I'm thinking about buying this from Parts Express as my first experience with Class D amp:
2x50W TDA7492 Class-D Amplifier Board 320-301
It is cheap enough and I have a power supply to run it. I'm just curious as to what it will sound like. They say the Mfg is Parts Express but it may just be a OEM item built by Sure.
Being cheap and Class D I don't expect a lot out of it. But you have to start somewhere.
2x50W TDA7492 Class-D Amplifier Board 320-301
It is cheap enough and I have a power supply to run it. I'm just curious as to what it will sound like. They say the Mfg is Parts Express but it may just be a OEM item built by Sure.
Being cheap and Class D I don't expect a lot out of it. But you have to start somewhere.
I just wanted something simple to start with. I have a 24V 4 amp power supply that will run this board. PE sells a IR2092 based amp for $69 but it requires two 56V power supplies at 12 amps. The amp puts out 250 watts. I am only wanting something in the 20-25 watt range. So the IR2092 solution does not work for me. As I said, I just want something that works at this point. It doesn't have to be high performance.
This guy did...
http://www.diyaudio.com/forums/class-d/194413-sure-electronics-new-low-cost-25w-100w-board.html
I have zero experience with class D amps. I have a Chipamp based on the LM3875 on a small desktop system connected to a uDAC-2 in a tiny case and it sounds way better than I would have expected.
It seems you have to buy the better and more expensive high-end Class-D amps to have them sound good.
If you just want to experiment, the PE product is cheap. Building a chipamp requires soldering and testing.
Like you said you have to start some where. But you also need to have an idea of where you want to go.
I started my DIY audio hobby building a 100W Aleph 2! And it worked! SO I'm stuck here now trying to figure out what to build next.
http://www.diyaudio.com/forums/class-d/194413-sure-electronics-new-low-cost-25w-100w-board.html
I have zero experience with class D amps. I have a Chipamp based on the LM3875 on a small desktop system connected to a uDAC-2 in a tiny case and it sounds way better than I would have expected.
It seems you have to buy the better and more expensive high-end Class-D amps to have them sound good.
If you just want to experiment, the PE product is cheap. Building a chipamp requires soldering and testing.
Like you said you have to start some where. But you also need to have an idea of where you want to go.
I started my DIY audio hobby building a 100W Aleph 2! And it worked! SO I'm stuck here now trying to figure out what to build next.
Dane,
Thanks for the reply. That is really good documentation of the performance of the amplifier. As I said, I suspect the PE one is an OEM build from Sure. I'll order it and run similar tests that he ran. If I see the same thing I can replace parts to correct it.
I've built a couple of LM1875 amps that perform very well. I've also built a couple of LM3886 amps running about 35 watts. In the case of the LM3886 ones I have one of them running on a linear power supply and the other one on a home built switching supply. The switching supply one will put out over 45 watts of clean sine wave signal into an 8 ohm resistive load. Due to the heatsink being fairly small it does go into thermal shutdown after a couple of minutes.
I was more interested in thermal performance of the Class D amplifier than I was sound performance. The PE one has a small heatsink with a small fan on top. I was going to put a 1 KHz sinewave into it and drive it to the clipping point into 8 ohms. I'll let it sit there and see how hot it gets. According to the data sheet it will only put out about 30 watts with 1% distortion with a 24 volt supply. If it doesn't get very hot at that power level I will take the fan off the board and see if it goes into thermal shutdown.
If I'm satisfied with the results I will lay out a board based on app notes and build one with a slightly larger heatsink and no fan. I have a lot of SMD building experience so assembly will not be a problem. Since I will be building it from scratch I can use the correct components to prevent the problems that Trevmar discovered. All this is just for the purpose of learning more about Class D stuff. Most of my music listening is at about the 2-5 watt level so I have no need for 100s of watts.
Thanks for the reply. That is really good documentation of the performance of the amplifier. As I said, I suspect the PE one is an OEM build from Sure. I'll order it and run similar tests that he ran. If I see the same thing I can replace parts to correct it.
I've built a couple of LM1875 amps that perform very well. I've also built a couple of LM3886 amps running about 35 watts. In the case of the LM3886 ones I have one of them running on a linear power supply and the other one on a home built switching supply. The switching supply one will put out over 45 watts of clean sine wave signal into an 8 ohm resistive load. Due to the heatsink being fairly small it does go into thermal shutdown after a couple of minutes.
I was more interested in thermal performance of the Class D amplifier than I was sound performance. The PE one has a small heatsink with a small fan on top. I was going to put a 1 KHz sinewave into it and drive it to the clipping point into 8 ohms. I'll let it sit there and see how hot it gets. According to the data sheet it will only put out about 30 watts with 1% distortion with a 24 volt supply. If it doesn't get very hot at that power level I will take the fan off the board and see if it goes into thermal shutdown.
If I'm satisfied with the results I will lay out a board based on app notes and build one with a slightly larger heatsink and no fan. I have a lot of SMD building experience so assembly will not be a problem. Since I will be building it from scratch I can use the correct components to prevent the problems that Trevmar discovered. All this is just for the purpose of learning more about Class D stuff. Most of my music listening is at about the 2-5 watt level so I have no need for 100s of watts.
I purchased the 320-301 from PE just to get my feet wet. The results of my testing are pretty much what I expected based on the datasheet for the TDA7492. Even though PE says it is their amplifier it came in a Sure box.
I powered the unit with a SMPS that puts out 27.8 volts. With an 8 ohm 50 watt resistive load on one channel and driving it with a 1 KHz signal to just below where distortion starts to show up the amp puts out 14.19 volts. This is 25.2 watts which matches the datasheet
Next I put a 5.8 ohm load on the same channel and drove it to just below the distortion point and got 13.81 volts out. This is 32.9 watts which is somewhat below the 40 watts that the datasheet says is obtainable.
Finally I wanted to try one other test. I figured if the amplifier blew up it would be no big loss. So I tied the two outputs in parallel and drove both inputs with the 1 KHz signal. With a 4.5 ohm load I got 13.7 volts out before distortion which is 41.7 watts. With the 5.8 ohm load I got 14.34 volts. This calculates to 35.5 watts which is less than I expected. Changing the load to 8 ohms I was only able to get 14.50 volts out with the outputs in parallel. This equates to 26.3 watts. Certainly not what I had hoped for. Apparently paralleling the outputs doesn't give any more output except for the 4.5 ohm load. I was afraid to go any lower than 4.5 ohms as I was concerned about damaging the amplifier.
I powered the unit with a SMPS that puts out 27.8 volts. With an 8 ohm 50 watt resistive load on one channel and driving it with a 1 KHz signal to just below where distortion starts to show up the amp puts out 14.19 volts. This is 25.2 watts which matches the datasheet
Next I put a 5.8 ohm load on the same channel and drove it to just below the distortion point and got 13.81 volts out. This is 32.9 watts which is somewhat below the 40 watts that the datasheet says is obtainable.
Finally I wanted to try one other test. I figured if the amplifier blew up it would be no big loss. So I tied the two outputs in parallel and drove both inputs with the 1 KHz signal. With a 4.5 ohm load I got 13.7 volts out before distortion which is 41.7 watts. With the 5.8 ohm load I got 14.34 volts. This calculates to 35.5 watts which is less than I expected. Changing the load to 8 ohms I was only able to get 14.50 volts out with the outputs in parallel. This equates to 26.3 watts. Certainly not what I had hoped for. Apparently paralleling the outputs doesn't give any more output except for the 4.5 ohm load. I was afraid to go any lower than 4.5 ohms as I was concerned about damaging the amplifier.
Are you sure the SMPS does not limit the current?
Are you sure your voltage meter (or whatever you used for measuring) can measure correctly at 1 kHz?
Why do you think "25.2 watts ... matches the datasheet"? On figure11 32W is shown at Vcc=25V, and your Vcc is said to be 27.8V, which implyes even more power!
Why did you expect more power with paralleled outputs? A correctly working amplifier is basically a voltage source! Paralleling does not change the voltage of voltage sources!
Are you sure your voltage meter (or whatever you used for measuring) can measure correctly at 1 kHz?
Why do you think "25.2 watts ... matches the datasheet"? On figure11 32W is shown at Vcc=25V, and your Vcc is said to be 27.8V, which implyes even more power!
Why did you expect more power with paralleled outputs? A correctly working amplifier is basically a voltage source! Paralleling does not change the voltage of voltage sources!
Are you sure the SMPS does not limit the current?
That is a possibility. That is why I was going to parallel 2 of the power suppplies. The suppy is a GPM55-28 listed here:
http://www.slpower.com/data/collateral/GFPM55_II.pdf
It is short circuit current limited but I'm not sure what limits are applied for instantaneous peaks.
Are you sure your voltage meter (or whatever you used for measuring) can measure correctly at 1 kHz?
The meter specifications say it measures out to 10 KHz.
Why do you think "25.2 watts ... matches the datasheet"? On figure11 32W is shown at Vcc=25V, and your Vcc is said to be 27.8V, which implyes even more power!
That is at 1% distortion. I was measuring for no distortion. Fig 12 shows .1% distortion as being 20 watts. So with my 27.8V supply I would expect somewhere between 20 and 25 watts. If I take it up to 10% distortion it slightly exceeds the datasheet specs. Figure 20 states typical 38 watts at 1% into 4 ohms with a 20V supply but the curve indicates 32 watts. I'm getting 41.7 watts into 4.5 ohms before distortion. This is with the 27.8V supply. The datasheet tends to indicate a maximum supply of 22V with a 4 ohm load so 27.8V is probably not safe at 4.5 ohms.
Why did you expect more power with paralleled outputs? A correctly working amplifier is basically a voltage source! Paralleling does not change the voltage of voltage sources!
You are correct. I wasn't thinking. The only way to get more power into the load is to increase output voltage, not current. That means a amplifier with increased supply voltages.
That is a possibility. That is why I was going to parallel 2 of the power suppplies. The suppy is a GPM55-28 listed here:
http://www.slpower.com/data/collateral/GFPM55_II.pdf
It is short circuit current limited but I'm not sure what limits are applied for instantaneous peaks.
Are you sure your voltage meter (or whatever you used for measuring) can measure correctly at 1 kHz?
The meter specifications say it measures out to 10 KHz.
Why do you think "25.2 watts ... matches the datasheet"? On figure11 32W is shown at Vcc=25V, and your Vcc is said to be 27.8V, which implyes even more power!
That is at 1% distortion. I was measuring for no distortion. Fig 12 shows .1% distortion as being 20 watts. So with my 27.8V supply I would expect somewhere between 20 and 25 watts. If I take it up to 10% distortion it slightly exceeds the datasheet specs. Figure 20 states typical 38 watts at 1% into 4 ohms with a 20V supply but the curve indicates 32 watts. I'm getting 41.7 watts into 4.5 ohms before distortion. This is with the 27.8V supply. The datasheet tends to indicate a maximum supply of 22V with a 4 ohm load so 27.8V is probably not safe at 4.5 ohms.
Why did you expect more power with paralleled outputs? A correctly working amplifier is basically a voltage source! Paralleling does not change the voltage of voltage sources!
You are correct. I wasn't thinking. The only way to get more power into the load is to increase output voltage, not current. That means a amplifier with increased supply voltages.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.