TAS6422-Q1: Question when using part in PVDD=24VDC application?

Thanks Gregg,

You are citing our exact problem.  We are great at 12vdc application and we use 3 of these amp's in our product.  We are 3 years into the design and it's all done now and will enter production for my client.  We had to use transformers on the PTBL outputs to kick up the driver AC voltage to handle a much higher impedance and that was fun as the transformer appears as a DC short, being the primary is under 1/10 ohm.  But it does work well in the end as once it plays the tone the Z goes way up and the amp. seems happy with this layout. 

Do you also support the TAS6584 part?  I'm waiting on TI to open the door for me to get the full data sheet to review.

One thing we had to do was put a current monitor IC on the output of the transformers to log that the driver speakers were getting the design amperage/voltage thus wattage.  This is a coastal audio horn to warn shipping lanes, etc.  Coast guard has to approve the design and today logging if your device is working is mandatory for any ensuing court cases of liability as you can image.  It took a bunch of parts to put this on our board with the 6422 part.  I see the 6584 part has IV monitoring, question, is this what we could use to log those values as our affidavit of performance.  I.E. can we read those actual speaker values with my firmware (via. I2S back channel) to capture them?

And loving the PVDD of 45VDC, you guys hit the snag before your clients did...hehe...

We only need around 30w(rms) per amp. chip and we use 3 to spread out the thermal signature as we used the 6424 and it would cook itself doing two channels for us.  I am suspecting we could toss the transformers when running at 24vdc PVDD as we need to hit our driver horns with 28Vac(rms) to get enough watts considering the impedance of the voice coils and mechanical plenum resonate chamber. 

I think that running the 6584 part at the lower PVDD of 12vdc we would maybe have a problem with the high impedance to get to the wattage we need, that's why we used transformers (about 1:1.8 ratio) to kick up the Vac to drivers on the 642x part family.  It was close but shy of reaching the spec. wattage we needed at the drivers.

I hope I'm not overloading you here.  We did get TI help, I think you even, about 2 years ago on the design with the TAS642x family parts.  Didn't see the 24V gotcha problem coming back then though!

Any recommendations or assistance would be great,

Thanks, Marc

OMG, I just looked back in my email trail, you did help us back in 2019, wow, almost 5 years ago, how time fly's by.  The TAS642x parts back then were also "per-production" at the time.

Hi Gregg, thanks for the reply.  I think this is where we'll go, finish the board testing with what we have and immediately being a new board with all new TI parts.  We will use the WiFi MCU CC3235SF and the TAS6482 (times 3 parts) to replace what we have on our current boards.  Looks like an easy switch over.  I did notice the I2S in that MCU is different but should be fine, I think it is using only FSYNC and SCLK, we only used SDOUT, but we sure could add SDIN to read the IV parameters. (We had MCLK, SCLK, FSYNC and a data line.) Setup: 48KHz sample rate, mono mode, left justified, etc....

Is there a private way to share the schematic of what we are doing and the transformer spec. with you that wouldn't be on the forum for everyone to read?  I'd like to share what we finally settled on over the years and you helped on about 4 years ago.

Thanks for some quick ideas on moving forward and TI support finally opened the door so I have downloaded the full data book and EVM software (I don't think I would get that module, we would go directly to boards as the company I use is fast at this, Komaru Technologies in San Diego.)  And at this point we have pretty good amount of experience with the TAS64xx part and the CC2642R1 MCU.

Thanks again,

Marc Yaxley

Hi Gregg,

Just reviewed the full data sheet on the 6584/2 parts.  Wow, lodes of issues were address. I was concerned about how to sync. 3 chips to play in phase and stay, the new part addresses that.  Also I had A/D lines to monitor PVDD, now I can read that via. I2C.  Just to many new goodies to mention, great work there at TI, you must listen to customers as things I've dealt with are now addressed. And of course maybe I could get rid of all of my IV output monitoring, although this would be on the primary side of the transformer, but still relevant to what the driver/speakers are getting.

Cheers,

Marc

Marc,

I will contact you directly offline.

Regards,
Gregg Scott