If the single chip tpa3116 on TI EVM board is 60 watt in stereo mode it is 61 watt in mono mode into the same load that gave 60 watt for stereo. That powerdifference is not audible. It only becomes 120 watt when load is halved.
Specified in same manner the dualchip board is probably close to a 1 watt ampboard and the single chip might be higher power before the 1% THD+N level is reached.
Specified in same manner the dualchip board is probably close to a 1 watt ampboard and the single chip might be higher power before the 1% THD+N level is reached.
As I wrote in post #9500, I tested at bench the dual chip red board, finding a distortion of about 1.5% at 10 W output (8 ohm load and 24V power supply), also after synchronization of the two chips; increasing power supply capacitor value did not change distortion.
I believe that increasing PS capacitor and/or replacing it with a lower ESR/higher ripple current one, has the only effect to help to avoid clipping at the lowest frequencies, caused by the internal current limitation normally present in every SMPS; in my opinion, this is the reason why many people think that increasing PS capacitor can extend the low frequency response of an amplifier.
Another reason could be that an increase of the PS capacitor can lower the output impedance of the amplifier at the lowest frequencies (try to believe), giving a better damping to the woofer and improving its response (speaker alignment is normally performed considering zero or very low amplifier output impedance).
I believe that increasing PS capacitor and/or replacing it with a lower ESR/higher ripple current one, has the only effect to help to avoid clipping at the lowest frequencies, caused by the internal current limitation normally present in every SMPS; in my opinion, this is the reason why many people think that increasing PS capacitor can extend the low frequency response of an amplifier.
Another reason could be that an increase of the PS capacitor can lower the output impedance of the amplifier at the lowest frequencies (try to believe), giving a better damping to the woofer and improving its response (speaker alignment is normally performed considering zero or very low amplifier output impedance).
I had 2 of these dual chip 3116 laying in my desk because they hissed a bit at half volume. My new DAC has volume control so I was able to turn the potmeter on the amps all the way up and control them with the DAC. I also lowered gain on both boards to 20dB. This resulted in a darker background than the Sanwu 3118 that I find to have some (very faint) high pitch hissing. Might also be because these dual chip boards have 22uH inductors. I think they are designed for 8 ohms which my speakers is. Will return with more detail about sound. Only listened to background.
Too much distortion.I listen on low volume and can't hear that distortion.Thanks to fulvio.
Sent from my HT3 using Tapatalk
I'm testing this board:
TPA3116 Digital Power Audio Amplifier Board DC12-24V 100W Single Channel BTL | eBay
I've found it is switching at 125KHz and, as a consequence, the inductors are hot, also without signal.
Gain resistors on the board are R1 = 100k, R2 = 47k (resistors designator are as for TI data sheet), so the TPA3116, is set to work in SLAVE mode (!!!).
If somebody has this board, would do better to check them (thet are labelled as R2 and R3 and are located near the input connector.
TPA3116 Digital Power Audio Amplifier Board DC12-24V 100W Single Channel BTL | eBay
I've found it is switching at 125KHz and, as a consequence, the inductors are hot, also without signal.
Gain resistors on the board are R1 = 100k, R2 = 47k (resistors designator are as for TI data sheet), so the TPA3116, is set to work in SLAVE mode (!!!).
If somebody has this board, would do better to check them (thet are labelled as R2 and R3 and are located near the input connector.
Understandable mistake but they didn't test I guess, picture on ebay has same slave setup. Any decoupling under heatsink on these ?
edit: I made same mistake populating my first 3116 ampboard, but that didn't function back then, went into protection I remember. Maybe some early 3116 didn't function as slave only, letters behind 3116xxxxx did change.
edit: I made same mistake populating my first 3116 ampboard, but that didn't function back then, went into protection I remember. Maybe some early 3116 didn't function as slave only, letters behind 3116xxxxx did change.
Last edited:
Under the heathsink there are the four boostrap C and other 3 C decoupling PVCC, AVCC and GVDD.
I solved my problem with a 10K resistor in parallel to the existing 100K; now the TPA is in MASTER MODE with a gain of 26 dB (and it is switching at 400KHz).
I am wondering how many boards have been sold with the same mistake (and the buyers have not noticed it)
This question is asked every few months.
The ic is already a BTL (or PBTL) and will drive 3.2R speakers (or 1.6R speakers)
If you are paralleling speakers to get a lower impedance load (less than 1.6R) then why not just use one amp per speaker?
To answer your question, first recap the basic facts.
(1) A single TPA3116 channel (1 of 2 on the chip) is rated at 50w. That is a thermal power dissipation limitation.
(2) Using the both channels in PBTL mode, the power dissipation rating for the output is 100w.
(3) The power output of any channel is "nominally" defined as P=((VxV)/2)/R. At 24V, the maximum nominal output power is dependent on the speaker impedance.
Examples: 24V @ 8 ohms, P = 36W. 24V @ 4 ohms = 72w. 24V @ 2 ohms is 144W. The only way to get more power out of a channel at 24V is to have lower impedance speaker.
Therefore, to get more than a 100w, you need a very low impedance load. The TPA3116 in PTBL mode is spec.'d for ~2ohm load. But it cannot do 144w @ 2 ohms because of the thermal power limitation of 100W.
So, could you take two chips, each in PTBL and parallel the outputs to provide 200w? Yes and no. First you would need a 1 ohm load. Second, paralleling the output of two chips in PTBL mode is not supported by the chip. However, it has been done. You need to: (a) Provide the exact same input signal to both chips; and (b) have one chip in master mode and the other in slave mode.
This has not been offered on any board to my knowledge.
If you are working with 8 or 4 ohm speakers, to get higher output, you need a chip that operates at a higher voltage such as the TAS5630.
(1) A single TPA3116 channel (1 of 2 on the chip) is rated at 50w. That is a thermal power dissipation limitation.
(2) Using the both channels in PBTL mode, the power dissipation rating for the output is 100w.
(3) The power output of any channel is "nominally" defined as P=((VxV)/2)/R. At 24V, the maximum nominal output power is dependent on the speaker impedance.
Examples: 24V @ 8 ohms, P = 36W. 24V @ 4 ohms = 72w. 24V @ 2 ohms is 144W. The only way to get more power out of a channel at 24V is to have lower impedance speaker.
Therefore, to get more than a 100w, you need a very low impedance load. The TPA3116 in PTBL mode is spec.'d for ~2ohm load. But it cannot do 144w @ 2 ohms because of the thermal power limitation of 100W.
So, could you take two chips, each in PTBL and parallel the outputs to provide 200w? Yes and no. First you would need a 1 ohm load. Second, paralleling the output of two chips in PTBL mode is not supported by the chip. However, it has been done. You need to: (a) Provide the exact same input signal to both chips; and (b) have one chip in master mode and the other in slave mode.
This has not been offered on any board to my knowledge.
If you are working with 8 or 4 ohm speakers, to get higher output, you need a chip that operates at a higher voltage such as the TAS5630.
Oh I see, the problem is that I have a 15" subwoofer which can be 2 or 8ohm, rated around 200w, well boss said that it can handle 1250rms, thing that I doubt, because with 100rms it slaps pretty hard. I can use a mono TPA3116, but I think I will need a little more in order to have it at max level. I need something like TDA7498, but as I read, its not suitable for 2R loads, then if I wire the sub on 8R, it would be almost the same than having it connected through a TPA3116 on 2R, is not it? So i prefer an amplifier board that can work on the 2R load. Would anyone recommend me an option? something around 200/300w on 2R, class D and single power supply. Is the only option a tpa3255? which is not popular and expensive, thing that will change in the near future, is that the only option? Although I also can think in wiring the sub on 8R and use any other class D able to handle big rms on a 8R, which I need to be around 200w. Thanks.
Last edited:
TAS5630 is an option.
This is a pbtl board.
It comes with a preamp which sums the signal to mono and has variable low pass filter.
You can also omit the preamp.
https://de.aliexpress.com/item/For-..._1&btsid=2de3248d-3973-44f2-bb54-0267913f2c58
And of course you can use the smaller TAS5611.
This board is 2x 4Ohm.
But you can feed the same signal into both channels and connect your driver with 2x 4Ohm.
https://de.aliexpress.com/item/Frss..._1&btsid=1ef5fdd3-07d7-431a-b042-14ea13d9e73b
Last edited:
Thanks. I did not know you could connect stereo amps to a single double coil driver, knowing this bring me more possibilities when looking for boards (now impedance its down to 4Rx2). I understand that the injected signal should be the same, acting as a mirror in the driver. Thanks bro, now can reconsider options that I rejected before for not being mono, but also that mini tas5611 seems pretty good option, and uses single psu, which is great!I am also considering this board:
TDA7498E 2*160W 2 Channel Audio Stereo Digital Power Amplifier Board 15-36V H7B8
Last edited:
I am also considering this board:
TDA7498E 2*160W 2 Channel Audio Stereo Digital Power Amplifier Board 15-36V H7B8
Another option here:
http://www.diyaudio.com/forums/class-d/299877-problem-matching-2-amp-board.html#post4898876
Fulvio, I have two of the V3 dual chip XH-M190 boards, purchased from the same vendor but at different times. The first one has the same resistors as listed above, and there is a capacitor in the C33 position (I don't know how to determine its value). I've never worked with SMD components, but assuming I can figure out how to remove C33, is it reasonable to assume that I will have the same result with regard to synchronization that you did? I don't have test equipment beyond an ordinary DMM.
The second board has the following:
R3 753
R4 473
R9 753
R10 473
R13 Not present
C33 Present; value unknown
What is this configuration?
Thanks.
For the first board: I’m 99% confident that the modification will work; anyway, you could ask jwicaksana if he did it and how it worked; removal of the capacitor is not difficult (you can also cut it in the middle and then remove each part separately).
For the second board: both the IC’s are in MASTER mode (with 36dB gain) so they work independently, but this is not always good, because the two switching frequencies could produce audible beats noise, that is spurious signals at a frequency equal to the difference between the two switching frequencies themselves; synchronization of the two chips makes these two frequencies equal, but requires, in your case, to replace/add some SMD components.
So, my advice is to connect two speakers and listen: probably you are lucky and don’t hear any beats noise.