Ok then. Using the information I have gathered from here and other resources, I have decided on this transformer to power the mono LM3886TF at 4 Ohms running a 70 Watt speaker.
Digi-Key - 237-1265-ND (Manufacturer - VPS20-2200)
Does that sound correct for use with the following Power Supply circuit?
http://chipamp.com/docs/lm3886-manual.pdf (page 4)
Digi-Key - 237-1265-ND (Manufacturer - VPS20-2200)
Does that sound correct for use with the following Power Supply circuit?
http://chipamp.com/docs/lm3886-manual.pdf (page 4)
Last edited:
Cory,
That transformer will only give +-14V so the chip will barely be working. I'd go for 40V centre tapped, ie 20,0, 20 to give +-28V supply rails. I would use a toroidal transformer.
The power supply circuit is fine but I wouldn't use the amp circuit because it has no DC blocking capacitor at input.
Frank
That transformer will only give +-14V so the chip will barely be working. I'd go for 40V centre tapped, ie 20,0, 20 to give +-28V supply rails. I would use a toroidal transformer.
The power supply circuit is fine but I wouldn't use the amp circuit because it has no DC blocking capacitor at input.
Frank
Cory,
slow down !!!!
You have a brief, maximum output of 60W to 70W into 4r0 resistor.
Now develop a specification. This is the theoretical design stage, using where necessary information gathered by others with experiments on what the chip needs and can do.
Post18 middle table.
68W into 4r0 using +-28Vdc. This is the required voltage at the power input pins while the amplifier is delivering that maximum power. When on quiescent duty the voltage from the supply will be higher. For simplicity let's assume that this voltage rise is zero.
pic 3 of the post shows the National recommended heatsink for a variety of operating conditions.
What is your worst case operating ambient temperature? Is it 20degC, requiring you to shut down when the air conditioning/cooling goes faulty or can't cope with an audience of 2000 dancing and singing fans? or 30degC for warm conditions and an external heatsink, or 50degC with an internal heatsink. You have to decide based on experience. This is called informed decision making. Don't guess.
Let's assume external heatsink and Ta<=25degC, Vcc=56Vdc shows 4ohm crossing the 56V point opposite 1.9C/W. I recommend doubling the heatsink size to 0.95C/W for T chipamp.
For TF take care. it might not survive.
Let's change that +-28Vdc + zero assumption. Mains voltage rise +6%, transformer regulation +8%, total rise ~ 14% of 28 to ~+-32Vdc, Vcc=64V, check that pic3 again. Oh ! we are off the top of the graph (the conditions that the graph assumes are T chip directly mounted on the heatsink without an insulator !!!!!!)
This tells me that a TF chipamp is probably not suitable.
National publish a spreadsheet showing their chipamps with various assembly components. Download it. It will show if National "red card" your brief.
Are you becoming aware of how slowly you work through the specification to arrive at the answers?
Can you now see why so many were cautioning you with your "Grande" proposal?
slow down !!!!
You have a brief, maximum output of 60W to 70W into 4r0 resistor.
Now develop a specification. This is the theoretical design stage, using where necessary information gathered by others with experiments on what the chip needs and can do.
Post18 middle table.
68W into 4r0 using +-28Vdc. This is the required voltage at the power input pins while the amplifier is delivering that maximum power. When on quiescent duty the voltage from the supply will be higher. For simplicity let's assume that this voltage rise is zero.
pic 3 of the post shows the National recommended heatsink for a variety of operating conditions.
What is your worst case operating ambient temperature? Is it 20degC, requiring you to shut down when the air conditioning/cooling goes faulty or can't cope with an audience of 2000 dancing and singing fans? or 30degC for warm conditions and an external heatsink, or 50degC with an internal heatsink. You have to decide based on experience. This is called informed decision making. Don't guess.
Let's assume external heatsink and Ta<=25degC, Vcc=56Vdc shows 4ohm crossing the 56V point opposite 1.9C/W. I recommend doubling the heatsink size to 0.95C/W for T chipamp.
For TF take care. it might not survive.
Let's change that +-28Vdc + zero assumption. Mains voltage rise +6%, transformer regulation +8%, total rise ~ 14% of 28 to ~+-32Vdc, Vcc=64V, check that pic3 again. Oh ! we are off the top of the graph (the conditions that the graph assumes are T chip directly mounted on the heatsink without an insulator !!!!!!)
This tells me that a TF chipamp is probably not suitable.
National publish a spreadsheet showing their chipamps with various assembly components. Download it. It will show if National "red card" your brief.
Are you becoming aware of how slowly you work through the specification to arrive at the answers?
Can you now see why so many were cautioning you with your "Grande" proposal?
Last edited:
Andrew,
I use the TF in an Australian summer (ie, a bit warmish as you can imagine) in an active set up and play it loud enough for the bloke in the next street to hear it. The chances of even getting close to maximum power at 35V rails are about zero.
From what I understand, Cory will going active and even if he does push the chip into protection mode, he will learn something. Even in a bridged TF configuration using 35V rails, protection hasn't cut in yet for me. Cory's post about bragging rights explains what he is after.
Frank
I use the TF in an Australian summer (ie, a bit warmish as you can imagine) in an active set up and play it loud enough for the bloke in the next street to hear it. The chances of even getting close to maximum power at 35V rails are about zero.
From what I understand, Cory will going active and even if he does push the chip into protection mode, he will learn something. Even in a bridged TF configuration using 35V rails, protection hasn't cut in yet for me. Cory's post about bragging rights explains what he is after.
Frank
It has been hard to find real world practical explanations of whats provided in the datasheets. I have a far better idea of what those are telling me now IMHO, after just completing an LM386, you should look at the 5 pin chipamps such as the TDA2050. They need about the same amount of components as the LM386 and you can run them off low voltage supplies (at reduced output power) until you get an adequate supply. The TDA2050 will operate at +5/-5 volts (or less) split rail supply for safe building.
As you move into these stronger chips, heatsinking and layout become important. First time power up should always be performed at lower voltages and limited current to avoid smoking your creation in case of a wiring error.
As you move into these stronger chips, heatsinking and layout become important. First time power up should always be performed at lower voltages and limited current to avoid smoking your creation in case of a wiring error.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.