any SC4 / SC5 and .CMP compressed firmware experts

[TrunkLatcher]

Has anyone seen SC4/SC5 used to modify DSP or audio behavior?

How about the CMP file format, anyone extracted or parsed it before?

 

[merlin]

Far from an expert but try to explain.
Before the MaCom big man passed away he developed and posted the use of SC4/SC5 files.
These files tie into the firmware/DSP files just like doing an upgrade.
The point of them is to get the radio to do things it isn't capable of with plain programming.
Most common use is sliding the operational band pass up or down in frequency (IE: HAM mods).
Audio behavior is done with Maintenance and editing the tracking data. includes squelch/mute and deviation levels.
USE WITH CAUTION so as not to violate FCC laws and type acceptance.
The compressed flash files,, That is a big can of worms. It stores in flash at very specific offsets.
This is a part of the CPU memory map. Booting loads the bootloader into RAM memory as uncompressed executable code.
This passes control to the firmware then loaded partially into ram and parts loaded and executed on an as needed basis.
The compression algo is proprietary to MaCom and don't know that it has been used outside of a radio, or the the companies that actually wrote the software.
Last 2 pages may be enlightening for you:

 

[TrunkLatcher]

Has there ever been an example SC4 or SC5 file in the wild doing anything with the DSP?

Or just the ham OOB stuff?


End goal is to make the P7200 thru XG75Ps sound a lot better by removing or lowering the dumb ass high pass filter they put on the RX audio in digital mode

 

[merlin]

Nothing in the SC4 or 5 files works directly with the DSP. Again, these are soft pots adjusted with tracking editing.
Been my observation, none of the portables prior to XG100 had a really good quality sound.
The latest firmware is likely the best you will get.
One point that may help is the radio should be in very good alignment for any DMR.
That high pass filter is needed to reduce digital noise in the audio bandpass.
The TMS320 is working overtime in what it has to do.

 

[TrunkLatcher]

I am confident that the radio can be made to sound a lot better with the existing hardware

In the P7200 they are using the TLV320AIC26

The TLV320AIC26 does have bass / treble control

Interestingly, this HPF is not applied to TX audio. TX audio on digital mode is full range, with no arbitrary HPF

The mobile radios also do not have this HPF

I suspect what happened is the high pass filter for RX audio was selected primarily for Analog, where it sounds nice and clear, and that little to no analysis of it was made for Digital. It was simply never revisited or given much attention by Harris.

If you solder to the speaker and play music, you will see the speaker is capable of producing much more full range audio like an XTS5000

The question is where the high pass filter is implemented

Looking at the service manual, the DSP and CPU are all in one

There is a separate 16 bit ADC Codec chip that produces the audio, which can be controlled by SPI registers

I am _hoping_ that the HPF is being done in the Codec

More research needed

 

[TrunkLatcher]

I guess its time to get the logic analyzer on the SPI bus and see if this bit is set or not...

 

[BMDaug]

It makes me wonder if this is to stop some weird intermodulation in the AF range… whether it’s the plastic housing buzzing, or some resonance affecting the oscillator.

Could have been an oversight, or it could be a workaround… definitely more research is needed!

 

[TrunkLatcher]

If it does end up being the Codec doing it, then there are two options:

1.) try to find the constants in the radio flash for the commands it sends over SPI every time radio turns on to configure the Codec. Edit them. By seeing exactly what all the registers are set to with a logic analyzer, it may be easier to find these constants in the radio flash

2.) Come up with an implantable, extremely small microcontroller to hijack the SPI bus for a second and retroactively "Fix" the registers in the Codec after the radio turns on. Figure out the safest time (delay?) to do this by watching the SPI bus activity during and after normal radio boot.


Either way, this could be a fun project

 

[merlin]

Correct, the OMAP CPU also has the DSP core. Unless I am off base here, the DSP also serves as the codec which passes the audio datastream to the DAC. After hysterisis filtering, audio pases analog switches for muting and then to the AF preamp and output.
Test points with a test setup, I would scope the audio chain to see where any distortion becomes a factor.
Unlike some earlier radios, the analog audio in P7200 and up are not very healthy. Overdriven preamp, aged coupling caps can certainly be problematic.
Harris did away with the AF analog processor as found in Orion and M7100 that without argument had the best sound.
That backed up by the 28 watt TDA audio chip.
You can try changing some of the DSP init's and maybe do better, but I would still keep an eye open for early distortins and bad levels.
Just food for thought.
I would be interested what you come up with.
PS: A stand alone ADC codec would make this easier, it will have its own registers to manipulate.
Cheers

 

[merlin]

Here is something I came across, worthy of checking:

 

[TDR-94]

M/A-Com "supposedly" brought in a BOSE engineer to "tweak" the audio of the P7200. The bass/treble control in the TLV320AIC26 most likely applies to the user audio adjustments only available when the P7200 is operating in OpenSky mode.

L3Harris collaborated with Harman Kardon on the design of the audio section of the XL-200P.

 

[TrunkLatcher]

M/A-Com "supposedly" brought in a BOSE engineer to "tweak" the audio of the P7200. The bass/treble control in the TLV320AIC26 most likely applies to the user audio adjustments only available when the P7200 is operating in OpenSky mode.

L3Harris collaborated with Harman Kardon on the design of the audio section of the XL-200P.

Do you know if the low cut in P25 mode, where absolutely 0 low frequencies of human voice get let through making it sound like ****, is performed in the TLV320AIC26 ?

I'm hoping hijacking SPI bus for a second and fixing TLV320AIC26 registers can fix this

 

[TDR-94]

I don't know if the low cut filtering is applied there or not.

The speaker drivers they used had a tendency to distort badly at louder volume settings, especially with digital voice, like ProVoice and P25.The low cut filter is probably being utilized to help keep the driver from being overloaded at louder volume levels with digital voice transmissions.

The P7200 platform had nemerous issues.Distorted audio and volume issues were common with these.