Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

tda2030 amplifies noise

Status
Not open for further replies.
Most of the problems were due to under-rated power supplies. You have to remember that in the 1980's consumer electronics boom, the markets were flooded with very low quality equipment that boasted specifications that were impossible to reach. There were no consumer protection laws in those days so a "Hi Fi" unit might have superb specifications but in reality be an under rated, badly designed, badly built cardboard PCB with 50mm bass speakers! If I remember correctly, one "20W per channel" system that passed through my repair workshop had a 1A rated mains transformer in it that also powered the receiver, cassette deck and record deck motors.

With 'noise' you have to be careful how you measure it. For example, 50mV of noise on a supply line might mean 50mV of noise across the loudspeaker but the same noise at the amplifier input would drive it into full power. Generally, the hiss kind of noise that increases with the volume control is generated in the pre-amplifier stages where the gain is normally much higher. It rarely originates in the power output stage. I have never seen any 'noise reduction' circuits in a power amplifier and usually by careful choice of components and gain distribution, it can be minimized in the earlier stages.

The LM1875 is almost a drop in replacement for the TD2030A but has much higher specifications. It is still in production and easily available. It's distortion specification is typically 50 times lower than the TDA.

Brian.
 
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
Most of the problems were due to under-rated power supplies. You have to remember that in the 1980's consumer electronics boom, the markets were flooded with very low quality equipment that boasted specifications that were impossible to reach. There were no consumer protection laws in those days so a "Hi Fi" unit might have superb specifications but in reality be an under rated, badly designed, badly built cardboard PCB with 50mm bass speakers! If I remember correctly, one "20W per channel" system that passed through my repair workshop had a 1A rated mains transformer in it that also powered the receiver, cassette deck and record deck motors.

With 'noise' you have to be careful how you measure it. For example, 50mV of noise on a supply line might mean 50mV of noise across the loudspeaker but the same noise at the amplifier input would drive it into full power. Generally, the hiss kind of noise that increases with the volume control is generated in the pre-amplifier stages where the gain is normally much higher. It rarely originates in the power output stage. I have never seen any 'noise reduction' circuits in a power amplifier and usually by careful choice of components and gain distribution, it can be minimized in the earlier stages.

The LM1875 is almost a drop in replacement for the TD2030A but has much higher specifications. It is still in production and easily available. It's distortion specification is typically 50 times lower than the TDA.

Brian.

Were thermionic valve audio amps available far prior to mentioned electronics boom duration amplifiers any better in quality? Seems many very old tube amps are still entirely functioning considering they only become slightly warmer when specs are over driven dislike semiconductor devices in which case they would incur irreversible damage.

The 1amp output transformer supply product must have been an imported product made offshore at minimum expenditure or was it a product from a mainland cooperation?
 

Many cheap amplifiers and speakers have the volume turned WAY UP so that the amplifier is clipping like crazy producing lots of awful distortion to make the power output number higher. Then they state the Peak or "Maximum" power instead of the continuous RMS power. There is a new power rating system that uses a squarewave input so that an amplifier that produces 10W at low distortion is rated at 35W. Often, the rated power number is when the output is only momentary before the amplifier or speaker melts.

The TDA2030A and LM1875 do not rate the continuous RMS output power at low distortion. Instead, the datasheet for the LM1875 lists the "typical" distortion at 0.05% but guarantee it is 0.4% or less. The TDA2030A lists the "typical" output power and distortion but do not say the maximum amount of distortion that the one you buy produces.

I built a vacuum tubes amplifier 59 years ago. It sounded good when the vacuum tubes were new but the power output dropped and the distortion rose after a few months. Macintosh amplifiers were very expensive but they sold many because they opened "clinics' where they tested the power output and distortion of your amplifier. Mine measured 25% distortion at half its rated power but measured very low distortion at its rated power with new vacuum tubes. The distortion and low power returned every 3 months.
Maybe that is why many Bands replace their vacuum tubes before each show.
 
Last edited:
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
I agree with Audioguru, tube amps were good but like all technologies, they get superseded with something better. Anything that is 'latest technology' is best, but only until technology moves on.

The product with the 1A transformer was a UK manufactured product from a company that made lots of consumer items at that time. It worked of course, but could never have met the specifications claimed for it. There was another popular product on the market at the time, a class-D power amplifier that used Germanium output transistors! The spec said it ran from 12V at 1A and gave 20W RMS output. Figure that one out....

Brian.
 
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
Were tubes detrimenting quality caused from vacuum pressure loss, electrode deterioration or was it from another issue? The tda amp board was recently submitted for fabrication. Is it fine to employ a few axial ferrite beads since 11uH only impedes 15Khz AC by 1R - one on high and one on low side of audio input? The beads are far below 1uH but they detriment high frequency noise which - even if no beads were in place and high frequency gain is prevented from being amplified by implemented pot and suggested 100pF capacitor the noise can cause ringing elsewhere in the board. It would have been far better to employ metal foil resistors since TDA IC is very susceptible to amplifying noise however they are very costly and there might be adequately sized resistors which fit this board which can be obtained later if this board doesn't have noise issues.

https://i.ibb.co/S3ZYmsB/tda2030.gif

https://i.ibb.co/ZBq2Gkh/tda2030-1.jpg

https://i.ibb.co/XpRcL2J/tda2030-2.jpg
 
Last edited:

Ferrite beads will have almost no effect at low frequencies, they don't really do anything useful until 100s of KHz and higher.

I'm not sure where this fantasy about resistor noise comes from. The IC will produce 100 times more noise than the resistors and even then the amount will be very tiny. This is a power amplifier, it has little voltage gain and a lot of current gain so it can drive a low impedance load. Without high voltage gain, the internal noise will not be significantly amplified. The feedback potentiometer only sets the knee frequency in HF roll-off, it isn't for nulling anything and from an audible perspective the most you would hear is a slight drop in highest notes as the value reduces to zero.

Brian.
 
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
Seems product class d with germanium transistors was liquidating on hand old stock with the advent of silicon parts they become obsolete. Many power germaniums had 5Khz f(t) far too slow for class D switching. But mentioned products' power supply does not output no were near enough to cause thermal issues for smaller germanium transistors and smaller parts tended to have larger f(t) values. Would low pass filter <=20Khz on preamp side help with noise reduction?
 

If the preamp creates the hiss then of course if you add a lowpass filter to its output then you cut the hiss and all other audio high frequencies.

Here we go again. I downloaded the datasheet of the TDA2030 that is not made anymore. They guarantee that it has fairly low noise and its typical noise is very low. Since it is not made anymore are you using a fake one from China?
 
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
Its not the iCs, the're authentic. Whats the typical RMS volts into the non inverting pin for this single supply amp?

Is this a decent filter or can values be improved?

https://i.ibb.co/QjF1FW5/filter.png
 

Attachments

  • filter.png
    filter.png
    7.4 KB · Views: 101
Last edited:

Seems product class d with germanium transistors was liquidating on hand old stock with the advent of silicon parts they become obsolete.
My point was that 12V at 1A means a power consumption of 12W (Watts = Volts * Amps) but it claimed 20W output power - so where did the extra 10W come from...

The audio frequencies you are interested in range from about 30Hz up to about 20KHz, a ferrite bead will have no audible effect at those frequencies. In fact they would have virtually no effect at 10 times audio frequencies.

Noise is something we have to tolerate but we design to make it as small as possible. There is no magic circuit that removes noise after it has been created. Human hearing is more sensitive to noise as the frequency increases (within audible range) but deliberately attenuating higher frequencies to make it less noticeable also reduces the wanted high notes as well. If you have a music system with bass and treble controls, turn the treble down and you will notice background noise also drops but so does the sharpness of the music. The treble control is basically a high frequency attenuation device, working like the capacitors you added to the TDA2030 circuit.

Brian.
 
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
It would be instructive to try with a battery ( and vol pot turned to zero) to see if the noise was quieter compared to the Meanwell power supply - it may be that the chips you have are an early copy batch and are just plain noisy and sold for cheap some where ...
 
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
Hi,

The schematic of post#29 has some issues.
* 200 Ohms may be too low for a standard signal input
* the pot misses the GND connection
* the capacitor in combination with the pot forms an LPF. But the LPF cut off frequency depends on source impedance and pot position. Usually one wants a stable cutoff frequency.

Klaus
 
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
Yes 200R pot should be 100k-ohms 470nF cap should be 4.7nF else you will attenuate a lot of the music frequencies ...
 
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
The non-inverting input level of your TDA2030 amplifier depends on its output level divided by your gain of 1 + (100k/4.7k)= 22.3 times. but since you are using the minimum supply of 12V the datasheet does not show the puny output level.
Their graph of output power vs supply voltage starts with a 16V supply producing an output of 2.2W into 8 ohms.

Then 2.2W into 8 ohms is 4.2V RMS which is 11.9V p-p. Then the total output voltage loss is 16V - 11.9V= 4.1V.
Your 12VDC - 4.1V= 7.9V p-p which is 2.8V RMS output just before clipping. So the input voltage is 2.8V/22.3= 126mV RMS. your output power into 8 ohms just before clipping is only (2.8V squared)/8= 0.98W.

The datasheet shows an input noise level of only 10uV and with your gain of 22.3 times then the output noise level is 223uV then the signal almost at clipping level to noise level is 12,600 which is about 82dB. But that is at full blast. Your average output level will be about 0.8V RMS then the signal to noise level is only 3400 times which is about - 72dB which might be audible.

So your noise problem is caused by using a low output power supply voltage which causes the low output power level of only 1W to be close to the noise level.
 
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
Hi,

The schematic of post#29 has some issues.
* 200 Ohms may be too low for a standard signal input
* the pot misses the GND connection
* the capacitor in combination with the pot forms an LPF. But the LPF cut off frequency depends on source impedance and pot position. Usually one wants a stable cutoff frequency.

Klaus

Why would a pot of this arrangement need its 3rd pin grounded which would draw current thru the entirety of the 200R resistor into ground making no use of it? The 200R pot was selected to demonstrate the low pass filter will infact remove the audio as the pot is lowered it will show its effect on noise. At 3R the filter passes ~112,875Hz and 200R it impedes >1,670Hz which is audible.

The non-inverting input level of your TDA2030 amplifier depends on its output level divided by your gain of 1 + (100k/4.7k)= 22.3 times. but since you are using the minimum supply of 12V the datasheet does not show the puny output level.
Their graph of output power vs supply voltage starts with a 16V supply producing an output of 2.2W into 8 ohms.

Then 2.2W into 8 ohms is 4.2V RMS which is 11.9V p-p. Then the total output voltage loss is 16V - 11.9V= 4.1V.
Your 12VDC - 4.1V= 7.9V p-p which is 2.8V RMS output just before clipping. So the input voltage is 2.8V/22.3= 126mV RMS. your output power into 8 ohms just before clipping is only (2.8V squared)/8= 0.98W.

The datasheet shows an input noise level of only 10uV and with your gain of 22.3 times then the output noise level is 223uV then the signal almost at clipping level to noise level is 12,600 which is about 82dB. But that is at full blast. Your average output level will be about 0.8V RMS then the signal to noise level is only 3400 times which is about - 72dB which might be audible.

So your noise problem is caused by using a low output power supply voltage which causes the low output power level of only 1W to be close to the noise level.

Going to check if amp is better with =>16Vcc with the bench supply, in which case a higher voltage power adapter or onboard boost should be obtained. Might try also with several series 18650s to make =>16v batteries would not eminate ripple into the amp unlike typical supplies. Perhaps the board will have an onboard DC converter from 12v adapter to 18v-22v or is this a guarantee for massive noise?
 
Last edited:

Undoubtedly the lowest power supply noise would come from a good linear regulated supply. As soon as you 'boost' the voltage electronically you risk introducing interference. Think about it, where does the extra voltage come from? It is done using switching techniques that use bursts of high current to charge a component, usually an inductor then releasing the charge as extra voltage. Those current bursts cause fluctuations in voltages due to drops in the wiring resistance so unless you are extremely careful, they reach the amplifier stages.

I'm still not convinced of the source of the noise you hear though. I appreciate what Audioguru explained about the noise being a bigger proportion of the output when the desired sounds are limited to lower volume but even so, it should be hardly audible.

Returning to your earlier question:
Were tubes detrimenting quality caused from vacuum pressure loss, electrode deterioration or was it from another issue?
It was mostly due to the electron emission from the cathode electrode falling as the tube got older. The reason for that was partly because of long term chemical changes in the cathode coating and partly because of slow release of oxygen molecules from the construction materials causing oxidization. Bear in mind the cathode would run continuously at several hundred degrees temperature in normal operation. Used within ratings they usually lasted for many years.

Brian.
 
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
Has the OP tried with batteries to see if the noise goes away?
 
  • Like
Reactions: Zak28

    Zak28

    Points: 2
    Helpful Answer Positive Rating
Boards enroute will be assembled in short order. Very likely =>16Vcc will alleviate noise from output.

My point was that 12V at 1A means a power consumption of 12W (Watts = Volts * Amps) but it claimed 20W output power - so where did the extra 10W come from...

The audio frequencies you are interested in range from about 30Hz up to about 20KHz, a ferrite bead will have no audible effect at those frequencies. In fact they would have virtually no effect at 10 times audio frequencies.

Noise is something we have to tolerate but we design to make it as small as possible. There is no magic circuit that removes noise after it has been created. Human hearing is more sensitive to noise as the frequency increases (within audible range) but deliberately attenuating higher frequencies to make it less noticeable also reduces the wanted high notes as well. If you have a music system with bass and treble controls, turn the treble down and you will notice background noise also drops but so does the sharpness of the music. The treble control is basically a high frequency attenuation device, working like the capacitors you added to the TDA2030 circuit.

Brian.

Apparently the carefully implemented and disguised free energy generators of the 80s in said products were not unveiled by you nor your technicians in your repair shop.

View attachment inadaquate-power-supply.zip

Also dont disregard corporations typically don't make fraudulent claims. They commonly make claims and counter claims called disclaimers to prevent issues from consumers. Those disclaimers tend to be not featured alongside claimed objects of products.
 
Last edited:

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top