Ham4CW
Administrator
A while ago I purchased a Xeigu X6100 transceiver as "spares or repairs" with an intermittent fault. The problem was that if the temperature of the X6100 was below about 14 deg C. it would not function correctly. Even just leaving the rig in a cool room would cause it to fail. The display would come on and everything would look OK, but there was no transmit or receive, and the waterfall display was blank (not showing any received signals).
Another curious property of the Xeigu fault was that once the rig had successfully booted up and was operating, even if the rig was then cooled down to near freezing point it would still continue to work.
So there is little information about these rigs on the web. A few folks have posted some items about things they have found and how they cured them, but Xeigu themselves give away nothing.
The only thing I knew about this fault was that it seemed to be related to the temperature of the PCB's within the rig. After a lot of experimentation I worked out that a clock pulse that was present on one of the connections to the power supply module was steady when the rig worked but it was erratic (or just like random noise) when the rig was not working.
I traced this clock signal to a small chip (marked 6965A) tucked away below a shield cover.
This chip turned out to be a programmable clock generator. I found that if just this chip was cooled the rig would not boot up correctly, but if I simply held my finger on it to warm it up the rig operated correctly when the power was cycled. The chip is used quite a lot in mobile phone applications apparently. I managed to track down a supplier and ordered two (just in case of any future issues!).
These chips are quite small, only about 4mm x 4mm, they have 24 pins (six along each side), and for added fun there is a larger pad in the middle of the chip which is used for a mixture of shielding and also cooling the IC.
The above photo is of the underside of the removed faulty 6965A, you can clearly see the large shielding/cooling pad in the middle. Don't forget that this whole thing is only 4mm across!
I also took a photograph of the PCB with the chip removed (see below). I placed the tip of one of my Fluke multimeter probes next to the pads to give some idea of the scale.
It was easy enough to remove the original IC, but this now left the problem of how to attach the new one since the 24 outer pins AND the inner large pad all needed to be soldered correctly.
After some thought I did the following:
1) Cleaned PCB pads and lightly tinned them with solder.
2) Added just a little more solder to the large central pad (since I would not be able to get to this once the chip was in place).
3) The underside of the new chip was covered with liquid flux.
4) The PCB was heated using hot air until the solder began to melt.
5) At this point I also began to heat up the chip (held in tweezers) so that when placed on the PCB it would not instantly cause the solder to solidify.
6) Once the chip had been heated and while the solder on the PCB was molten the chip was put in to place and carefully aligned with the pads before the solder cooled.
7) Once the board had cooled a little and the chip was being held in place by the solder under the large pad I then began to flow solder around all of the 24 outer pins.
After multiple checks to ensure all of the pins were soldered correctly the surplus flux was washed away using isopropanol.
The above photograph shows the chip after replacement (the small square just above the middle of the view). Also visible is the PSU module mentioned earlier. There is a row of five connections on the edge furthest away from the camera, and it is on the leftmost pin that the clock pulse should be found. It's something like 3.3V p-p (on an oscilloscope), while on a multimeter it shows up as a steady 1.6V DC (this is the average). When the rig is not working this voltage jumps around all over the place.
Here's another view after I had refitted the shields on to the PSU and also on to the clock generator section.
After all that effort I was rewarded with a working Xeigu X6100! After the chip was replaced it did not matter how cold or warm the rig was, neither did it matter how many days the rig had been powered down (I left it switched off for just over three weeks), the rig would always boot up correctly.
I've noticed that there were quite a few comments posted around the web about this or a similar issue, and it makes me wonder if the rigs may have been produced with a faulty batch of IC's or maybe more than an average amount were sub-standard?
I've tested the rig after leaving it to cool down to just below freezing point, and again it boots up correctly now (before it would take multiple power cycles before the rig would operate correctly.
Hopefully this information might be of help to someone.
73, Mark...
Another curious property of the Xeigu fault was that once the rig had successfully booted up and was operating, even if the rig was then cooled down to near freezing point it would still continue to work.
So there is little information about these rigs on the web. A few folks have posted some items about things they have found and how they cured them, but Xeigu themselves give away nothing.
The only thing I knew about this fault was that it seemed to be related to the temperature of the PCB's within the rig. After a lot of experimentation I worked out that a clock pulse that was present on one of the connections to the power supply module was steady when the rig worked but it was erratic (or just like random noise) when the rig was not working.
I traced this clock signal to a small chip (marked 6965A) tucked away below a shield cover.
This chip turned out to be a programmable clock generator. I found that if just this chip was cooled the rig would not boot up correctly, but if I simply held my finger on it to warm it up the rig operated correctly when the power was cycled. The chip is used quite a lot in mobile phone applications apparently. I managed to track down a supplier and ordered two (just in case of any future issues!).
These chips are quite small, only about 4mm x 4mm, they have 24 pins (six along each side), and for added fun there is a larger pad in the middle of the chip which is used for a mixture of shielding and also cooling the IC.
The above photo is of the underside of the removed faulty 6965A, you can clearly see the large shielding/cooling pad in the middle. Don't forget that this whole thing is only 4mm across!
I also took a photograph of the PCB with the chip removed (see below). I placed the tip of one of my Fluke multimeter probes next to the pads to give some idea of the scale.
It was easy enough to remove the original IC, but this now left the problem of how to attach the new one since the 24 outer pins AND the inner large pad all needed to be soldered correctly.
After some thought I did the following:
1) Cleaned PCB pads and lightly tinned them with solder.
2) Added just a little more solder to the large central pad (since I would not be able to get to this once the chip was in place).
3) The underside of the new chip was covered with liquid flux.
4) The PCB was heated using hot air until the solder began to melt.
5) At this point I also began to heat up the chip (held in tweezers) so that when placed on the PCB it would not instantly cause the solder to solidify.
6) Once the chip had been heated and while the solder on the PCB was molten the chip was put in to place and carefully aligned with the pads before the solder cooled.
7) Once the board had cooled a little and the chip was being held in place by the solder under the large pad I then began to flow solder around all of the 24 outer pins.
After multiple checks to ensure all of the pins were soldered correctly the surplus flux was washed away using isopropanol.
The above photograph shows the chip after replacement (the small square just above the middle of the view). Also visible is the PSU module mentioned earlier. There is a row of five connections on the edge furthest away from the camera, and it is on the leftmost pin that the clock pulse should be found. It's something like 3.3V p-p (on an oscilloscope), while on a multimeter it shows up as a steady 1.6V DC (this is the average). When the rig is not working this voltage jumps around all over the place.
Here's another view after I had refitted the shields on to the PSU and also on to the clock generator section.
After all that effort I was rewarded with a working Xeigu X6100! After the chip was replaced it did not matter how cold or warm the rig was, neither did it matter how many days the rig had been powered down (I left it switched off for just over three weeks), the rig would always boot up correctly.
I've noticed that there were quite a few comments posted around the web about this or a similar issue, and it makes me wonder if the rigs may have been produced with a faulty batch of IC's or maybe more than an average amount were sub-standard?
I've tested the rig after leaving it to cool down to just below freezing point, and again it boots up correctly now (before it would take multiple power cycles before the rig would operate correctly.
Hopefully this information might be of help to someone.
73, Mark...
