I'm curious about the PSU schematics, could you post them or send them to me?
I'm also curious about the switching frequencies you are using for the UC3854 and UC3875
Also I think that the heatsinks and the inductors shown in the pictures look pretty small for a 1,000W PFC since usual PFC efficiencies aren't better than 95% and this means 50W of dissipation at full load. Also 1,000W output requires 16A@90V AC [plus ripple] peak current from the switch and the diode, and also an inductor allowing that current without saturation
In the other hand, I think this amplifier won't draw much more than 100W@8ohm and 200W@4ohms on average with music signals
Are your heatsinks and inductors sized for music use or for worst case?
I'm also curious about the switching frequencies you are using for the UC3854 and UC3875
Also I think that the heatsinks and the inductors shown in the pictures look pretty small for a 1,000W PFC since usual PFC efficiencies aren't better than 95% and this means 50W of dissipation at full load. Also 1,000W output requires 16A@90V AC [plus ripple] peak current from the switch and the diode, and also an inductor allowing that current without saturation
In the other hand, I think this amplifier won't draw much more than 100W@8ohm and 200W@4ohms on average with music signals
Are your heatsinks and inductors sized for music use or for worst case?
I have to give props to EVA on here. She is very perceptive and correct. The heatsinking is definitely sized for music use. I did a quick test with a dummy load at 1000W output but not for too long because of the heat sinking. The large black heatsink is the heatsink for the diode and the two extremely inadequete pieces of aluminum hold the boost FET. The particular MOSFET used there is a SUPER-247 package so it has no mounting hole. Keep in mind too that what you are looking at is a prototype unit that I did about 2 and a half years ago. There are many things I would do differently but at the time was limited by cost, time, and space.
The circuit for the PFC is not that much different than that specified in the TI datasheets and app notes. If I remember correctly, the switching frequency of the PFC is around 100kHz. The switch is a IRFPS37N50A. The boost inductor is a high flux MPP core with around 80-85 turns of 13AWG wire. The boost capacitor is only 450VDC (folks don't under rate parts this close) 1200uF.
The power supply will be shown next and I will go through some of the details for that at that time but here is a preview:
Zero Voltage Switching
UC3875
IRFBA22N50A transistors
isolated gate drive magnetics
isolated feedback
cutsom transformer and inductor design
400VDC preregulator input
+/-45V output.
200kHz switching frequency (100kHz for magnetics design).
STAY TUNEND!
The circuit for the PFC is not that much different than that specified in the TI datasheets and app notes. If I remember correctly, the switching frequency of the PFC is around 100kHz. The switch is a IRFPS37N50A. The boost inductor is a high flux MPP core with around 80-85 turns of 13AWG wire. The boost capacitor is only 450VDC (folks don't under rate parts this close) 1200uF.
The power supply will be shown next and I will go through some of the details for that at that time but here is a preview:
Zero Voltage Switching
UC3875
IRFBA22N50A transistors
isolated gate drive magnetics
isolated feedback
cutsom transformer and inductor design
400VDC preregulator input
+/-45V output.
200kHz switching frequency (100kHz for magnetics design).
STAY TUNEND!
I also wanted to add that when running the PFC at full power the peak current ratings begin to be approached for critical components, especially the diode. The PFC unit shown was never tested as low as 90VAC but closer to 110-110VAC only. Also remember that there are 4 amplifier channels as the final load and each amplifier will do 190WRMS@4 ohms and several hundred watts each for a pair when bridged. If this system were tested with sinewaves as the audio with all the channels driven to their maximum output power then there would be some major power supply thermal failures. That would include the boost diode, the boost transistor, and the current sense resistor. As the engineer, we can either design for all possible cases or to design for regular use cases. In this case, that would be with regular audio programming. The same thing holds in this case for some of the switching transistors in the phase-shifted bridge.
uh...yeah
Either I chose the wrong school or I am a complete idiot. I am going into my senior design this fall. your project looks way too much for me. If I were taking only senior design instead of the two 16 hours semesters ahead of me I think i could get it. Did you send out for those PCB's or do them your self. I am at Purdue University (well actually not the main campus but its sister campus) and we don't have the ability to make a PCB of that quality here. The main campus has just about anything though. Anywho.....I guess I will stick to my much easier class AB amp unless you have any other suggestions that wouldn't be too difficult to get finished and perfected. I have a bit of a problem in that if I build something I want it to operate at 100% of its capability. Othetwise I will pick something else. Anywho.......AWESOME PROJECT! I bow to your superior knowledge and know-how.
Either I chose the wrong school or I am a complete idiot. I am going into my senior design this fall. your project looks way too much for me. If I were taking only senior design instead of the two 16 hours semesters ahead of me I think i could get it. Did you send out for those PCB's or do them your self. I am at Purdue University (well actually not the main campus but its sister campus) and we don't have the ability to make a PCB of that quality here. The main campus has just about anything though. Anywho.....I guess I will stick to my much easier class AB amp unless you have any other suggestions that wouldn't be too difficult to get finished and perfected. I have a bit of a problem in that if I build something I want it to operate at 100% of its capability. Othetwise I will pick something else. Anywho.......AWESOME PROJECT! I bow to your superior knowledge and know-how.
Hi BeanZ,
I just dropped in to see how the project was going.
I was wondering whether or not you could create a simple block diagram with the different modules that you are using. In particular, I'm interested in the power supply section. I have a TA0105 eval board and need to create a power supply for it.
Am I correct in seeing that you're going to be using a switched power supply to create your +/- rails for the Tripath Parts? One thought I had was to do this and have the ability to modify the output voltage, to control the volume levels (60-90V). This could remove the need for a preamp. I don't know how much volume control this would infact provide and/or whether it would cause issues with the operation of the amplifier.
Are you still going to need large cap resevoirs for the rail voltages with this system? I've been looking into a regular troidal transformer PSU, and the CAPs can get quite pricy (as do the troidal transformers.)
Thanks,
kiwi
I just dropped in to see how the project was going.
I was wondering whether or not you could create a simple block diagram with the different modules that you are using. In particular, I'm interested in the power supply section. I have a TA0105 eval board and need to create a power supply for it.
Am I correct in seeing that you're going to be using a switched power supply to create your +/- rails for the Tripath Parts? One thought I had was to do this and have the ability to modify the output voltage, to control the volume levels (60-90V). This could remove the need for a preamp. I don't know how much volume control this would infact provide and/or whether it would cause issues with the operation of the amplifier.
Are you still going to need large cap resevoirs for the rail voltages with this system? I've been looking into a regular troidal transformer PSU, and the CAPs can get quite pricy (as do the troidal transformers.)
Thanks,
kiwi
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.