757

 Erwin Hackl OE5VLL said

"How it came about: Due to a momentary idea, I bought a defective FT757GX2 on ebay. In addition, there was the statement "Frequencies are displayed, noise comes from the loudspeaker, no test option". Of course, you can get such a device much cheaper than a fully functional one, but it needs to be repaired and I wanted to accept this challenge. First function tests: A first simple reception test showed that the device apparently works, but a station that was well received on a comparison device could not be heard on the 757. It soon turned out that the receiving frequency deviated from the displayed frequency by a few kHz. This made it clear that the crystals that determine the frequency were already showing major deviations due to their age of around 30 years. For common crystal oscillator circuits, this means that the series capacitance must be reduced (parallel to the tuning capacitor).

Frequency processing: If you start to “crank” the frequency-determining components of a radio set, you should know what you are doing, as otherwise it can very easily happen that there is much more “twisted” afterwards than before. Unfortunately, the service manuals contain adjustment instructions, but no frequency processing plan. But only such a display can offer an overview of the “frequency events” in the device. Now, I've done plans like this for radios in the past, so I knew that it can be quite a bit of work, but it gives you a very good overview of the device afterwards.

The tables above give a very simple insight. However, these say nothing about the actual details. Thus, on that occasion, I created a frequency processing plan. It is then included in the report in reduced form. This is for a quicker overview. However, it will also be available for download in the form of two DIN A4 PDFs. If these two A4 pages are printed out and glued together, you have an easy-to-read plan at hand.

Description of the frequency processing plan: This plan only serves as an overview of the frequency processing and is therefore not complete and is only shown schematically. The diodes on the connecting lines are only intended to show that switching diodes are used for switching, the arrows show the direction of the signal flow. The varactor diodes shown show that the frequency of the respective oscillator can also be changed using varactor diodes. For reasons of clarity, the XF1001 filter has been drawn in both the receive and the transmit branch - it actually only occurs once and is integrated into the respective circuit branch by means of switching diodes. The square fields correspond to those of the block diagram. The upper designation (e.g. "Q1021") also corresponds to the designation of the module in the block diagram. If there are two designations in a module in the block diagram, e.g. Q1006 and Q1007, only the first of the two is entered for reasons of space. It should also be noted that the four-digit designations of the block diagram (e.g. Q2017) are to be read in such a way that the first two digits define the plan (in this example the "20" stands for the "Local Unit Schematic Diagram"), and the " 17" for assembly "Q17" in this plan. It must therefore be noted that there is also a module "Q17" in the "RF Unit Schematic Diagram", which is shown as "Q1017" in the block diagram. It also happened to me a few times that I accidentally "was on the wrong plan". The colored stripes are for better and quicker orientation (red --> mixer, yellow --> oscillators and VCOs, green --> PLL). All quartz oscillators are summarized in the lower left area. Q2005 (15.000MHz) is the main oscillator for frequency generation. Among other things, it is the reference frequency for both PLLs. If this oscillator is not tuned to its correct frequency, virtually all of the radio's frequencies will be wrong. In order to prevent misunderstandings here, it should be mentioned that the lowest 500 kHz are not officially "available", but the device can very well be set to 400 kHz, for example, but "nothing is guaranteed anymore" there. The frequency processing also works in this range. That is why the frequencies are also listed for this range. 02/21/2017 Page 4 of 29 Repair report Yaesu FT757GXII

The oscillators: The device contains more than 5 crystal oscillators with a total of more than 7 crystals, since switching diodes are used to switch between two crystals in two of the oscillators. In the area of frequency processing, these are 5 oscillators or 7 crystals. Other crystals are used to generate clocks for the microprocessors, but are not relevant for this consideration. The main oscillator is "Q2025" with the crystal "X2006", frequency 15,000 MHz. Among other things, it serves as a reference frequency for both PLLs. In addition, its 3x or 4x frequency is used as a mixing frequency in the feedback circuit of the two PLLs. This oscillator is also responsible for the FM branch. The oscillators Q2009, Q2015, Q2016 and Q1024 are used to generate the frequencies for the various mixer assemblies. Oscillator Q2009 generates a 700 Hertz offset for CW (quartz 15.0007 MHz) and is used for FM modulation (quartz 15.0000 MHz) depending on which crystal is "switched on". This FM is also the reason why there is a second 15MHz crystal, as the other 15MHz crystal must not be varied in frequency. Q2015 is involved in generating the frequencies for the 2nd LO and the 3rd LO. Quartz frequencies: X2003: 6.7834 MHz for USB and X2004: 6.7866 MHz for LSB according to plan. In fact, however, this oscillator is responsible for generating three frequencies. The third frequency is achieved by connecting a capacitance diode, hence the circuit symbol for this diode in the frequency processing plan. More on this topic below. Q2016 (quartz X2005 - 32.060 MHz) is involved in generating the frequency for the 2nd LO. In addition, the 100 Hz frequency steps are generated via this oscillator using a capacitance diode. In fact, even smaller frequency steps are generated, but this cannot be shown on the device's display. Q1024 (quartz X1001 - 8.670 MHz) generates the mixing frequency for the 3rd mixer in the receiving branch. He is also involved in generating the mixed frequency for D1075 (demodulator for SSB and CW, also called product detector). Its frequency can be varied by approx. +/- 2.9 kHz with the "Notch" potentiometer. If the potentiometer is locked in the end position (counterclockwise), it delivers its nominal frequency.

The PLL's: There are 2 PLLs, whereby the first PLL (Q2031) is responsible for generating the 1 kHz steps, the second PLL (Q2042) covers the entire frequency range. Both PLLs use Motorola's MC145157. First PLL: The entire frequency range of 30 MHz is divided into 60 individual ranges of 500 kHz. Within each of these 60 ranges, the first PLL generates a frequency between 34.41 MHz and 39.40 MHz. The frequency of 39.40 MHz begins at the lower end of each of the 60 individual areas. With each 1 kHz step towards higher frequency, the first PLL generates a 10 kHz lower frequency until after 500 steps the lowest frequency of 34.41 MHz is reached. These frequencies are then mixed at 45 MHz (generated by the frequency tripler Q2028) and result in 5.60 MHz to 10.59 MHz, which the PLL IC MC145157 then compares with the reference frequency 15 MHz and uses it to control the VCO Q2032. Furthermore, the frequencies generated are fed to a 1:10 frequency divider (Q2035) after buffer Q2034 and thus generated at 3.940 to 3.441 MHz in 1 kHz steps. These are fed to mixer Q2037 and mixed at either 45MHz (0 to 14.999MHz) or 60MHz (14.500MHz to 29.999MHz). This creates frequencies from 41.060 MHz to 41.559 MHz (0 to 14.999 MHz) and 56.060 to 56.559 MHz (15.000 to 29.999 MHz). These are then amplified via separate bandpass filters and amplifiers (Q2038 or Q2039) and fed to the mixer Q2040, which already belongs to the second PLL. In addition, these frequencies can be tapped at measuring point 04. Second PLL: The second PLL Q2042 controls one of four VCOs depending on the frequency range. The frequency generated by the VCOs is 47.060 MHz higher than the frequency set on the radio. This 47.06 MHz is identical to the frequency of the first intermediate frequency (1st IF).

The small table above explains why there is a tripling or quadrupling of 15 MHz. The reason is that the PLL-IC MC145157 is not specified for higher frequencies at the f-input. With this "artifice" you stay below 20.5 MHz, which, strictly speaking, already exceeds the specification of the IC. The frequency generated is decoupled and amplified via the modules Q2055, Q2057 and Q2058 and is then available as the 1st LO after passing through a bandpass filter. In the case of reception, this is fed to the first mixer, in the case of transmission to the last (third) mixer. Further information on PLLs: If you are more interested in the function of a PLL, you can find the report on "SRD4000B from the Motorola transceiver cassette SLF3710A" at www.oe5.oevsv.at/technik/verbindungen/d-netz also described in detail how such a PLL component can be programmed. Receive branch: Depending on the type of modulation, different assemblies are run through in the receive branch. In addition, different frequencies are sometimes generated by the oscillators. This is taken into account in the following frequency information with "approximate information". Which exact frequencies are used is explained in a subsequent chapter. After the Rx/Tx switch, the reception branch leads to the bandpass filter either directly or via the optionally switchable 20 dB attenuator and a low-pass filter. This is used both on the receiving and on the transmitting side. Switching takes place by means of switching diodes (not shown). The received signal then reaches the first mixer stage (Q1006) either directly or via an optional switchable 20 dB preamplifier (Q1003). Here the received signal is compared with the signal from “1. Local Oscillator” (1st LO) mixed and then amplified in the IF amplifier Q1008 (47.060 MHz). After passing the 1st IF filter (XF1001 - 47.06 MHz), the 2nd mixer stage Q1010 follows, where the signal is mixed down with the 2nd LO of approx. 38.84 MHz. The indication "approx. 38.84 MHz" is because the frequencies are different depending on the type of modulation. This also applies to other assemblies. After passing the XF1002 filter with a bandwidth of around 7 kHz, the further signal path splits into four different paths.

AM: 3rd mixer in the receiving branch. The 8.215 MHz signal is mixed with 8.67 MHz by oscillator Q1024 and passed through ceramic filter CF1001 (455 kHz). The alternative notch filter is ignored here. After amplification in Q1017 and Q2018, the signal path splits again. With AM, the buffer Q1020 is run through and demodulated in D1068. The CMOS switch Q1030 (CD4066) transfers the signal to the filter Q1032 and then the paths of all modulation types meet again. This is followed by the LF amplification in Q1035 and Q1037 and then the LF is passed to the human ear via loudspeakers. SSB (LSB and USB) and CW-w: With SSB and CW-w reception, the signal passes through the narrow-band filter XF1004 (bandwidth 2700 Hz) after the broad-band filter XF1002. Then again the same way as with AM: mixer Q1016, filter CF1001, Q1017 and Q1018. After that, the signal is passed to the product detector D1075, which is actually just a mixer, but with the difference that the output signal is already the AF. Mixing takes place at a frequency of around 455 kHz, which varies depending on the type of modulation. It is made by mixing the signals from Q2015 (approx. 6.78 MHz) and Q2025 (15 MHz) which produce a frequency of approx. 8.21 MHz (3rd LO), this is transferred from Q1024 to the specified around 455 kHz down-converted and fed to the D1075 product detector. After that, it is transferred to the audio amplifier via the CMOS switch Q1030 and the filter Q1034. The further way again as with AM: Q1035, Q1037 and speakers. CW-n (CW-narrow or narrow-band telegraphy): With CW-narrow, the signal passes through the XF1004 filter, which has a narrower 600 Hz bandwidth, instead of the XF1004 filter. It should be noted here that this filter has a different center frequency of 8.2159 MHz instead of 8.2150 MHz. More detailed explanations on this in a later chapter. The further path of the signal is the same as with SSB but with different frequencies. FM: FM takes a completely different approach after the XF1002 filter. It leads via the Q1021 buffer directly to the Q1022 assembly with the MC3359 IC. This takes over the IF amplification and the FM demodulation and, after Q1023 and Q1033, passes the signal on to the AF modules Q1035, Q1037 and the loudspeaker.

Transmission branch: After amplification of the AF from the microphone (Q2001 and Q2002), the transmission branch is divided into FM and AM-SSB-CW areas. The AM-SSB-CW section consists only of the Q2003 amplifier. With FM, the further amplified AF (Q2007, Q2008) is fed to the oscillator Q2009 as a modulation voltage and impressed by a capacitance diode as a frequency modulation of the oscillator frequency of 15.000 MHz. Then these 15 MHz are mixed down with 6.78 MHz to 8.21 MHz in Q2012 and forwarded to Q2004 via Q2014. There the FM branch meets again with the AM branch. It should be noted that this is the oscillator Q2009 with the crystal X2002 and not the oscillator X2006 with the crystal X2006, which is the reference oscillator and the frequency may be varied! Both crystals oscillate at 15.0000 MHz! This is followed by the buffers Q2005 and Q1040. The signal is then mixed up in Q1042 with the 2nd LO (38.84 MHz) to 47.06 MHz and passes through filter XF1001, which is also used in the receive case. For the sake of a better overview, this filter is drawn in both the receive and the transmit branch, although it is actually only present once and is connected to the respective branch by means of switching diodes. After further amplification in buffer Q1090, the transmission signal is mixed one last time, namely in D1090 with the 1st LO (47.060 to 77.059 MHz) to the actual transmission frequency of 0 – 29.999 MHz. The signal is then amplified in the Q1049 buffer, passes through the band filter that is also used in reception and reaches the output stage via the Q1050 amplifier. This is followed by a low-pass filter and the directional coupler for forward and reverse power measurement. With the transmit/receive switchover, the end of the transmit branch is reached and the signal is emitted via the antenna.

Detailed explanation of the processing of the frequencies: The processes in the PLLs have already been explained in a previous chapter. Now, using frequency tables, among other things, an attempt is made to explain the detailed processing of the frequencies in this radio. AM: The table on the left shows the mathematical relationships between the frequencies. In the right part, the dependencies of the frequencies are represented by the arcs. Strictly speaking, the reception frequency Rx is actually dependent on the 1st LO and 1st IF, but was given the other way around here for the sake of better representation. It is essential that it is very easy to optically record how the frequencies are generated by mixing.

LSB: This table again shows the dependencies of the frequencies. The reason why the NF is 0 Hz is that only the theoretically occurring case of an SSB signal without modulation is shown here. In fact, the carrier is suppressed with SSB, but this is irrelevant for this presentation, since only the dependencies of the frequencies are important here. But you can also see very clearly that the dependencies in SSB compared to AM are already much more diverse and complicated.

The second table shows the mathematical relationships. It also shows the frequencies that result from five different modulation tones. Apart from that, the display "without sound" is only of a theoretical nature, as already mentioned in the previous table.

USB: The representation for USB is the same as for LSB, except that the frequencies are different for the other sideband.

The same applies to the second table.

CW-narrow: A very special problem arose for CW-n. After I had adjusted the oscillators "according to the instructions", all modulation types (AM, FM, LSB, USB and CW-w) worked except CW-n. I spent many hours trying to unravel the mystery, but concluded that CW-n cannot work with the frequencies specified for matching. In the adjustment instructions, the following frequencies are to be set for point "J2008", which is the 3rd LO: LSB with TC2005 at 8214.4 kHz CW with TC2004 at 8215.9 kHz USB with VR2006 at 8216.6 kHz LSB and USB OK. But if for CW (strictly speaking, only CW-narrow is meant here) is adjusted to 8215.9 kHz, then a frequency of 8215.2 kHz results for the 2nd IF. But this is outside of FiltersXF1003. Its specification is 8215.9 kHz and 600 Hz bandwidth, giving a frequency range of 8215.6 kHz to 8216.2 kHz. The resulting 8215.2 kHz can never get through the filter because 400 Hz is removed from the passband of the filter. After a lot of calculations and many attempts, I put the horse before the horse, so to speak, and calculated the frequencies from back to front, and came up with a frequency of 8216.6 kHz to be set for CW-w, which is the same as the frequency for USB is. When I set this frequency, CW-n also worked. If someone can prove to me that my considerations listed here are wrong, I'm willing to look at the arguments - I can't rule out the possibility that I've made a mistake somewhere. Side detail: It is also somewhat confusing that the frequency of the oscillator Q2015 is actually adjusted with the specified controllers with the crystals X2003 (6783.4 kHz) and X2004 (6788.6 kHz), but the measured by mixing with Q2025 ( SSB, 15000.0 kHz) or Q2009 (CW-n, 15000.9 kHz) resulting frequency of 8.21 MHz. See the table of dependencies below.

In the two tables below, on the left, the frequencies that resulted from my calculation, on the right, the display that results from the adjustment instructions.

This table shows the calculation of the frequencies. 0 Hz is only meant symbolically here, since there is no modulation at all with CW. It should only be expressed that the remote station is received with a deviation of 0 Hz. The columns to the right show how the frequencies behave when the remote station is received with a frequency offset.

FM: The same frequencies actually apply to FM as to AM. The only difference is that the Q1022 module with its own ceramic filter CF1002 (455 kHz) is responsible for the 3rd IF in contrast to AM.

Additional information on frequency mixing for SSB and CW: If two frequencies are mixed, both the additive and the subtractive products are clearly produced. The desired frequencies are only obtained through the subsequent filtering. As can be seen in the following drawings, the upper and lower sidebands can be "swapped" with SSB. However, this only happens if the frequency to be converted is mixed down with a higher frequency. The other 3 options are not mixed up! The following image is intended to clarify the reception conditions in the FT757GXII.

Transmission and reception frequencies for CW, SSB, AM, FM: There are always difficulties as to which frequency is actually used for transmission or reception. With AM and FM this is clear: carrier frequency is the nominal frequency (frequency set on the radio). This of course applies to Tx and Rx. If the signal is modulated, the two sidebands are added. CW: When transmitted, a carrier "appears" on the nominal frequency. There is no modulation of the carrier in that sense - "modulated" (the information transfer) happens through "carrier in - carrier out". USB: If there is no modulation, there is also no carrier signal, since precisely this carrier (and the second sideband) is suppressed. If a 1 kHz tone is transmitted, a carrier "appears" 1 kHz above the nominal frequency. If a 2 kHz tone is transmitted, a carrier "appears" 2 kHz above the nominal frequency. However, this apparent carrier is only a part of the upper side band in this case. However, since "only a tone" is sent in this example, the signal sent "looks like a carrier". LSB: No carrier like USB. But now the lower sideband is radiated. If a 1 kHz tone is transmitted, a carrier "appears" 1 kHz below the nominal frequency. If a 2 kHz tone is transmitted, a carrier "appears" 2 kHz below the nominal frequency.

The actual repair: As already described above, the first thing to be noticed was that the reception worked in principle, but stations that could be received on another device were not audible. A check with the measuring transmitter then showed relatively quickly that the frequencies were shifted by a few kHz. So it went to the frequency adjustment. Of course, we started with the reference oscillator Q2025, quartz X2006, nominal frequency 15.0000 MHz. The problem with the signal being too weak for the frequency counter and its solution using an MMIC amplifier is described in a later chapter. The problem to be solved here was that the 30-year-old quartz could no longer be pulled to its target frequency. The remedy then was that the 20 pF capacitor connected in parallel with the balancing capacitor had to be replaced by one with a smaller capacitance. The existing 20 pF capacitor was pinched off on one side and twisted away, and after a few attempts a 2 pF capacitor was soldered to the component. The purpose of leaving the original capacitor on is that the original condition could have been restored relatively easily. In addition, with this procedure, the circuit board did not have to be removed and made accessible from the soldering side. The oscillator was then set as precisely as possible to its target frequency of 15.0000 MHz. In the picture you can see the balancing capacitor on the right next to the crystal and above it the new gray 2-pF-C. To the left of this is the brown original C with a black "cap", which is still soldered on on one side.

With the crystals X2001 to X2004, it was sometimes necessary to completely remove the fixed series capacitor in order to get the correct frequency. A special problem arose with the “Q2015” oscillator. According to the adjustment regulation, not only the two crystals X2003 (8.2134 MHz) LSB and X2004 (8.2159 MHz) CW have to be adjusted, but also a "third frequency" (8.2166 MHz) for USB, which can be set via the trimmer potentiometer VR2005 is. But the potentiometer was already at the stop (resulted in about 1.25 volts). A closer look at the circuit showed that the 10k potentiometer is fed via a series resistor of 47k ohms. Reducing the series resistor to about 10 kOhm brought the desired result and the "third frequency" could be set. At this point in time, I had no idea what problems there would be with this oscillator, but that has already been sufficiently described above. In the middle of the picture you can clearly see the small parallel soldered resistor.

According to the circuit diagram, the oscillator "Q1026" (X1001) has no component for tuning. However, it turned out that the series inductance does have a "compensation core" with which the exact frequency can be set. No component changes were necessary with this oscillator. 02/21/2017 Page 22 of 29 Yaesu FT757GXII repair report Checking the frequencies generated: In order to be able to check the frequencies generated in the device and also to deepen the understanding of the device, I created an Excel table in which I entered various frequencies. It turned out that something is wrong with the 1st LO signal. The spectrum analyzer showed that the signal was sometimes very "impure" and the amplitude was also very small in some cases. Especially in the upper frequency ranges and there, among other things, with the set reception frequencies around 23.120 MHz. The error was also audible (1 kHz signal from the measuring transmitter). The frequency processing plan, which has now been completed, was extremely helpful for troubleshooting. A quick check of the signal at test point TP-04 showed that this signal was OK in the event of an error. Now there was not much more to consider - basically only the second PLL. The cause could be narrowed down very quickly with cold spray – the cause was the “Q2040” mixer IC, a SN76514 from TI. The error could often be reproduced by warming up the IC with hot air. Subsequent cooling with a very small amount of cold spray ensured that the IC functioned correctly again. Since I didn't have this IC in stock, I ordered it from Hong Kong without further ado. When the IC arrived 11 days later, I wanted to reproduce the error again before replacing it. Interestingly, this was no longer the case. Various attempts did not bring the error to light again - maybe there is something like a visit to the dentist with ICs - when the time comes, it doesn't hurt anymore - hi. According to the principle "never change a running system" decided me not to exchange the IC for the time being but to put it in stock. If necessary, it can still be changed relatively quickly. Incidentally, this type of IC is also present in the other PLL circuit. The said IC (SN76514N) can be clearly seen in the center of the photo.

Dial lighting: The dial lighting was in a sorry state as shown in the lower left picture. Replacing the two light bulbs with LEDs shows the success in the lower right image. The two LEDs come from a disassembled LED strip. The series resistor of 150 ohms on the LED strip (intended for 12 volts, 20 mA) was supplemented by a resistor of 47 ohms, which gives the LEDs a current of around 23 mA at 13.5 volts, but this is still in the range of tolerance is. They shine just a bit brighter than the 12 volt original. Since the LED strip can only be divided after 3 LEDs, the third LED was simply left in the device - so to speak as device interior lighting - hi. Saving electricity is not essential at this point, but the following should be noted. The two bulbs used 160 mA at 13.5 volts, which results in 2.16 watts. The 3 LEDs including the series resistor consume 0.31 watts.

Additional "service" on the radio: The four double potentiometers (e.g. Mic and Drive) on the front panel sometimes cannot be operated in such a way that the second axis does not rotate. This is not only annoying under certain circumstances. It is usually caused by gummy fat or oil. In such cases, removing the knobs and oiling the axles a little is a good idea. After the subsequent installation of the rotary knobs, the device can be used properly again.

Lowest possible transmit power: Since I bought the device specifically for WSPR activities, power control was a very important feature of the device for me. Measurements have shown that the transmission power can be reduced to less than 10 milliwatts with the “drive potentiometer” (maximum transmission power 100 watts). 48 µWatt were measured for FM and 21 µWatt for CW. This should suffice for all low power requirements in WSPR activities. "Activation" of the device: In order to be able to operate the device "broadcasting", e.g. in the 60 m band that may be released in the future, the device must be "activated. This is done by a switch. However, this switch is not shown in the official plans and is very well hidden in the device.

In the left picture you can see the open radio from above, front left. The location of the "unlock switch" is below the orange component. On the right picture an enlarged picture. You can see the black switch between the arrow and the orange component. Position of the slide switch here on the picture: Above is unlocked, pushed down is original condition. 02/21/2017 Page 25 of 29 Repair Report Yaesu FT757GXII Attention trap: no transmission possible! After I had already reassembled the device and put it to its preliminary use, a "new error" suddenly appeared: sending was not possible. The red transmission LED on the display lit up, but no HF could be detected at the antenna socket. After checking all the settings on the front panel of the device, I came to the conclusion that, maliciously, a new error must have occurred. While searching for the error, I came across the Q1042 / Q1043 mixer. Both "coming" signals (IF at 8.21 MHz and 2nd LO at 38.84 MHz) were there, the mixed product (47.06 MHz) was detectable, but far too small. Since there is a very similar mixer in the reception branch (Q1010 / Q1011), I decided to use this for comparison measurements. The DC voltage measurements on the two transistors showed that when receiving, a voltage of about 0.1 volts was measured at the gates. When transmitting, however, a voltage of -3.44 volts occurred. With the other pair of transistors, this negative voltage (here -4.4 volts) could be measured in both cases. The search for the cause then led via Q1063 to the CMOS switch Q1050. Here, an "incorrect high" of approx. 1.5 volts could be detected at pin 9 of the MC14011, which should have been at least 5 volts to cause a switching process at the gate. There are 5 possible signal sources as the cause of the error. Q1059 and D1059 could be ruled out and the two diodes D1097 and D1098 were not even present in the device. So only D110 remained. The signal tracking then led to board 2 (local unit) and on this board to the Q2073 assembly. Among other things, this is controlled by the "linear switch". This can be operated from the back of the radio and is used to control an external PA when transmitting without the internal PA also being controlled. When handling the device, I had apparently activated this switch without noticing it. This is a so-called "push button". This works like a ballpoint pen, press once is "ON", press again is "OFF". If it is "pressed", you can't see it that easily, since the button cap still protrudes a few millimeters in this case. Only when you press the switch again do you notice that it was "switched on". This means that the "wrong" switch position is not noticeable when you look at it normally. That's why in the headline "Caution, trap". This "source of error" is very annoying, because e.g. B. when lifting the device this switch can very easily unintentionally operate. So that this can't happen to me in the future, I blocked the switch mechanically.

Poorly accessible ("poorly connectable") measuring points: When measuring devices, it often happens that the targeted measuring point is difficult to reach, e.g. because the "connecting pin" of a transistor is extremely short or a single IC pin. So-called "micro clamps" have proven very useful here. With such clamp connections can be made, which for various reasons (too coarse, too heavy, etc.) are not possible with e.g. normal oscilloscope probes. If such a "micro clamp" is provided with a connection wire of approx. 1 cm length, a normal probe can then be connected to it. These clamps are so small that they can be used to clamp a single IC lead "in the middle" without touching the neighboring leads. However, this does not apply to the actual connection process. The power supply of the device should be switched off under certain circumstances, since unusual short circuits can have unusual effects. I know from experience that smoke that has escaped from an IC, for example, can hardly ever be brought back into the IC. In such cases, it's stupid if you don't have a replacement IC at hand. These two pictures show the use of the micro clamps. In other cases, it can also be helpful to solder a piece of wire to the desired location for the measurement work. However, one should not forget to desolder it after completing the work. If measurements are to be carried out on cables whose ends are not accessible or only with great difficulty, a pin pushed through the insulation can be helpful. Of course, this only applies to single-core cables. But since pins often only have a relatively coarse point, I sharpen them on a bench grinder. These can then still be used relatively well for very thin cables. Such needles can also be used as an "extended measuring tip" for probes.

Problem: Signal too weak for the frequency counter: In order to be able to measure the frequency at specific points with a frequency counter, the signal had to be amplified. Therefore, a simple MMIC amplifier was built for this purpose on a circuit board that was already “in stock”. These boards are designed so that MMICs including associated (SMD) components can be assembled within minutes. The MMIC was selected based on two criteria: 1. The frequency that is to be amplified. It should be noted here that many MMICs are only specified from 0.1 GHz, so it was necessary to choose a model that amplifies best from DC (0 Hz). 2. The gain should be as high as possible. 3. The MMIC should also be in my possession. The MSA-0686 (marking code "A06") was found. This MMIC is specified for DC - 0.8 GHz, Gain 18.5 dB. In practical use as a measuring amplifier, it has proven to be beneficial for the intended purpose that the circuit board was quickly soldered to a shielding plate inside the radio with its own ground surface (see the two soldering points on the right edge in the photo). The connection to the measuring point was then made using a short piece of wire (solder point in the middle on the left). The frequency counter at the center right edge. The circuit board has the dimensions 37.4 x 37.4 mm. Apart from the IC, only two capacitors (on the input and output side), a resistor and two solder terminals are soldered on.

It later turned out that the frequency measurements with the spectrum analyzer could also be carried out without a measuring amplifier, which greatly simplified the measurements on the device. 02/21/2017 Page 28 of 29 Yaesu FT757GXII repair report Closing word: Why is such a report actually written? One reason could be that you want to show others what you can't do. But to achieve that, the effort would be quite high. There are also plenty of others who can do it far better. However, it is a certain self-affirmation if you were able to successfully complete such a repair. Afterwards you just have the good feeling of having achieved something again. But that alone would not be enough. This is where the “social streak” comes into play. If you already put in a lot of effort ("the repair"), then you can push the effort even higher ("the report"). Then others will benefit too. In particular, "others" should also be encouraged to possibly tackle such a repair themselves and turn "scrap" into a usable device again. Especially in times of the "throw away society" you can set a "good sign" here. And it is precisely for those people that this report is intended to offer support and help. But also for those who are only interested in the inner workings of such a device. And who doesn't want to be among the "better informed". I would be happy if there were many OMs who liked or used this report. Feedback via email is welcome. Many thanks also to those who helped me to create this report with as few errors and inconsistencies as possible. Erwin Hackl OE5VLL wishes you lots of repair and information fun, email: erwin.hackl@pc-club.at "