Audio line level to CAN bus converter: debugging pop/crack

[flybbya23]

Hello all,

First post here.  Extensive software background but love tinkering with electronics, but still consider myself a beginner: finally putting it to use.

I have a unit for my F-150 that takes line level audio in and ships it to the factory head unit stereo via CAN bus (my truck is old enough to not have a line in or bluetooth input).  Luckily a 3rd party created this nice little unit.  It's worked fine for years but recently the left channel started popping and cracking to the point where it is unusable.  Unfortunately, the 3rd party doesn't provide schematics (or should it be "fortunately" for my electronics/circuit learning plan?!).

In a nutshell, I plan on using this thread to document my debugging activity such that it may be useful to someone else in the future.  At the same time, I might get some pointers on where to go next.  I know that I might just be able to reflow the solder for all of the components and that just might do it, but I wouldn't learn much.  I have an old Wavetek 145 waveform generator and a Rigol DS1052E to work with amongst other odds and ends.

So far I have passively probed the audio input circuitry just to connect the dots.  Here is what I have come up with in a very general sense:

(1) The line level audio (which comes in through a pair of RCA jacks) signal (call it L+ for Left plus) channel is routed to a capacitor (fillin: value and where other side of cap goes [channel shield I think]).

(2) The L+ continues to pin 3 of a 4558 op amp (JRC).

Question: why does the line level audio even need to go through an amp if it gets amplified at the head unit?  Probably find out later.
Answer (found out later): It looks like the amp (what I thought was a pre-amp) is probably a circuit that converts single-ended line in to differential out.  The head unit has an optional DVD L/R +/- inputs: it looks like this is used as a psuedo line-in and the CAN bus is just used to tell the head unit that it's a DVD player playing?!

(3) More on the circuit later, but here are a few big parts on the board: Microchip MCP 2515 (stand-alone CAN controller with SPI interface), a Microchip microcontroller (there's a sticker over it so I haven't taken it off to identify it ... not sure if I'll need to know), a large multi-pin POT (fillin: number of pins) to adjust gain, a 10-pin (verify) MOLEX connector which ties into a harness that breaks into the Ford head (fillin: details of harness and head unit pinout).

(4) I have more of the op amp (line in pre-amplifier?) network mapped out, but not on me right now: will fill in later.

(5) I have the waveform generator set up to input a 1 volt signal into each audio channel on the board.  My thought was to put a known signal in and look at it through the pre-amplifier and see what I saw at different points: maybe nice and clean up to a certain point and then it breaks up, pointing to the culprit.  I looked up typical line levels and 1-1.5 V is what I found.  However, I decided not to trust that data and play a 440 Hz through my phone headphone out jack and look at it with the Rigol: it's only 400 mV peak to peak at the highest volume!

Question: is this 400 mV more typical of a line level and/or headphone output?  Is this what I should use as input to the pre-amp?  I also saw quite a bit of noise on the signal: expected?  If I use a filter on the Rigol will I miss anything?

(6) Just because I was interested, although I haven't mapped out the pre-amp circuit in its entirety, I did trace the outputs from the 4558: they actually go on PCB traces underneath the MOLEX connector to the head unit.  I thought they would go to the CAN chip or microcontroller so I was surprised.  Is the analog/amplified audio actually going directly out to the head unit and/or speakers?  I will be getting the stub harness between this unit and the head unit/vehicle harness out of the truck tonight which may answer the question.

Next:
(a) Retrieve converter to head unit/vehicle stub harness: map pinouts,
(b) Determine what I need to power up the card (I believe it's +12V on one of the MOLEX pins which would be driven from the , power it up and inject the test signal; probe pre-amp intermediate and output points; compare with right (good) channel.

[David_AVD]

I don't know a lot about CAN, but wouldn't it be controlling the audio instead of being a transmission method?

[flybbya23]

I think you're right.  I found out the following tonight:

Question: why does the line level audio even need to go through an amp if it gets amplified at the head unit?  Probably find out later.

Answer (found out later): It looks like the amp (what I thought was a pre-amp) is probably a circuit that converts single-ended line in to differential out.  The head unit has an optional DVD L/R +/- inputs: it looks like this is used as a psuedo line-in and the CAN bus is just used to tell the head unit that it's a DVD player playing?!

[David_AVD]

The radio in my previous car used to display some info on another display in the dash.  I'm guessing it used CAN or similar to do that.

[SeanB]

The digital side is used to emulate a CD/DVD player to the radio, so that it will then switch the audio in to it. That will not cause noise on audio. Most likely causes are the opamps , capacitors or dry joints.