I really enjoyed making my first one.....so I sold it so I could make a second 
This time its dual mono and regulated or high-cap with snubber, still haven't decided.
BrianGT amp PCBs, Digi01 reg PCB, 25Vx2 160VA transformers and 2x 6800uF caps per LM3875, custom aluminium case work by me
.......but I think like the the last one it will be about 20 mins of soldering and 20 hours of casework 
This time its dual mono and regulated or high-cap with snubber, still haven't decided.
BrianGT amp PCBs, Digi01 reg PCB, 25Vx2 160VA transformers and 2x 6800uF caps per LM3875, custom aluminium case work by me
.......but I think like the the last one it will be about 20 mins of soldering and 20 hours of casework
They did arrive and 2-3 months later Im donemaxw said:Now im waiting for my BrianGT LM3886 kits to arrive
maxw said:Still haven't decided whether to use regulation of high-cap on these thoughbut I guess I will have to try both and decide.
I have decided I like the snubber high-cap one better.
So, all done! I get 20mV of DC offset in each channel and I'm not using the Ci caps. Its made totally from recycled metal so the finish is not perfect at all but I quite like it this way. The rear panel doesn't even get warm when in use and it sounds great
.Attachments
Assuming a same unloaded voltage for the PSU: Could it be because with an unregulated supply the supply drops during use and thus the voltage across the IC is less? With a regulated supply the same amount of voltage is always supplied to the IC so when power is drawn the current increases but voltage doesn't drop. That means higher power is dissipated by the IC. Also, an unregulated supply, with voltage droop, will clip sooner than a regulated suppply which lowers the power dissipated by the IC.
The heatsink or box may be getting warmer because the voltage regulator circuit is dissipating power (that's how it regulates the voltage)- i.e. generating heat.
If the regulator chips/pass transistors are mounted to the same box/heatsink as the amp it is easy to be fooled into thinking the amp chips are running hotter. The extra heat is coming from the regulator, not the amp chips. Linear voltage regulators are not very efficient - they are class A amplifiers of the reference voltage.
If your unregulated supply voltage drops dramatically while the amp is playing, to the point where the heat generated by the chip is noticably lower than that from a "same voltage" regulated supply, your unregulated power supply is grossly underdesigned and needs a bigger transformer and/or bigger storage capacitors. It seems to me that If you can feel the difference in temperature there must be a huge audible difference in performance and it is very hard to imagine that the difference in performance when using an under designed power supply is an improvement.
Amplifiers driving low impedance loudspeakers need low impedance power supplies. That means you need a lot of storage capacitance.
Capacitance is cheap. There is no reason not to use a lot of it.
I_F
If the regulator chips/pass transistors are mounted to the same box/heatsink as the amp it is easy to be fooled into thinking the amp chips are running hotter. The extra heat is coming from the regulator, not the amp chips. Linear voltage regulators are not very efficient - they are class A amplifiers of the reference voltage.
If your unregulated supply voltage drops dramatically while the amp is playing, to the point where the heat generated by the chip is noticably lower than that from a "same voltage" regulated supply, your unregulated power supply is grossly underdesigned and needs a bigger transformer and/or bigger storage capacitors. It seems to me that If you can feel the difference in temperature there must be a huge audible difference in performance and it is very hard to imagine that the difference in performance when using an under designed power supply is an improvement.
Amplifiers driving low impedance loudspeakers need low impedance power supplies. That means you need a lot of storage capacitance.
Capacitance is cheap. There is no reason not to use a lot of it.
I_F
Nuuk said:I don't know how to put it any more simply!
Yes, I know the voltages are actually lower (for the same transformer) but my chips were running hotter. I'm sure that I have heard other people report the same.
I suppose you are talking about the LM338?
Depends on the resistors you use, but certainly with typical values there is some constant current that may be higher that what the chip needs at idle.
So, that's normal.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.