The emphasis was on the word "was" as in: "it was attached with an adhesive thermal pad but it is not anymore". Please read "with " and not "by" which is indeed wrong. So now a non native english speaker is correcting another non native english speaker ? 😀
https://www.youtube.com/watch?v=aOd1jIDCeN8
https://www.youtube.com/watch?v=aOd1jIDCeN8
Last edited:
Not exactly sure what you mean but with the english language following grammar advice from native speakers might not be the best idea 🙂
Anyone played with this blue version of the Sanwu TPA3116? It looks like an upgrade from the red board version. Apart from the net ground plane, it seems to follow the datasheet quite well. There seems to be a power plane rather than traces, and decoupled with smd caps close to the chip. Looks like Pin 1 is tied to gnd through R24 so it is BD mode and there is no power limit. I can't read the resistor values in the pic to know the gain. There are some smd resistors/caps under the toroids - not sure if they are bootstrap snubbers, CRC filter...? The AM pins are all tied to gnd so it's 400khz - is there much SQ to be gained from 1.2Mhz? https://world.taobao.com/item/44785...&spm=a312a.7700846.0.0.9V0gJ6&_u=j10d5b0gd588
Last edited:
Last October I visited London and while there I bought a package deal from a diyaudio member. 2 tube preamps/buffers, several small power transformers, some SMPS’s, a couple of electronic crossovers and 5 Class D amplifiers.
3 of the class D amplifiers are TPA3116 2.0
1 is a TPA3116 2.1
1 is a dual TRipath TPA2050 2.0 board
The tripath board worked for about 10 seconds before it quit. I’ll look at it later.
1 of the 2.0 TPA3116 boards did not work. A bit of solder on one of the pins of the TPA3116 fixed it.
All of the TPA 3116 boards worked well. I found that under heavy load or continuous sine wave testing they would all momentarily shut down. I replaced all of the tiny heat sinks with an old CPU heat sink that I cut up to fit the boards.
The TPA3116 2.1 board has a left and right volume control a sub frequency control and a sub volume control. There is no master volume. I cut a couple of traces and soldered some wires from the LR volume wiper pins to the sub Volume control.
I tested all the boards with a 5.5 amp bench power supply. I was surprised that the current limit on the bench supply would turn on when driving the 3116 amps to clipping and 4 ohm loads and power supply set to 24V
I am putting the 2.1 TPA3116 in a case for my daughter.
The amplifier will have a 2 position input switch. The transformer that I could fit in the case that I had lying around will give me about 22VDC.
I made a little anti turn on/off click and pop circuit. I will describe it in the next post.
3 of the class D amplifiers are TPA3116 2.0
1 is a TPA3116 2.1
1 is a dual TRipath TPA2050 2.0 board
The tripath board worked for about 10 seconds before it quit. I’ll look at it later.
1 of the 2.0 TPA3116 boards did not work. A bit of solder on one of the pins of the TPA3116 fixed it.
All of the TPA 3116 boards worked well. I found that under heavy load or continuous sine wave testing they would all momentarily shut down. I replaced all of the tiny heat sinks with an old CPU heat sink that I cut up to fit the boards.
The TPA3116 2.1 board has a left and right volume control a sub frequency control and a sub volume control. There is no master volume. I cut a couple of traces and soldered some wires from the LR volume wiper pins to the sub Volume control.
I tested all the boards with a 5.5 amp bench power supply. I was surprised that the current limit on the bench supply would turn on when driving the 3116 amps to clipping and 4 ohm loads and power supply set to 24V
I am putting the 2.1 TPA3116 in a case for my daughter.
The amplifier will have a 2 position input switch. The transformer that I could fit in the case that I had lying around will give me about 22VDC.
I made a little anti turn on/off click and pop circuit. I will describe it in the next post.
This circuit is based on anti turn on/off pop circuits that I used to make 45 years ago. I left out the DC protection because the DC protection circuit does not work with BTL amplifiers.
I am using an unregulated power supply for the TPA3116 2.1 because that’s what I have lying around. It gives me 22 volts for the amplifier.
The relay is an old Potter & Brumfield KUP14D15-24 a 24VDC relay 3poles and good for 10amps AC or DC. I will say now that except for the TPA3116 boards, all of the parts are from my junk box. I have a lot of junk.
The circuit works on the principle that the relay will turn on a few seconds after the power supply and amplifier stabilize and the relay turns off before the power supply drops to a low enough voltage to cause the amplifier to pop.
I installed 200ohm resistors on the output of the TPA3116 board to provide a load to the amplifier. The 200 ohm resistors are not shown in the schematic.
Q1, Q2 are darlington transistors.
On Power Up
Q2 turns on as the power supply is turned on.
Q1 turns on after a couple of seconds of C3 charging up through R1.
The Zener Z1 will pass current to Q1 after C3 reaches 14.5 volts.
When the voltage across C3 reaches about 16-17 volts enough current will flow through R3 and Q1 will turn on and since Q2 is already on the relay will energize.
At power off, as soon as the power supply drops below about 17-18 volts Q2 turns off and the relay de-energises and disconnects the speakers and no pop.
R2 discharges C3.
Again for this to work C1 must be large enough to keep the amplifier playing for a second or so.
C2 has to be large enough so that the relay does not chatter.
I tried 15 volt zeners but when driving the amplifier hard the power supply voltage would drop a bit and the relay would click on/off. Undersized power transformer.
All Relays will turn on well below the rated voltage and will stay energized at an even lower voltage. With this particular relay, turn on is about 17 volts and turn off is around 7.
If you are using a regulated power supply, a SMPS power supply or batteries you will have to have another diode in series with the power supply and a large filter cap after the diode. C2 will not be needed.
If you are using a power supply lower than 22 volts or different transistors, relay etc you may have to adjust R1 R3 R4 values and the zeners.
I normally use a LED and a resistor across the relay coil to indicate speakers on/off.
Q1,Q2 have internal diodes across the CE junction.
I drew the schematic after I built the circuit. It should be acurate.
I am using an unregulated power supply for the TPA3116 2.1 because that’s what I have lying around. It gives me 22 volts for the amplifier.
The relay is an old Potter & Brumfield KUP14D15-24 a 24VDC relay 3poles and good for 10amps AC or DC. I will say now that except for the TPA3116 boards, all of the parts are from my junk box. I have a lot of junk.
The circuit works on the principle that the relay will turn on a few seconds after the power supply and amplifier stabilize and the relay turns off before the power supply drops to a low enough voltage to cause the amplifier to pop.
I installed 200ohm resistors on the output of the TPA3116 board to provide a load to the amplifier. The 200 ohm resistors are not shown in the schematic.
Q1, Q2 are darlington transistors.
On Power Up
Q2 turns on as the power supply is turned on.
Q1 turns on after a couple of seconds of C3 charging up through R1.
The Zener Z1 will pass current to Q1 after C3 reaches 14.5 volts.
When the voltage across C3 reaches about 16-17 volts enough current will flow through R3 and Q1 will turn on and since Q2 is already on the relay will energize.
At power off, as soon as the power supply drops below about 17-18 volts Q2 turns off and the relay de-energises and disconnects the speakers and no pop.
R2 discharges C3.
Again for this to work C1 must be large enough to keep the amplifier playing for a second or so.
C2 has to be large enough so that the relay does not chatter.
I tried 15 volt zeners but when driving the amplifier hard the power supply voltage would drop a bit and the relay would click on/off. Undersized power transformer.
All Relays will turn on well below the rated voltage and will stay energized at an even lower voltage. With this particular relay, turn on is about 17 volts and turn off is around 7.
If you are using a regulated power supply, a SMPS power supply or batteries you will have to have another diode in series with the power supply and a large filter cap after the diode. C2 will not be needed.
If you are using a power supply lower than 22 volts or different transistors, relay etc you may have to adjust R1 R3 R4 values and the zeners.
I normally use a LED and a resistor across the relay coil to indicate speakers on/off.
Q1,Q2 have internal diodes across the CE junction.
I drew the schematic after I built the circuit. It should be acurate.
Attachments
I have glued this to Sanwu pbtl pcb bottom, exact fit, that is why, no idea of effectiveness, amp is used by someone else, personally don't see direct need to add heatsink, but it doesn't hurt probably.
Aavid Thermalloy 374624B00032G
Aavid Thermalloy 374624B00032G
If you look at IR class D layout design advise or Hypex and compare what is writen by TI you will notice Sanwu outputfilter layout meets TI and differs from IR/Hypex in that outputfilter gnd according to TI should be close to powersupply (mostly interpreted as board entry) and accordeing to IR/Hypex should be close to local powersupply. Sanwu pbtl is small, but my first attempt design with huge filmcaps going back to chip/local powersupply is closer to IR/Hypex and closer to TI tpa3251EVM than most other TI interpretations of where powersupply is. I tend to think the high frequency current will "travel" the entire ampboard easier than "travelling" to external powersupply 🙂
The Hf return currents directly travel under their corresponding trace. So best practice is to have the return path as short as their conductor to keep the radiation loop small. Termination should be on the point of source.
So Sanwu pbtl increases path lenght here by splits in groundplane and vias, do I understand/see that correctly ? It still is a very small ampboard, so even with these splits total path lenght might be far smaller than other ampboards, still...
I talk about the highest frequencies, high currents that TI in appnote 32xx inputstage also identifies for increased distortion and bass deterioration, they travel through groundplane and through outputcapacitor to ground, not around traces ?, you might refer to even higher frequencies?
Higher MHz you can't filter LCR, probably just hope they won't be amplified too much by the CLR filter then ? Lower MHz should be through capacitor still acting as a capacitor with inductor still acting like an inductor. I think LOL
That's because at 24v into 4 ohms, you're getting 70+wrms per channel, x2 is 140Wrms. 140W divided by 24v is 5.833 amps, more than the PS can deliver. So driving into clipping (where even more power is used) the PS should shut down.
yes I know the math. The power supply went into current limit before clipping occurred. I had to use only one channel to get the max output voltage before clipping. The output of the chip is nowhere near 24 volts, more like 16-16.3. Higher than about 16 there is obvious distortion on the peaks.
Are you measuring rms voltage at the output. ~16+ Vrms would be about the max output for a 24v DC supply after accounting for semiconductor losses.