That's a shame. I don't know if it's software, firmware or hardware issue. As a last ditch effort. maybe try the system flush update then follow up with the standard Humax firmware (and install the custom firmware later), reboot, factory reset - just to see if it does anything different. This may be a daft question - but do the replacement tuners have the same part numbers?
Thanks for the interest. I have already tried all the above , but no change.
The basic part numbers are the same yellow label FT2115 21933720 SV2111 and orange label FT2116 21933730 SV2111, followed by 9 other digits, presumably serial numbers/date codes.
Last edited:
do the replacement tuners have the same part numbers?

I don't know if it's software, firmware or hardware issue.
The only way I can conceive this being a firmware issue is if there is a failed bit or bits in the flash in the relevant section of code, which "fortunately" does not cause a firmware crash. I have no idea whether Humax have included any built-in test, but when I was doing that sort of thing it was standard to checksum the EPROM (in my day) at boot.

If the flash is corrupt or failed, I doubt a reflash would succeeed (the write process should be followed by a veriify scan). For my money it is a vanishingly small chance this is a firmware issue (and I have no idea what you mean by "software").
According to my notes one of my HDRs has these
FT 2115B 21960610 SV2011xxxxxxxxx aerial in (L, when viewed from back)
FT 2116B 21960620 SV2011xxxxxxxxx aerial out (R, when viewed from back)
If the flash is corrupt or failed, I doubt a reflash would succeeed (the write process should be followed by a veriify scan). For my money it is a vanishingly small chance this is a firmware issue (and I have no idea what you mean by "software").
I think there is something about that here
Poking about with a voltmeter, pin 5 is tuner 5v supply, which appears for about 15 seconds, then drops to zero. Perhaps this is enough to get the initial strength, but then gets turned off almost immediately. More work required.

IGNORE Red Herring.
Last edited:
Just to confirm:
1. The signal strengths and quality shown when manual searching are genuine, not remembered - they vary when the aerial signal is attenuated or disconnected.
2. Pass through works fine.
3. Secret menu shows no signal on tuners 1 and 2.
4. 5 volts are present on pins3,5 and 10 as expected.
5. Antenna power is available, when switched on.
To me this indicates that the tuners are OK and the problem lies upstream. Next step is to dig out my trusty 'scope (not used for about twenty years) and look at the control lines to the tuners, although this will only tell me that something is happening or not.
As you may have gathered, this has become a purely academic exercise as I have picked up an immaculate replacement T2 for twenty quid.
Last edited:
Early results of 'scope probing over the Xmas break.

Tuner 1 FT2115 Antenna in, Tuner 2 FT2116 loop through r/f out.

Pin 1(r/f out) of tuner 1 is directly connected to pin 1 (r/f in) of tuner 2
providing loop through which is working. As the centre pins of the sockets are not directly connected, there is presumably some buffering/amplification between the sockets and the pins.

Tuner 2 FT2116 Antenna out
There is activity on the data and clock pins 6 (SCL Clock) and 7 (SDA Data) of the tuner's I2C interface during auto search, consistent with the search pattern. That is, lots of action when searching target channels and minimum activity during others, but no channels were registered. My old 'scopes trigger cannot cope with the erratic spacing of the data packets, so I can't resolve detail of the packets.

Tuner 1 FT2115 Antenna in.
No activity on clock or data pins during search.

Conclusion is that tuner 2 is used for the search, which seems to be happening, but nothing is being registered.

Tuner 2 was replaced by the old damaged unit and the above tests repeated.

This time tuner 1 was used for the search, with no activity on tuner 2.
Possible conclusion is that tuner 2 is indeed faulty and the search is performed with whichever tuner is available. Again no channels were registered.

Before I could work out my next steps, family pressure became impossible to resist, so further progress will have to wait until the new year.
  • Like
Reactions: /df
I eventually got back to my faulty machine yesterday and made some progress. Previously I had thought that the unit was searching for channels but nor accepting any. Having overcome the triggering problem with my scope, I can now see that instructions are indeed being sent to the tuners via the I2C bus both during an auto tune and a manual tune, but a "no channels detected" message is received. While I was recording the waveforms, I repeatedly set an autotune as this gave me lots of time to set the scope up and record many transactions. I decided that the tuners were probably all right.

I was just about to pack up for the day when I noticed that some tens of channels had appeared,, all labeled "unknown" and appearing in the 8oo series. All were completely scrambled and unwatchable. The hidden menu indicated that they were from Channel 24 (Winter Hill local) and 39 (Winter Hill SDN).

This morning, to my surprise, I successfully tuned all Winter Hill Muxes manually with no problem, everything in the right place and solid as a rock.

Suspecting bad connections or temperature problems, I spent the morning with a hair dryer and ice cubes in plastic bags, all to no avail. I even tried a method used by an old friend who used to fault find valve TVs with a wooden lollipop stick, loudly singing "Strangers in the Night" "Knock, Knock, Tap, Tap". He claimed it always found such faults!

Could all this have been caused by some corrupted data base which has been fixied by the current channel changes?

I will leave the unit on soak for a few days and see what happens.

PS. All the above was with the original tuners refitted.
A little further progress.

Removal and replacement of the tuners would eventually cause pcb damage, so I fitted sockets so that the tuners could be changed at will. Further work with the four tuners available proved that the fault was independent of the tuners, as I had previously suggested.

I quickly found that applying heat in the area indicated below using a hair dryer, would cause the machine to burst into life. whereas cooling with a small fan caused it to fail. Since I managed to tune successfully the fault has now become more like those high lighted by Luke elsewhere.
Because the heating/cooling test is not at all selective, I have tried various methods to localize the culprit with freezer spray and a soldering iron , but I am not getting anywhere. The problem is latency or hysteresis in that there can be a delay of up to ten minutes minutes before anything happens. Sometimes, once started, it may fail after several minutes sometimes it will run all day, everything working fine. I am running out of ideas.


  • heay.jpg
    475 KB · Views: 19
Last edited:
Yes, I have similar thoughts, but the supplies look identical in the failed condition or not. I did notice that Newcoppiceman's capacitor was running about 15 degrees hotter than all the others that averaged about 37 degrees, but decided that as it is associated with the 12 volt supply, it probably wasn't relevant. I might try changing it, more out of hope than anticipation.. I did give all the caps in turn a liberal dose of freezer all to no avail. There are so many of them that I find the task of measuring them in circuit somewhat daunting. Maybe when the weather gets better!