LM4970 unused inputs?

Second question regarding the LM4970; what is it expecting to see on the audio inputs for levels? Currently, for test, I'm using headphones right up to the input mic. At the device pin (after the cap) it shows 400mVp-p. MIDGAIN is set to "maximum" (whatever that means) and summing amp is set to 12dB. This results in the PWM section of the device seeing 1.6V "peaks", no? I hate to say it (being a fanboy of TI and all); this data sheet is "seriously lacking" in info. 

For the record, settings:

PWM Freqt: 15KHz.

HPF cutoff: 3.5KHz. (with a 0.1uF cap for Cfilt).

Pattern select: audio sync. (I can blink the leds via I2C)

Current select: 2x, 2x, 2x

MGAIN: maximum

SGAIN: 12dB

Thanks.

Hey, TI?

So far, this post has been moved twice, so I figure someone there has at least seen it. You all may be far more concerned with getting DLP's or whatever other high dollar devices you might be hawking get addressed, but there those of use out here that have to design the $11.99 stuff as well. It may not seem like a big deal to you guys, but I've got my butt in a sling because I dropped this part on a board, supposedly to go to a show in March where about 6000 of our customers (a third of which are our distributors from around the world) will see it to place orders. However, it doesn't work, and I can see no logical reason why. If I had done something that was grossly wrong, I'd figure okay, my problem. But when a data sheet is so lacking that very simple questions like "What is the thing expecting to see for input levels?" goes unanswered I have to start wondering if anyone gives a damn whether these devices get designed in or not. This is a consumer item that would be retailed through Walmart, but if I have nothing to show them, I very well can't sell them, can I? 

To re-iterate; I2C is working; I have full control of the leds in that mode. Flipping the bit to audio sync does nothing. The only thing that I did differently than shown in the data sheet is to ground the "stereo" inputs. The data sheet says:

"Audio signals coupled in from MIN, LIN, and RIN are mixed together by an audio input summing amplifier". Grounding these inputs shouldn't make a difference, but nothing /  no one will tell me whether this is correct or not.

 I'm re-rolling the board to include the two "extra" caps such that all of the inputs will be fed from the audio input. I've jacked the input gain up so that the peaks are around a couple of volts, which increases the noise on the audio, but given that its blinking leds, I couldn't care less. I have four boards that are all behaving exactly the same way. Any help, at all?

Nothing? Nothing at all? Wow... so much for the "We try to answer every post in 24 hours or less", huh?

 Hello Mike,

It seems like an audio device issue. I will suggest you to ask question in the audio forum. It would be more help there than asking here in LED forum.

Best regards,

Victor

This is a joke, right? This is an LED DRIVER DEVICE! Is this or is it not the LED DRIVER FORUM? Did you even read the post? This post has been bounced around to three different forums. Here's what I suspect at this point; you guys look at the data sheet for something obvious, then post an obvious response. When there is a question outside of what's in the data sheet, it gets ignored, or as in this case, you pass the buck. 

Are you telling me that no one in the LED DRIVER support team knows nothing about this device? Something as simple as "What is the audio level input requirements?" This is clearly missing from the data sheet. The device WORKS UNDER I2C, but when a single bit gets flipped, nothing. And no one there has any insight? Wow...

Maybe if I were Samsung wanting to put this part in a cell phone? You'd tell me to go to another forum? I guess million piece orders aren't good enough for TI...