LM48511 : THD +N %

Hi Shantharoopan,

I will look into this issue and try to offer you some answer.

Andy

Hi Shantharoopan,

Here are some comments for you:

"To get 2W into 8 ohms load with 0.1% THD, it is theoretically impossible at 5V Vdd without booster. Even no loss on the Class-D portion, it is impossible. With no loss, supply voltage to Class-D amp will be around 5.66V. So if customer needs to get 2W with 0.1% THD into 8 ohms load, it needs to activate the booster and set PV1 to 7V. Then customer can get following performance.

But please note that if Vdd down to 3V, it can’t get 2W."

Andy

Hi Andy,

Thanks for your time and response.
The reason for not using the boost converter is that amplifier gain is adjusted by using TPL0202 Digital potentiometer. TPL0202 is supplied with 3.3V, hence I am unable to send the amplifiers feedback output (7.6V) to the digipot when boost is ON. When boost is OFF, am able to get 3.3V peak to peak from the feedback, which is ok for digipot. Also am able to get 2W output power but with 30% THD+N.
My customer requirement is to get THD+N% below 7%. So, may I know is it possible for me get THD+N% below 7, without using boost, with VDD=5V?

Thanks in advance,
Shantharoopan TSB
Phone : 96296-95476

HI Andy,

I would also like to know the realtionship between the Power supply to the H-Bridge and THD+N%, please help.

Thanks in advace,
Shantharoopan TSB

Hi Shantharoopan,

I will try to reply with some answers soon.

Andy

Hi Andy,

No problem. Would like to let you know that even for 250mW and 500mW the THD+N% is above 10%.

I have attached the schematics for your reference. Please provide your inputs.

Regards,

Shantharoopan

Hi Shantharoopan,

I don’t think it is possible to get 2W into 8 ohms load with THD below 7% on any Class-D amp. And without using the booster function, I don’t quite understand why LM48511 is chosen in your design. It will be better to choose a simple Class-D amp like TPA2011D1. But even with TPA2011D1, it is not possible to get 2W within THD=7%. If you really want 2W within 7% of THD, I will suggest you use 6 Ohms or 4 ohms load.

Andy

Hi Andy,

Thanks for your response.

I belevie you might have checked into the LM48511 datasheet, in which it says it would support upto 3W output into 8ohms at 5V with THD+N=1%.

Am very well surprised upon why LM48511 cant be used to get THD+N<7%. Am I wrong in some basics? Please help me regarding this and let me understand the relation between THD+N and power supplied to it (Boost) and Load resistance.

Thanks in advance,

Shantharoopan TSB

Hi Shantharoopan,

Take a look at the key specifications at same page and you will find the paragraph below.  The PV1 = 7.8V when Po is 3W with a 8 Ohm load, which means the boost is on.

Andy

Hi Andy,

Thanks for your kind response. As per our design, we wont be able to switch on the booster. 

Would like to inform you that even for 250mW is THD+N% is around 10%.

Isnt it possible to get THD+N% below 7% without doing any HW design change?

Moreover, please see the below graph. I guess its similar to my situation, but the THD+N% is well below 1% for 500mW even when filter is used. Whereas we are able to get 10%.

Regards,

Shantharoopan TSB

Hi Andy,

Can you please provide your feedback?

Thanks in advance,

Shantharoopan TSB

Hi Shantharoopan,

I will take a look at your question today.

Andy

Hi Shantharoopan,

My colleague took a look at your schematic. Here are some comments from him:

1) The customer uses TPL0202 as gain setting for LM48511 amplifier under 3.3V supply voltage and Vdd=5V of LM48511. In this case, the DC bias voltage on VGO on LM48511 is Vdd/2 which is 2.5V. The signal peak voltage on VGO exceeds 3.3V such as 4V or 4.5V. Maybe the signal on VGO will be clamped by ESD Di on TPL0202. This is no good operation for TPL0202. Please look at attached scope picture about VGO waveform at Pout=2W with VDD=5V and Vbst=6.5V.

2) I also checked what Vbst voltage will be required to get 2W into 8 Ohms load within 7% of THD+N. Please take a look at image below, which is showing Pout Vs THD+N with Vbst and VDD voltage are parameters. To get 2W into 8 Ohms load, Vbst required is around 6V to 6.5V.

3) I’ll suggest that the customer set fixed gain like R5=R7=20k, R6=R8=40K and then input the signal from TPL2020 WA and WB via coupling cap, LA and LB are AC GNDed via cap and LHA and LHB are from Line out. In this case TPL2020 works under 3.3V.

Andy

Hi Andy,

Sorry for the delay. 

I would like to thank you and your team from the bottom of my heart for such a detailed study of my design.

Thank you so much for the inputs.

Would like to know how did you measure THD+N%? Any specific tool for that?

Thanks,

Shantharoopan TSB

Hi Shantharoopan,

I don't think so. We measure the THD+N% using some Audio Precision machines in our lab. It is easy to get the THD + N% using the software provided by the Audio Precision machine.

Andy  

Hi Andy,

I removed the DPOT and tested using with resistors of value Rf=42K and Rin=15K for power rating of 2W. When measured the THD+N% it was 54%.

Then used an external LC filter of cut-off frequency Fc=28.9KHz as below and the THD+N% improved suddenly to 4%.

Please let me know if this is the right procedure or correct me pls.

Thanks in advance,

Shantharoopan TSB

Hi Shantharoopan,

I suggest you take a look at this application note : http://www.ti.com/lit/an/sloa068/sloa068.pdf, which should answer your question.

Andy