Hi all. Bluetooth stopped working, and after turning off the navigation screen turned on itself while driving.
Self-diagnosis shows
HFL NG
B1792 ERROR
Which means the module is over.
Research on the Internet showed that the input filter capacitor may have lost capacity or the solder balls have lost contact. I looked from the side and it was clear that the quality of the factory soldering could have been better, but first I unsoldered the capacitor and looked at its capacity. The capacity was normal 460 microfarads, but just in case I changed it to the same one, it did not give any result.

The next step was to put the module in the oven at 400 Fahrenheit for 10 minutes and see what happens. I got a completely frozen screen on the navi, as soon as I disconnected the module, everything worked, but when I connected back, everything froze. The stove didn't oven in my case.
Then I order stencils, solder balls and flux. It’s good that I have a Soldering Hot Plate and a heating station on my job.

Next, I desoldered all the elements, with Chemtronics Soder Wick completely clearing all the remnants of the old solder.


If anyone is interested, all BGA elements are on 0.45mm balls.
It's time to roll new solder balls



It turned out something like this, don’t judge strictly, this is my first experience in such work.



The lower heating ball is set to 200 Celsius, the soldering station is at maximum temperature and low air flow speed. After planting the chips, I moved each chip with a dental probe to make sure it returned to its place and the flux would do its job with the balls.

Do you know what happened after 4 hours of torment??????
The screen stopped freezing, but...B1792 again.
On the side of the chip, the balls clearly began to look good and there is nothing to say that I did anything wrong. But I also can’t know 100% that all the balls are soldered.
Now meet the new used module from eBay for $115. Yes, he is working.

I'm wondering what could have broken the original module? The only thing that comes to my mind is the failure of the chip due to overheating or the failure of the firmware. Again, perhaps the problem is again in the balls and need to re-roll the soldering balls again, but I don’t have the strength or desire to do this.
I took out the board and looked at it with a thermal camera after an hour and 15 minutes of Bluetooth operation.

I can draw one conclusion: the processor on the board gets hotter than anyone else; in general, it is cold and does not heat up above 100 Fahrenheit in the open air.
I couldn’t think of anything smarter than drilling holes in the old case and putting everything back in place.




What are your guesses as to what breaks Honda's HFL modules?

 

I have no guesses with respect to the unit failure. Mine still works, although I use it infrequently. Your writeup is interesting, and shows diligence. To be honest, I thought solder balls was a venereal disease... Anyway, good info for those with the time and tools to try it. Thanks.


RFT!!!
Dave Kelsen
--
Let's start referring to 'The War Against Terror' by it's initials.

 

My guess is the RISC processor chip itself is bad and I base this on the known problem with the HFL module draining the battery. The HFL module in my car was draining my battery and I found a warm spot on the processor (IR thermometer) when it should be in sleep mode and drawing little to no current. There's an Acura service bulletin about these modules drawing 250mA in sleep mode and draining the car battery or not working at all and the remedy is for the dealer to check and replace the HFL module if needed. For the processor chip to start drawing the higher current while in sleep mode it makes me suspicious there's a reliability defect with some of these processors in the HFL module and over time they start drawing more current and still function or they fail completely.

 

Most likely this is the case, but my HFL died without battery drawing, the car was eat 35mA. in sleep mode.