This DC/DC converter based on a LM46002 powers the AVCC, OP Amps and the 3V3 DC/DC including all 3V3 components (also the display for example)
The converter and its inductor is indeed oversized but I have really good experiences with this converter and inductor combination so I decided to use this combination here, too.
It's a TPS563200.
Yes, for high frequency applications or fullrange in combination with lower power supply voltage / "smaller" TPA chips.
But of course you can use the big inductors (MA517x) for both channels, too. The MLC1565 are just a option for lower power applications.
Regards
Bernhard
Bravo, what an excellent project!
What's the target power output for the boards pictured?
What kind of heatsinking is required to drive a 50W max 8ohm tweeter and 150W max 8ohm woofer? For example, I noticed in-development projects like the Sinewave G6H have massive heatsinks with a fan, and then doctormord designs use small heatsinks. I just assumed each design is targeting a different power requirement.
What's the target power output for the boards pictured?
What kind of heatsinking is required to drive a 50W max 8ohm tweeter and 150W max 8ohm woofer? For example, I noticed in-development projects like the Sinewave G6H have massive heatsinks with a fan, and then doctormord designs use small heatsinks. I just assumed each design is targeting a different power requirement.
Isn't the TPA325x a bit of an overkill for a "not the high grade sound quality" solution?
//
With 90% efficiency, this will be 20W power dissipation. At 30C ambient temperature and 125C max die temp there will be:
125C - 30C - 20W*0.3K/W (Rjctop) = 89K
89K / 20W = 4.45K/W
4.45K/W for the heatsink and thermal bridge (paste or pad).
This calculation is for continuous output power. At normal music levels you can recalculate this for like ~4W dissipation due to crest factor.
Good day 
I continued with some tests on my amplifier and can give you some news:
First test was a kinda stresstest under the following conditions:
The "heatsink" (just a piece of aluminum 160x100x3mm) had 77°C when the amplifier went to thermal shutdown. This was after 15 minutes!
The test setup:
Amp with temp sensor, channel 2 driven.
DC PowerSupply with 48V / 4.8Amps
THD+N measurment
Temp measurment
4 ohms resistive dummy load with active cooling
Note: This test is far away from what you need in real life. Nobody will listen to 200Watts 1kHz sinewave. The most speakers and ears will die in this situation! It just shows what's possible with just a simple and cheap heatsink.
For a normal "home use" it's totally enough to just mount the amplifier inside a aluminum housing. If you want to use this amp under high loads you will need a dedicated heatsink and a ventilated housing.
The next test was the THD+N measurments of the small MLC1565 inductors.
I need to say that these inductors are not recommended for signals < 2kHz. THD increases drastically in this case! For signals between 2kHz and 22kHz THD+N was quite low at around 0.008% but between 20Hz and 2KHz it went up to 0.22%.
So if you want to save some money and just use this channel for high frequency speakers the MLC1565 inductors are a good choice, but I will not recomment them for a fullrange speaker.
So yes, stay tuned ;-)
Regards
Bernhard
I continued with some tests on my amplifier and can give you some news:
First test was a kinda stresstest under the following conditions:
- 1kHz sine wave input
- ~210Watts into a 4 ohms load (one channel driven)
- ~1% THD+N
- Amp was supplied with 48V / 4.8Amps
- Temp Sensor was mounted on the corner of the aluminum
- T(ambient): 22°C
- TPA3255 chip
The "heatsink" (just a piece of aluminum 160x100x3mm) had 77°C when the amplifier went to thermal shutdown. This was after 15 minutes!
The test setup:
Amp with temp sensor, channel 2 driven.
An externally hosted image should be here but it was not working when we last tested it.
DC PowerSupply with 48V / 4.8Amps
An externally hosted image should be here but it was not working when we last tested it.
THD+N measurment
An externally hosted image should be here but it was not working when we last tested it.
Temp measurment
An externally hosted image should be here but it was not working when we last tested it.
4 ohms resistive dummy load with active cooling
An externally hosted image should be here but it was not working when we last tested it.
Note: This test is far away from what you need in real life. Nobody will listen to 200Watts 1kHz sinewave. The most speakers and ears will die in this situation! It just shows what's possible with just a simple and cheap heatsink.
For a normal "home use" it's totally enough to just mount the amplifier inside a aluminum housing. If you want to use this amp under high loads you will need a dedicated heatsink and a ventilated housing.
The next test was the THD+N measurments of the small MLC1565 inductors.
I need to say that these inductors are not recommended for signals < 2kHz. THD increases drastically in this case! For signals between 2kHz and 22kHz THD+N was quite low at around 0.008% but between 20Hz and 2KHz it went up to 0.22%.
So if you want to save some money and just use this channel for high frequency speakers the MLC1565 inductors are a good choice, but I will not recomment them for a fullrange speaker.
So yes, stay tuned ;-)
Regards
Bernhard
Last edited:
Ah I was talking about that this amp is not equipped with a high quality ADC / DAC converter, "high quality input caps" and super voodoo op amps.
This amp is equipped with good quality solid parts and audio grade input caps. Enough for most application with the eye on a balance between audio performance, efficiency and power but infact not a high class audiophil amplifier.
Sorry, my mistake. It was ~1% THD+N. I corrected it.
Hmm good point, I will make this test again with the sensor directly at the plate connection of TPA.
77°C trip would indicate a non-optimal heat transfer, you should get higher temps at the heatsinks hotspot before OTP trips.
(rougly 21W*0.3K/W Rjctop -> 6.3K)
OTE trips at ~155°C, so there is 155°C - 6.3K - 22K = 126.7K which would lead in 126.7K/21W= 6K/W for the heatsink + thermal bridge. Seems legit if the amp way layed down on the "heatsink" without spacers. Having the pcb+heatsink in 90° should give some more minutes before OTE. I'd expect ~120-130°C at the hotspot (bottom side heatsink direct unter the chip).
125
(rougly 21W*0.3K/W Rjctop -> 6.3K)
OTE trips at ~155°C, so there is 155°C - 6.3K - 22K = 126.7K which would lead in 126.7K/21W= 6K/W for the heatsink + thermal bridge. Seems legit if the amp way layed down on the "heatsink" without spacers. Having the pcb+heatsink in 90° should give some more minutes before OTE. I'd expect ~120-130°C at the hotspot (bottom side heatsink direct unter the chip).
125
Yes you're right, thanks for your support on my project.
I will keep my eye on this.
My new website is online: EasyAmp :: very easy to use Class D Amp Projects
Just a couple of information online for now but I will keep updating this site.
I also orderd the parts for two more amplifiers and designed new spacers for TPA <-> heatsink.
Just a couple of information online for now but I will keep updating this site.
I also orderd the parts for two more amplifiers and designed new spacers for TPA <-> heatsink.
The new spacers will have a lot better connection to the heatsink, this should keep the thermal resistance low.
Next point is that the new spacers can be mounted on the heatsink directly, what makes the mounting of amp easier. Actually the threads are in the "heatsink" (this piece of aluminum you saw in my posts) and the spacer just gets pressed to the heatsink with the screws what goes through the spacer.
The new ones will be screwed to the heatsink directly and will have threads for screws
so the mounting of amp can also be done by the customers.
Unfortunately I don't have a milling machine so I need to wait for a friend to make the spacers.
Next point is that the new spacers can be mounted on the heatsink directly, what makes the mounting of amp easier. Actually the threads are in the "heatsink" (this piece of aluminum you saw in my posts) and the spacer just gets pressed to the heatsink with the screws what goes through the spacer.
The new ones will be screwed to the heatsink directly and will have threads for screws
so the mounting of amp can also be done by the customers.
Unfortunately I don't have a milling machine so I need to wait for a friend to make the spacers.
Hi guys,
short update:
I did a lot of tests and measurments the last time and am now working on a final, improved version of this amplifier.
The DSP board will be replaceable (not fix soldered). There will also be a version without DSP and pure analoge volume control.
The amplifier also gets a 12V fan control circuit and a improved input stage.
Attached to this post you see the new replaceable DSP board with analoge volume control (optional) and RCA in/out jacks.
Stay tuned, I will order the new boards around mid of Jan!
Regards
Bernhard
short update:
I did a lot of tests and measurments the last time and am now working on a final, improved version of this amplifier.
The DSP board will be replaceable (not fix soldered). There will also be a version without DSP and pure analoge volume control.
The amplifier also gets a 12V fan control circuit and a improved input stage.
Attached to this post you see the new replaceable DSP board with analoge volume control (optional) and RCA in/out jacks.
Stay tuned, I will order the new boards around mid of Jan!
An externally hosted image should be here but it was not working when we last tested it.
Regards
Bernhard
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.