Deze pagina mag afgedrukt worden voor thuisgebruik. Commercieel gebruik of verdere verspreiding is niet toegestaan.
Dit is de afdrukbare pagina, klik hier voor het artikel op de volledige website.

Icom IC-706mkIIG power loss

Last year I bought an [Link:]: Icom IC-706 and within a couple of days I was offered to buy a broken [Link:]: IC-706mkIIG. The fault description was that the output power was unstable. Since this seems a repairable problem I bought the IC-706mkIIG. I sold the IC-706 since the 70 cm addition of the mkIIG is rather convenient for me. It was a guess if I could repair the ďnewĒ one, butís thatís a risk Iím willing to take. And the problem is found. A cold solder joint of the driver FET Q101 was causing the problem. Below is the process of repairing the IC-706mkIIG shown.
inpection and cleaning
The first thing I usually do is a visual inspection. During the visual inspection a couple of things are found. One of them is that the knobs are rather dirty. I donít like dirty rigs so I removed the knobs and brushed them clean using an old toothbrush, water and cleaning agent. The result is nice clean and fresh knobs. The before and after image is shown below. Then taking a look inside, some dust is found. Thatís no surprise for a radio this age. The filter board and fan are removed and cleaned using a dry paint brush and compressed air. The result is a dust free radio. (Luckily the radio is a non-smoker radio, otherwise the dust was much harder to remove.) So far so good. Thereís a 200 Ohms resistor found which was added to the fan circuit. Thatís a known [Link:]: fan modification. The fan is originally only turned on while transmitting. Since the rig generates approximately 30 Watts of hear during reception, therefore the rig tends to get very hot. By adding the 200 Ohms resistor, the fan is always running a little bit (and at full power when transmitting) and therefore the rig stays much cooler. The resistor mod was carried out not that pretty, so thatís a project for ďcleaning opĒ later.

fault diagnose
First was checked if the radio turns on and is reception is as it should. And that turned out to be fine as it should. The transmitter is checked and the power loss was intermittent as the original fault description said. The transmit signal frequency is as it should, so the synthesizer is in working order. Therefore the cause of the lack of power is likely between the local oscillator (LO) mixer and the antenna connections based on the block diagram (shown below). It turned out that the intermittent power loss is at all bands. Thatís an important clue. Since the HF and V/UHF final amplifiers are separated, the chance that both amplifiers are faulty is rather small. Therefore itís more likely the problem cause is else. Both local oscillator signals are checked and turned out to be fine. The integrity of the 30 MHz signal and the variable local oscillator signal is fine during receive and transmit. Even when the power drops, the signal is stable.

test setup
Itís confirmed that the problem is between the oscillator mixers and the antenna connections. Therefore itís rather logical to check the power amplifier and mainboard signal by separating both board by disconnecting the (red marked) TX coax. The test setup is shown below. Left is the signal generator, in the middle is the IC-706, at the right is the digital oscilloscope and a IC-7300 is behind the test setup for monitoring the spectrum.

mainboard check
The mainboard feeds the generated (and modulated) RF signal to be amplified by the PA board for the antenna. I connected a test lead to the mainboard, terminated it with a 50 Ohms resistor and fed the signal into a digital oscilloscope. Now the signal from the mainboard is visible during transmit. Since thereís no signal fed into the power amplifier, itís not sure if the problem is gone, hiding or not in the main board. Therefore the power amplifier should be checked at the same time.
PA board check
The disconnected TX coax to the PA board is hooked up to an externa signal generator in inject a suitable RF signal. I used the [Link:]: Rohde & Schwarz SMT 02. I generated an 14,2 MHz signal, modulated with 1 kHz audio tone in FM. The signal level was set to -20 dBm. During transmit, the signal level is increased to approximately 0 dBm. The power output of the rig was approximately 90 Watts as it should. During repeatedly transmit the mainboard signal, shown on the scope, war stable and the PA output power was not stable. The power output was fluctuating between 0,5 Watt and 100 Watts. The TX coax connection of the test setup is shown below. The white test coax is connected between the mainboard and the (50 Ohms terminated) scope. The red clip feeds the external signal generator signal into the PA board coax.

detailed fault search
Based on the test, the fault is at the PA board Šnd the problem is at all bands. That means that only common components for all bands are suspect of causing the fault. Based om the block diagram are only two active components; Q101 and Q121. Based on this knowledge, the radio is built together and turned over since the PA board is at the other side of the radio. The transmitter is activated and with a screwdriver the two components are mechanically stressed. Q101 reacts rather strong to a little force on top of the 2SK2854 MOSFET. It seems that the MOFET is in working order, but the soldering is likely cracked. Based in the visual inspection of the two driver MOSFETís, reflowing would be wiseÖ The problem is found and a bad soldering of Q101 is the culprit. The component is marked om the image below. This explains why the problem appeared more often when the radio became warmer. The MOFETís tend to get hot and expand an shrink due to thermal stress. After a long time this can cause a racked soldering. Since the soldering is not that neat around the components, I planned to remove the PA board, remove the MOFETís , clean the board and solder all the components back in place. The diagnose is complete and the repair is planned. To be continuedÖ

© 1984...2021 - - Afgedrukt op 2021-12-07T09:35:35+01:00