Reducing the beeps from the JBL amp

I would bet that the bips are generated inside the amp following an "order" from the mfd. It seems that Mike made a nice approach of this behaviour.
My idea would be to use:
1 * 6708 IC
1 * any controller with a lot of programming.
1 * programmable audio controller (I2C)

the interesting point is :
with a total control of the avc lan,the "I agree" & so on could be managed; and we are nor far from controlling the MFD (visual & feedback).

 

I believe the generic name of the AVC protocol is IEBus. What we may need from Toyota is its implementation, in other words, the device IDs and their command sets. Much of this is already available from the diagnostic information.

As for managing AVC communications, you can manage it as much as you can manage conversations between a few people in a room. You can answer someone, you can eavesdrop on conversations. Unless you jam the airwaves, you can't prevent one person from talking to another and the other not hearing it.

From what I understand in my cursory look at the 6708 documentation, the chip can respond to calls addressed to its programmed ID, its ID group, or broadcasts. I don't think it has a permiscuous mode.

 

"Chuck, interesting concept. Probably would be easier designing our own AVC interpreter and then feeding its control to VCAs".

If building a AVC interpeter is easies, Yes. From there as you suggest control the VCA's.

"Would need to find VCF though for analog based EQ".

If we had the AVC iterpeter, this could be easy (my engineers always tell me it is easy for a sales/marketing person to say something is easy for engineering) as we can use the EQ control signal to control VCA's or some type of control circuit to control the 3 individual bands. If it were found to be difficult, the EQ control would be a feature I would be willing to give up and replace with a external equalizer, possibly in the space below the CD player.


"What did you mean by nav tone? If you mean the nav lady speaking, that goes directly from NAV box straight to front driver woofer".

I am refereing to the beep itself. Someone mentioned it was at 3kHz. A notch filter could be used to pass the tone from the AVC interpeter or amplifier to a mixing circuit with level control trim pot. I would also suggest that this signal be gated so no stray signals except the 3k tone be allowed to pass when it was above a preset level. We could do the same with the nav lady (no filter needed) to mix her into the analog signal rather than having to depend upon interrupting a speaker path. It would be easy (that word again) to design a ducking circuit that would lower the L-R input signal when the nav lady was speaking.

"One possible problem with your concept though. When the MFD sends a beep request to the JVC amp, that would probably cause all your VCAs to max out during that period. I suppose though you could use a notch filter to detect that beep and hold the VCAs' control voltages during that time."

The fixed frequency oscillator generating the reference input to the amplifier could be 1kHz or 500Hz and we would have notch filters before the rectifiers to make sure that that frequency only was getting to the rectifiers and controlling the VCA's. The tone should have minimum effect as it should be some 40db or greater down if steep filters were used..

I am floating this idea only as an alternative direction if your AVC to VCA control interpeter could not be made. My desired OPTIMUM device would be the same as yours; translate all of the AVC control signals including EQ and process and mix the signals in ANALOG (including tone and nave lady). Inputs would be the L-R signal from the head unit with separate input from the nav lady (and a separate i-pod input?) and the tone received internally from the AVC.

 

I'm posting my latest diagram about what I intend to make with brosnan's significant help. I think he and I can completely disconnect the JBL from the rest of the system with our device acting as a telephone that "drops" beep packets. It would also allow us to inject packets of our choice, as well as possibly acting as an "I agree" defeat. Big plans, I know, but I am pretty confident that it is possible.

http://gunn.ath.cx/projects/prius/avc-lan/...ages/layout.jpg

 

hello Mike & engunneer.
If you want to get completely rid of the JBL amp; it could be possible to translate the commands from avc_lan to I2C to control something like a TDA7416 or TDA7419 and use a standard amp.
That means , in your schematics ,replace TX2/RX2 by a "car audio processor".
So there would be no tampering of the analog signal in JBL amp and SynergyGuy would be happy :flame:
By the way, how do you manage the timing of TX1 to avoid collisions?

 

Frenchie,
I'm personally not interested in replacing the amp. Our circuit will output serial , that someone else could use to control a different amp. It would just be a matter of programming to convert all the different commands and codes both ways.

I'm looking forward to seeing what we can log for transmissions in the system, as we may be able to figure out what most of the commands are. Mike's already figured out alot of them, but I bet there are other interesting things out there to do with the AVC-LAN.

As far as TX timing, The spec is that if the line is clear for more than 100ms, then it is clear to transmit. I think we'll figure that one out when we get there.

 

Is this beep-firewall still needed?

I think I have production hardware that I can make a message firewall out of.

Which is needed more -- the raw audio out, before the amp, or the beep blocker?

Sean

Peripheral Electronics
www.peripheralelectronics.com

 

see next post...

 

Personally I think that raw audio out is more important, and I am very happy to see PIE considering working on this.  Note that if you tapped into the raw audio out from the JBL head unit it would naturally be beep free so you'd be killing two birds with one stone (my that metaphor sounds violent these days .
There seem to be a few problems with the audio signal going out of the JBL head unit:

1. It is balanced, whereas most aftermarket amps/processors expect unbalanced.
2. When sources are changed (e.g. CD to FM or to Tape), some static clicks appear on the signal during the source change.  At the same time a mute signal line uses a low voltage signal (between 2-3 Volts) to signal the JBL amp to mute the signal.  If you built an interface you would also need to do that muting.
3. The signal is only 2V whereas it would be better to get 4V or 8V output for noise reduction purposes.
4. The factory volume control doesn't affect signal level.

Sean FYI I've written a FAQ on the JBL audio system with some details at:

http://tom-morrow-land.com/tests/priusaudio/

and I'd be happy to answer any questions you had if it would help get an audio interface that gave a clean signal for aftermarket audio.  My email is in that document.  This really belongs in a different thread, so feel free to start another thread if you have more questions with the audio.