Yes. This Sunday at Burning Amp '18 Wayne Colburn is presenting a really cool discrete opamp that can be doubled up for balanced.
Hang tight for a few more days.
I am going to resurrect my blown F5T. Don't ask how. It is embarrasing.
Anyway, for this build, I will be removing 1/2 of the output devices, and going to two pairs of outputs per channel, biased into a 4 ohm load.
In the original build I used the DIYaudio store boards and a toroidal power transformer with 5 center tapped secondaries.
I used only 4 secondaries to created a sort of "dual mono" F5T V3 in a single enclosure. If memory serves, 48 volt rails.
I have a couple of questions:
Since I have a free secondary, it occurred to me..... can I use that to power the input boards for both channels?
I used one diode bridge ground decoupler per channel ( That thing on the Primary side of the transformer ).. How many should I actually use?
Will I need a third for the input board supply? or should I only have used one in the first place?
Any help would be greatly appreciated.
THANKS IN ADVANCE!
Anyway, for this build, I will be removing 1/2 of the output devices, and going to two pairs of outputs per channel, biased into a 4 ohm load.
In the original build I used the DIYaudio store boards and a toroidal power transformer with 5 center tapped secondaries.
I used only 4 secondaries to created a sort of "dual mono" F5T V3 in a single enclosure. If memory serves, 48 volt rails.
I have a couple of questions:
Since I have a free secondary, it occurred to me..... can I use that to power the input boards for both channels?
I used one diode bridge ground decoupler per channel ( That thing on the Primary side of the transformer ).. How many should I actually use?
Will I need a third for the input board supply? or should I only have used one in the first place?
Any help would be greatly appreciated.
THANKS IN ADVANCE!
Thanks Hiraki1! I was trying to find that reference, but did not get to the f5 thread yet.
There is going to be 2 pairs of outputs per channel.
As I understand it, the heat sinking poses the class A limit. I was running 4 pairs/channel into an 8 ohm load until the incident.
4 ohm bias is by using a high wattage 4 ohm resistor on the speaker outputs.
They will be hooked up to a pair of speakers that are nominally 4 ohms.
The 48 Volts is just because I have the transformer. I was using it as an isolation transformer providing balanced power, but I did not like it.
As I understand it, the heat sinking poses the class A limit. I was running 4 pairs/channel into an 8 ohm load until the incident.
4 ohm bias is by using a high wattage 4 ohm resistor on the speaker outputs.
They will be hooked up to a pair of speakers that are nominally 4 ohms.
The 48 Volts is just because I have the transformer. I was using it as an isolation transformer providing balanced power, but I did not like it.
To me, 4 ohm bias means full class A until clipping into 4 ohms.
This means 2.5A bias, good thing you have two pairs of mosfets so they can share the load and spread the heat.
The problem is achieving 2.5A bias with 48v rails. It’s going to require big, big monoblocks or some sort of creative heat management. Creative heat management is always interesting to see on this forum.
This means 2.5A bias, good thing you have two pairs of mosfets so they can share the load and spread the heat.
The problem is achieving 2.5A bias with 48v rails. It’s going to require big, big monoblocks or some sort of creative heat management. Creative heat management is always interesting to see on this forum.
Green Cooling | Liquid Cooling | Green Revolution Cooling
"OEM servers are immersed in a cooling rack filled with a proprietary, nontoxic, non-conductive coolant called ElectroSafe™, which provides 1200X the heat capacity of air. Heat from the servers is absorbed by the coolant and quickly removed from the rack. The result is a supremely energy-efficient data center cooling system with a cooling capacity up to 100 kW/rack."
You need a small version of this using "ElectroSafe™" or mineral oil or what ever it is to move the heat outside of the house. Just immerse the amp board and heat sink. Use a thermostatically controlled pump to circulate the oil through a radiator outside, like an air conditioner does, once the amp reaches "warmed up" temperature.
Or lower the rail voltage.
Rush
Just as a reference point. I have monoblock F5V3 (4 pairs of fets) with total bias of 2.1 amps for about 70w class A into 8ohms. I’m using 2 chassis that are 16”deep and 6.5”high with 2” heatsinks. They run at a reasonable temperature. You can keep your hands on the heatsinks without burning them so probably in the mid 40c range. I kept the power supply heat out of the amplifier box by using separate power supply boxes. This helps to keep things a bit cooler as the amp chassis does not have to deal with rectifier and transformer heat and the transformer also impedes airflow through the chassis. I have good ventilation for convection through the amp chassis which also helps to keep things cool.
Water cooling is a neat idea but who wants to deal with heat transfer fluid, pumps, radiators fans, tubing, noise from fans and pumps etc. Look at some of the offerings from Pass Labs. Some of the class A push pull amps do not run huge class A power. You can look at the size of the amplifiers and get a good idea what conservatively designed heatsinking looks like for a given class A output level.
Water cooling is a neat idea but who wants to deal with heat transfer fluid, pumps, radiators fans, tubing, noise from fans and pumps etc. Look at some of the offerings from Pass Labs. Some of the class A push pull amps do not run huge class A power. You can look at the size of the amplifiers and get a good idea what conservatively designed heatsinking looks like for a given class A output level.