Please note the decision to reverse two electrolytic capacitors was not lightly taken to sort the headphone audio. So please don't follow this blindly. Please make sure you are satisfied you have found electrolytic capacitors that have reverse voltage on them. I have added a photo to showing the two reversed electrolytics.
Intro:
I was investigating spurious emissions from an R1 mkII that are caused by U1 3.3v switch mode power supply and a switched line to control the LCD backlight brightness.
I concluded the antenna was so close to the above, pickup of spurious was unavoidable on FM and DAB (not such an issue with DAB which uses error correction).
R1 mkIIs use switched electronics to conserve power as there is a battery option.
The antenna is easily detachable so an external one can be used. I plugged in an external dipole via coax - French stations were coming in clearly.
So I thought I would give the R1 a once over and found the headphone volume was very low, unbalanced and distorted (See oscilloscope images). And I found this has further implications as I did the following.
1. I purchased another R1 MkII and ALL of the above was the same!
2. I traced the audio path (see diagram I came up with).
3. I found severe audio attenuation on capacitors C95 and C93 (confirmed 2.2uF 50v) electrolytics which feed the headphone amplifier.
4. The puzzle was why just these pair broken? There are may more of identical values around the PCB all working?
5. On the diagram I observed an audio combiner comprising R201, R96, R102 and R121. This takes stereo FM, DAB or AUX and combines stereo to mono for the TDA7265 amplifier to drive the single mono speaker. Importantly this combiner should be grounded by Rx8 and Rx9 10K resistors.
6. Because the combiner is grounded, you would expect any electrolytic capacitors to have their minus pins connected to the combiner. But C95 and C93 had + connected to the combiner and I was seeing about 6v reverse voltage on C95 and C93. The PCB silk screen was also indicating this was the way to put them in! So I put two new capacitors in the opposite way to what the silk screen was indicating, resulting in a much higher headphone audio output.
7. But that was not the end of it. The higher audio output was still severely unbalanced on the headphones.
8. C58 was attenuating audio going into the combiner. But C58 was connected the right way. I replaced both C58 and C59 4.7uF (L and R) as a pair and now all is working.
Conclusion
I think because C95 and C93 had reverse polarity, they started conducting and lifted the potential of the combiner that should be grounded. This put reverse polarity on C58 and C59 causing one of those to fail.
If you don't use the headphone output, then you might not think this is much of a problem. But the mono audio amp was receiving a highly unbalanced L/R input. This would surely impact the fullness of sound quality at the speaker?
I could not believe I have two units with the same fault and wonder how many more there might be.
Now I have got two R1 mkII's, I plug a high-end turntable into the pair of these and use them like wired speakers - left channel into one R1 and right channel into the other R1. They sound absolutely superb. Now the fault is fixed the sound level is identical on both for the same volume setting on each.