Replacement for TDA2050v

[budm]

Re: Replacement for TDA2050v

I see, so you use the Power MOSFET as the low side switch to sink the load.

[JonathanAnon]

Re: Replacement for TDA2050v

How does VS- get to the circuit GND? I see it goes to the Center pin of the 3-pin connector. What is the 3-pin connector for?

Hi Bud, I want it so that when I push a button the amp stays on for only a short period of time, then auto shuts off. So I'm using the FET, with an RC timer at it's gate, to control the path to ground...

I got this idea from an Everyday Practical Electronics (magazine) timer which they used as an auto shut off for a cheap multimeter. But switching the ground might not work so well for an audio circuit.

[budm]

Re: Replacement for TDA2050v

How does VS- get to the circuit GND? I see it goes to the Center pin of the 3-pin connector. What is the 3-pin connector for?

[JonathanAnon]

Re: Replacement for TDA2050v

Next, connect the input ground to the VS- of the amp chip, directly. The way you have it here, you have a ground loop! Basically, higher currents - compared to the current of the input signal - will cause voltage drops across the GND wire, which basically will feed back into the input signal.

I see what you're saying about the ground loop. The chips only path to ground is through the FET, and there will be a slight voltage drop, so that the grounds of the chip and the grounds of the input/output will be at slightly different voltages. But because I'm switching on/off the ground side of the circuit I cant connect the battery ground to pin 3 of the amplifier chip as this will make the RC switch redundant. Maybe I should switch the positive side?

Also with regard to the interference between outputs/inputs/power, I was trying to fit the circuit on a very small space, but this is clearly to the detriment of the operation.

I personally would use a comparator with some hysteresis for the timer, but that's my way of doing things :P That may be another reason for bad sound quality, if the FET doesn't turn off cleanly, then it will have a high voltage drop across it, and the voltage at the amp will be less, which could lead to distortion at higher signal levels.

The circuit does get noisy as the RC circuit (triggering the FET) voltage goes lower. It's not something that annoys me, but was an optimization I was gonna work on.

I'm struggling a little with the software, but here's the circuit with a few corrections.

[budm]

Re: Replacement for TDA2050v

"The second after posting this I spotted a few mistakes, but I'll correct them later ..." Please do because it does not look right at all.

[ben7]

Re: Replacement for TDA2050v

Would you believe me if I said that I'm still running it without the heatsink and it's still working perfectly. I have an infrared temperature detector and it didnt get that hot. But the plan is to fit in a proper heatsink anyway.

Ahh, well, I guess if you aren't taxing it then it won't get that hot.
I was like for a second there! xD

This is the circuit that I am playing with at the moment. The left side of the circuit is for a RC circuit + momentary switch, so I can hit the button and the speaker stays on for a few minutes (while I'm in the shower) then switches itself off automatically. I'm building it on stripboard, cos I'm poor and dont have any proper PCB tools :-)

Also, cut traces so that none of the speaker or power traces are going near the input traces (as best as possible). That will help reduce any unintended noise/feedback.
Like (L,A), (L,B), (N,K), (N,M), (N,N), (N,O) ... etc...
Especially (L,B), this will decrease the coupling from the GND to the inverting input.
Point format: (Column/X, Row/Y) The strips on the board are the rows. Just saying, 'cause I don't want any confusion :P

Next, connect the input ground to the VS- of the amp chip, directly. The way you have it here, you have a ground loop! Basically, higher currents - compared to the current of the input signal - will cause voltage drops across the GND wire, which basically will feed back into the input signal.
This can cause distortion and even oscillation (worst case). That might be the source of your poor sound quality. I'd be surprised if the TDA2050 worked well with it like that!

Also, that 680uF capacitor on the feedback, shouldn't that be more like 10uF to 22uF?

You might want to add a capacitor to the 2x 22K resistor divider, to keep it more stable. The LM1875 datasheet shows a 10uF. Of course you don't have to follow the datasheet, but it isn't a bad idea to at least try.

I personally would use a comparator with some hysteresis for the timer, but that's my way of doing things :P
That may be another reason for bad sound quality, if the FET doesn't turn off cleanly, then it will have a high voltage drop across it, and the voltage at the amp will be less, which could lead to distortion at higher signal levels.
I'm surprised how it stays on for a few minutes using only the gate capacitance of the FET!

Also, the reference board diagram in the LM1875 datasheet is really messed up. The components are out of alignment with the board. They didn't do quite as good of a job of eliminating ground loops as I would like ... weird.

-Ben

[JonathanAnon]

Re: Replacement for TDA2050v

You ran it without a heatsink previously?!

Would you believe me if I said that I'm still running it without the heatsink and it's still working perfectly. I have an infrared temperature detector and it didnt get that hot. But the plan is to fit in a proper heatsink anyway.

Finally, you can try using a lower impedance speaker if you can, that will give you more power output with the same input voltage.

I've had trouble finding a chip to drive the 4 ohm speaker with the same power as the (now defunct) TDA2050V... The TDA2050V is perfect but no longer in production and doesnt seem to have an equivalent. Not using the 12V anyway, which is why I'm talking about DC-DC converters. I think you may be right about the 1875 I think it just needs a higher input supply to work properly.

This is the circuit that I am playing with at the moment. The left side of the circuit is for a RC circuit + momentary switch, so I can hit the button and the speaker stays on for a few minutes (while I'm in the shower) then switches itself off automatically. I'm building it on stripboard, cos I'm poor and dont have any proper PCB tools :-)



The second after posting this I spotted a few mistakes, but I'll correct them later ...

[ben7]

Re: Replacement for TDA2050v

@redwire
I've had it up and running for about a day now..... This circuit with the 1875T will simply not work without input filter caps, unlike the very similar circuit with the TDA2050V (which I had breadboarded and working fine without any filtration on the voltage source)

I've thought that you should always have some supply decoupling capacitors...

I also initially left out the feedback cap on the TDA2050V circuit initially, thinking that it was not mandatory, but the circuit would not work at all without it in place...

Yeah, leaving out the feedback cap definitely would cause issues.

I had some problems with it sounded distorted / muffled after a period of time, but I added a temporary heatsink to the IC, and a pot to limit the input and it's been running fine for the last five hours.

You ran it without a heatsink previously?!

I find that if I allow too much signal on the input, the output will distort badly... But I guess this is due to a higher gain in the 1875T circuit.... Gonna keep testing it with voice and music inputs for a few days...

The output is probably clipping. Basically, it's getting too close to the supply rails for the amp to handle, and the signal gets distorted, or "clipped" off, making the sound less than desirable.

Also, reading the datasheet, I see that it is suggested to keep power supply traces/wires away from the input wires, to reduce coupled currents.

I tried out the 1875 for a few weeks, and I just dont like it.... I built the two circuits and the TDA2050V is somehow more powerful, much less noisy and it just works. So I'll probably buy a few on the cheap from China.

I'll work with a view to using a DC to DC converter in future projects and a chip that works with a higher voltage.

DC to DC converter might make noise in the output, especially depending on how your grounds are wired, and the filtering on the output. But it is possible (think: car audio power amps)

I suggest trying a D class chip in the future. Less power loss in the chip itself, so you don't have to worry about dissipating that as much as with a regular amplifier.

Finally, you can try using a lower impedance speaker if you can, that will give you more power output with the same input voltage.

Oh well, one last thing is, you can do a bridge mode where you have two amplifiers driving one speaker, using a single ended power supply (no negative voltage rail). That will give you more output power, as the speaker will be able to get +-12V, instead of +-6V. (I think that is right, isn't it?)

-Ben

[JonathanAnon]

Re: Replacement for TDA2050v

I tried out the 1875 for a few weeks, and I just dont like it.... I built the two circuits and the TDA2050V is somehow more powerful, much less noisy and it just works. So I'll probably buy a few on the cheap from China.

I'll work with a view to using a DC to DC converter in future projects and a chip that works with a higher voltage.

[JonathanAnon]

Re: Replacement for TDA2050v

@redwire
I've had it up and running for about a day now..... This circuit with the 1875T will simply not work without input filter caps, unlike the very similar circuit with the TDA2050V (which I had breadboarded and working fine without any filtration on the voltage source)... I also initially left out the feedback cap on the TDA2050V circuit initially, thinking that it was not mandatory, but the circuit would not work at all without it in place...

@budm
input voltage is 12.2V and the voltage at the point between the resistor ladder (R1 and R2) is 5.98V, so it's pretty much 1/2Vcc

I had some problems with it sounded distorted / muffled after a period of time, but I added a temporary heatsink to the IC, and a pot to limit the input and it's been running fine for the last five hours. I find that if I allow too much signal on the input, the output will distort badly... But I guess this is due to a higher gain in the 1875T circuit.... Gonna keep testing it with voice and music inputs for a few days...

[redwire]

Re: Replacement for TDA2050v

I've never had a TDA circuit stable without adding a cap across the feedback resistor (R6) and an RF cap across the input (R4). The IC will oscillate or pick up AM radio, or both. It sounds like distortion but on an oscilloscope you can see HF oscillations. If you can't get things to make sense or sound decent:
In parallel with R6 add a 33pF cap (for 24kHz -3dB).
In parallel with R4 add a 220pF cap (for 70kHz -3dB).

[budm]

Re: Replacement for TDA2050v

OK, I thought you do the direct drop in with just the modification on the feedback, I did not know that you also added the biasing circuit.
So is the output pin sitting at 1/2VCC?

[JonathanAnon]

Re: Replacement for TDA2050v

You need to Bias the LM1875 pin1 so the output will be at 1/2 VCC since you are not using Dual power supply (+/-), you are using single power supply, it is also picky about how the circuit ground traces are routed. See app notes on how it is used, see Fig. 2.
https://cdn.badcaps-static.com/pdfs/...e9ec9f67d4.pdf

This is the single supply circuit that I'm working with.



From what I'm aware the two resistor R1 and R2 put the non-inverting input's biasing voltage at 1/2VCC. C1 and R3 form a high pass filter on the input. And I presume that the 1M ohm resistor is just to discharge any stray voltage when the chip is off. C4 and C7 are filtering caps on the input voltage. R7 and C5 is the Zobel network. R5, R6 and C3 set the gain and C6 blocks any DC from reaching the speaker. I'm open to correction but that's my reading of it.

I have the circuit built pretty much the same as it is above, but I'll take into account what you said about the chip being fussy, and I'll build it with the EXACT values (well, within 5%) on the datasheet...

[budm]

Re: Replacement for TDA2050v

The datasheet for that 1875T says that it's input voltage is 16V-60V.. I tried swopping it in to the same circuit in place of the
TDA2050V, and changed the resistors to 200k and 10k in the feedback part of the circuit... But it sounds very low and completely distorted. I dont think it's gonna work without a DC-DC converter to boost the voltage >16V..

I didnt choose the TDA2003 initially cos it was only 10W, and I was planning on using the same circuit for a small guitar amp in the same plastic box / 4 ohm speaker combination. However, it may be a good option. There's a single supply circuit in the datasheet, and the chip operates with Vs of 8v-18v.



Of course I still have the option of just buying the TDA2050v on ebay or something... The chip has been around for years, and they only get rid of the thing when I decide to use it !!!

You need to Bias the LM1875 pin1 so the output will be at 1/2 VCC since you are not using Dual power supply (+/-), you are using single power supply, it is also picky about how the circuit ground traces are routed.
See app notes on how it is used, see Fig. 2.
https://cdn.badcaps-static.com/pdfs/...e9ec9f67d4.pdf

[stj]

Re: Replacement for TDA2050v

places like Dalbani probably have it, but those obsolete parts places charge a lot!

[JonathanAnon]

Re: Replacement for TDA2050v

The datasheet for that 1875T says that it's input voltage is 16V-60V.. I tried swopping it in to the same circuit in place of the
TDA2050V, and changed the resistors to 200k and 10k in the feedback part of the circuit... But it sounds very low and completely distorted. I dont think it's gonna work without a DC-DC converter to boost the voltage >16V..

I didnt choose the TDA2003 initially cos it was only 10W, and I was planning on using the same circuit for a small guitar amp in the same plastic box / 4 ohm speaker combination. However, it may be a good option. There's a single supply circuit in the datasheet, and the chip operates with Vs of 8v-18v.



Of course I still have the option of just buying the TDA2050v on ebay or something... The chip has been around for years, and they only get rid of the thing when I decide to use it !!!

[stj]

Re: Replacement for TDA2050v

i just realised something.
the arcade stuff mostly used TDA2003 on a 12v supply.

it's almost a dropin for the TDA2050 with the same w/v curve.

[tom66]

Re: Replacement for TDA2050v

From a measurement of THD+N perspective, you may have a point, the THD+N figure for class D is worse, but I would contend the overall audio performance as humans perceive, and in the measurable 0-20kHz range, is better in most respects.

Class D really only has one disadvantage, and that's the noise components at higher frequencies due to switching noise. Most class D amplifiers switch around 300~400kHz to minimise filter sizes. This is well beyond any human hearing range so I fail to see how this could be considered an issue. Bear in mind most audio is only sampled up to a maximum of 192kHz. There are few subharmonics created by a stable clock source (such as a crystal) so even those should not be of a concern.

The process of generating audio is much more linear. The ideal class D amplifier would be purely DSP driven and would achieve perfect linearity. Most class D amplifiers use analog electronics (comparators and voltage ramps) to generate the switching pulses which introduce linearity errors although good design can minimise these. The only nonlinearities introduced by the switching process would be from the power supply sag and inductor saturation... both of which can be minimised by using proper design techniques and air-core inductors.

Contrast that to class AB and similar amplifiers. There is crossover distortion, which to eliminate requires large bias currents. Most manufacturers try to minimise this because it causes high power dissipation when operational. Also, the transistors themselves are quite nonlinear devices, something the feedback loop must compensate for.

[redwire]

Re: Replacement for TDA2050v

I've used TDA2030, 40, 50's in BTL (with a single-supply) but the turn-on thumps were huge and almost killed small speakers.
I would use one of the cheap DC-DC converters on eBay or DX to get higher voltage.
You need a heatsink on the LM1875T in your pic's, and negative voltage transient protection on power coming in.
Cloth acoustic-damping material in the box would help a lot.

[kaboom]

Re: Replacement for TDA2050v

If you are constrained by a single supply of 12-14V, why not go BTL?

Guaranteed output of 20-25W into 4 ohms, no DC-DC converter, and you'll save board space without the output coupling caps.

Try a TDA7370 (dual-btl) or TDA7240 (single-btl).