Hi,

I have recently purchased an amplifier module based on the TDA8950 device from eBay and am experiencing distortion at all output levels.
(eBay: http://www.ebay.co.uk/itm/252534422644)

The module includes a bridge rectifier and smoothing caps and only requires a (stated) 22V-0-22V AC supply.

The circuit board does not appear to have any form of voltage regulation and according to the NXP spec sheet the device can handle a wide range of supply voltages from ±12.5 (min) ±35 (typ) ±40 (max).
(NXP spec sheet here: http://www.nxp.com/documents/data_sheet/TDA8950.pdf

I have a been powering it with a toroidal transformer supplying ±30V AC.
When I first set it up on the bench I connected it to a small set of 4.8Ω speakers and fed it with a line level signal from a CD player. At all volume levels there is distortion.

If I set up the amplifier with a signal generator producing a 1KHz test tone at 0.4V and drive 4.7Ω power resistors (or the speakers) and look at the input and output signals on an oscilloscope the output trace apears ‘fat’ when compared to the input. The output is not clipping - although that can be achieved.

The user connections on the board are:
AC Power In: 22V-0-22V AC
Signal Input: (via what looks like a JST type connector) Left-Ground-Right
Outputs: Out-L+GND & Out-R+GND

There was no supporting paperwork supplied with the module. (I have briefly described the problem to the Chinese supplier and their brief response was “friend, could you check your connect and try it again”)

Any ideas and thoughts would be most welcome.

Thanks

Mods: Please forgive me if this thread is in the wrong forum - please move if required.

 

Only thing i can suggest is the transformer is not giving enough current out, can you measure the voltage across the two smoothing capacitors, when you apply the signal, otherwise try using a dual DC power supply.

 

Can you determine if the distortion is at the PWM frequency of the Class D? How have you set up the oscillator frequency?

 

If you select x10 magnification on the scope you can see the 'fat' output trace is indeed made of a high frequency sine wave. I can't be 100% sure but I (think) I have measured a peak to peak time of 3.04 uSec . There is no user input to externally clock the device on the board and my assumption was it was self clocking. In the eBay description there is no mention of users having to setup an oscillator freq.

Dave: The voltage across each 3300uF/50V cap is 40.5V

 

What happenes if you sweep the frequency from 150Hz upto 20Khz??


Think your psu is Too high voltage!!!!

 

Hi Dave. The 2 traces below display 120Hz (top) and 21KHz (bottom).
Iv'e also swapped the mains transformer for one that outputs 21-0-21V to see if that made a difference - it did not.

 

I've also made a short audio recording (CD music playback) which can found here:
https://dl.dropboxusercontent.com/u/14584363/TDA8950 Distortion-01.mp3

 

Hi,

I have recently purchased an amplifier module based on the TDA8950 device from eBay and am experiencing distortion at all output levels.
(eBay: http://www.ebay.co.uk/itm/252534422644)

The module includes a bridge rectifier and smoothing caps and only requires a (stated) 22V-0-22V AC supply.

The circuit board does not appear to have any form of voltage regulation and according to the NXP spec sheet the device can handle a wide range of supply voltages from ±12.5 (min) ±35 (typ) ±40 (max).
(NXP spec sheet here: http://www.nxp.com/documents/data_sheet/TDA8950.pdf

I have a been powering it with a toroidal transformer supplying ±30V AC.

right there. 30 vac will give you about 45v dc

 

The "fat" waveform is the 345kHz PWM signal. This is normal. Your speaker cannot respond to such a high frequency.
My guess is the apparent distortion is from something else.

 

There is no user input to externally clock the device on the board and my assumption was it was self clocking. In the eBay description there is no mention of users having to setup an oscillator freq.

You should read the data sheet and ignore everything you read on eBay. See section 8 regarding the oscillator.

 

Modular amplifiers can be a tough challenge to diagnose. I do not have much experience on class d amplifiers but have built plenty of class ab amps and they are very sensitive to becoming an oscillator instead of an amplifier. From looking at the scope it looks like its oscillating from feedback. All audio amplifiers must have a high frequency feedback blocking filter usually a capacitor about 470 pico farrad from output to inverting input to prevent excessive high freguency gain. Also check the power ground ang get it as far away from the input as possible. Check all or your small capacitors. I hope this helps.

 

Is your circuit similar to the one in figure 10 of the datasheet?

 

Hello Joe jester, so far I have not tried to put a class d amp kit together. I was referring to a class ab simple push pull but I don't know if a class d has feedback protection. You could have simply a defective ic.

 

The TS said they were using the kit found on ebay based on the TDA8950. I'm asking is the kit wired similarly to the schematic on figure 10 of the data sheet. I know it will be difficult since the heat sink is firmly mounted on the TDA8950.

The component values will give the biggest clue.

 

The TS said they were using the kit found on ebay based on the TDA8950. I'm asking is the kit wired similarly to the schematic on figure 10 of the data sheet. I know it will be difficult since the heat sink is firmly mounted on the TDA8950.

The component values will give the biggest clue.

Yes

 

Hi All,

Many thanks for all suggestions of help with this but I'm personally at the end of the road with this amplifier module.

If anyone in the UK would be interested in doing some hands on diagnostics please email me.

Thanks.

nevillet01@gmail.com