Band 1 adjustment can be done with the internal calibrator, but for bands 2 and above, you need some microwave signal source of exact known level, preferably up to 21 GHz. Good RF power sources for lower microwave bands are those unlicensed low-power 2.4 and 5.6 GHz FMTV transmitters which exact output power is first measured with a calibrated microwave RF power meter (such as HP-432).

The other faults generally reported, are related to digital display memory, either H or V memory board. Errors 59 and 60 may intermittently show up sometimes and are related to phase-locking difficulties - if it goes away easily by turning span switch - ignore it, the design was not exactly perfect.

RF-to-DET&LOG AMP chain RF signal reference levels when a -30 dBm signal is fed to analyzer input connector, no preselector and no front-end attenuator is used:

Based on your previous experiences using contact treatment chemicals (oil and grease), you may apply them on trimmer resistors and connectors, etc. They will not remove oxidation without some abrasion, but they may prevent oxidation (to proceed) for a while by sealing small gaps and covering contact surfaces.

Spare parts and manuals are found both from US and Europe for a price, use Google to find, but be aware the serial number range issue; a manual for low serial numbers (below 2xxxxxx..) is only about 85 % valid when going through the calibration procedure with a unit that has serial # beginning with 4xxxxxx and note: the Service Manual does not have any schematics included. Some TEK docs are found on the web too at some free manual download sites (like KO4BB). Norway Labs in Oregon, USA, manufactures and sells extension card kit for the TEK 49x series, mail to Matt North. Spares are available at "Tektrotom" in Europe. The manufacturer's support has expired.

The above repair tips may perhaps apply also to TEK 494, 495 and 496 models. These equipment are so old you are bound to have some troubles (before long) and need to fix them, whatever promises you get when buying one. When purchasing an analyzer, get some warranty or R.o.R. Don't pay a top-price if the risk is all yours -  25 kg of electronic junk to dispose of, is not what you need. If the price is less than 1/3 of valid price, suspect it is a fraud attempt - never pay with Western Union wire transfer and never deal with sellers that have no name, no address, no bank account. Ask for equipment serial number and find out if it can be valid. Look up from web's forums if the dealer has a good reputation.

Incidentally, if you have found a second-hand non-US-made (Japanese) analyzer, it might as well develop or have problems, but you probably find no documents, no source for resonably priced spare parts if any and the only option may be the official service where the repair costs more, than you paid for the analyzer...very hard to keep such equipment  in working condition.

The new TTi PSA-2701T is handy and small, but lacks dynamic range, see the comparison image above.
The harmonics you see with it, are not there when you use an analyzer with more dynamic range.

Rev. 2010-01-07
© I. Yrjölä, 1999... 2010