TAS6754-Q1: 2in1 inductor for 1L modulation

Hi Kuramochi-san

  We have tried this way, and the cross talk at 20KHz could reach to -70dB, with the best inductor we could find. With a worse inductor, the results could be further affected. If customer could accept it, then there's no other side effect.

Hi Shadow-san,

Could you please share about test condition(schematics/layout/part#) and measurement result?

Best regards,
Kazuki Kuramochi

Hi Kuramochi-san

Tested on the EVM, to use 1 dule inductor replaced two channel's normal inductor. Part number is DBL8087H-3R3M-R.  Test results as below

Hi Shadow-san,

Thank you for sharing data.

This is just for confirmation but how do you think about how to connect 2in1 inductor to EVM?
The EVM seems that this board cannot mount 2in1 inductor directly.
So, I'd like to make sure for eliminating cross talk from wiring.

Best regards,
Kazuki Kuramochi

Hi Kuramochi-san

Kazuki Kuramochi said:

This is just for confirmation but how do you think about how to connect 2in1 inductor to EVM?

Have to use the wire, seems no other way.

Kazuki Kuramochi said:

So, I'd like to make sure for eliminating cross talk from wiring.

It's hard to provide solid proof. We do tested other inductor type with same method, like VAMV06077 3R3, the results as below.

Hi Shadow-san,

I appreciate to share data.

To close this discussion, could you please share about same board and channel combination's measurement data under individual inductor condition?

Best regards,
Kazuki Kuramochi

Hi Shadow-san,

Does your provide measurement result is from 2p0 samples?

Best regards,
Kazuki Kuramochi

Hi Kuramochi-san

Default results as below. This parameter doesn't change among different sample version.

Hi Shadow-san,

Thank you for providing waveform.

The measurement result at VAMV06077 3R3 is only measured with 1p1 or 1p0 sample due to measured data is 10/25/2023.
Is it right?

DBL8087H's measurement result is 6/24/2024 so I guess it is measured with 2p0.
Did this data measured at same board(not meaning same design board, meaning same board physically) of single inductor?

I'd like to make sure about how compare those data.

Anyway, I understand that the main concern is channel separation but there isn't any other serious concern(wrong fault detection, pop noise, oscillation etc...)

Best regards,
Kazuki Kuramochi

Hi Kuramochi-san

Kazuki Kuramochi said:

DBL8087H's measurement result is 6/24/2024 so I guess it is measured with 2p0.

2P0 comes back at around 10/5/204. So all tested with 1P1 or 1P0 then. All tested on the EVM.

Hi Shadow-san,

I understand that VAMV06077 is measured with 1p0 and the other condition are measured with 1p1 probably.

I made comparison as below.

It is little difficult to make conclusion due to single inductor has bad channel separation at middle frequency range(1k~8kHz).
Also other range's performance difference is less than 10dB.(it is almost less than 5dB)

Is this 5dB difference from inductor difference or chip to chip performance difference?
Why there is channel separation improvement at middle frequency range by 2in1 inductor?

Best regards,
Kazuki Kuramochi

Hi Kuramochi-san

It's a little hard to tell. We don't really spend lots of resource to research into this direction. Please contact Kevin if you think it's valuable to your customer, thanks.

Hi Shadow-san,

I understand our situation.

At this time, we generally think 2in1 inductor should have worse channel separation from the view point of theory.

However, we cannot issue any comment for how much impact is there.

Also, we don't think 2in1 inductor cause serious concern.

So using 2in1 is depends on customer.

Best regards,
Kazuki Kuramochi