If it's of interest, the factory version (First Watt F3) gets to about 45*C at about room temp 28*C - my 'clone got 'upped' a bit and runs a bit higher current and I found it sounded slightly better at about 55*C on the sinks - used those Thermo isolation pads so the transistors were only a couple of degrees hotter t(han using mica & goop) until mid summer and they got to about 65*C plus -
I imagined they sounded better but I did same thing with small fan and cooled it down a bit 'just in case'
There's just something 'right' about the sound of this amp,IMO ....
I imagined they sounded better but I did same thing with small fan and cooled it down a bit 'just in case'
There's just something 'right' about the sound of this amp,IMO ....
Is James' "better" due to extra output bias current, or due to hotter devices?
Do hotter devices with less power capability, perform better than cooler devices that might last ten times longer?
I suspect the "better" is due to extra output bias current and the more effective cooling gives the ten times longer life while the devices have the same power capability due to running at the cooler Tc.
Do hotter devices with less power capability, perform better than cooler devices that might last ten times longer?
I suspect the "better" is due to extra output bias current and the more effective cooling gives the ten times longer life while the devices have the same power capability due to running at the cooler Tc.
I'm not sure Andrew - I've held the opinion over the years that keeping power fets hot will give you a better sound, but this obviously depends on all sorts of other things too, and this is a very simplistic (and limited) view of fet operating reality
I vaguely remember some discussion about this also in the F5 development and I think the highest safe operating temp (for the Fet junctions) was below 100*C so we were encouraged to use the best thermo transfer insulator material available.
Now, the sound difference with the extra temp is a bit of a puzzling thing - when I ran it at about 1.5A and low 40's temp, it certainly lacked some life and when I boxed up the sinks and raised the temp to mid 50's, it certainly came more alive (sorry about vague terms)
However, with higher 1.7A current, the sound quite definitely became more dynamic, etc as you would expect, and the heatsink temp got to about the mid 50's with open fins so it appears to be a combination of higher current and higher junction temp - I'm hesitating to say an opinion here as I again pushed the current up past the 1.8A and again achieved an improvement in the amp's sonics and the heatsink get to over 60*c.
In the summer here on hot days, amp h/sink can get well over the 70's but didn't notice any change in the sound, but when I tried fan cooling to low 50's, sound deteriorated a little, became 'thicker' or perhaps slower and lost some detail which is one of this amp's better qualities and I set it at about the high 50*C point and that's where it now stays.
Curiously, I built a slightly different circuit of Patrick's (EUVL) with the same/similar components (the DAO headamp) and it runs about 350mA but have the heatsinks sizzling about 70*C all the time (estimate junction temp over 110*C) and this does make quite a difference in the sound via the AKG k701 heaphones (this is added just to confuse!!)- I do think I'm pushing things a bit here, but ....
Sorry this meandering chatter is so imprecise and lacking in measurable details, test results, etc - [it's Sat night here and my mind is on holiday!]
I built this second F3 with Peter D's boards and they're more compact than the factory amp and have different resistors, the original design electro signal caps (for active current source) are now big Propylene caps and some other component differences have made it a much "faster" amp (sorry about the vague description)
I vaguely remember some discussion about this also in the F5 development and I think the highest safe operating temp (for the Fet junctions) was below 100*C so we were encouraged to use the best thermo transfer insulator material available.
Now, the sound difference with the extra temp is a bit of a puzzling thing - when I ran it at about 1.5A and low 40's temp, it certainly lacked some life and when I boxed up the sinks and raised the temp to mid 50's, it certainly came more alive (sorry about vague terms)
However, with higher 1.7A current, the sound quite definitely became more dynamic, etc as you would expect, and the heatsink temp got to about the mid 50's with open fins so it appears to be a combination of higher current and higher junction temp - I'm hesitating to say an opinion here as I again pushed the current up past the 1.8A and again achieved an improvement in the amp's sonics and the heatsink get to over 60*c.
In the summer here on hot days, amp h/sink can get well over the 70's but didn't notice any change in the sound, but when I tried fan cooling to low 50's, sound deteriorated a little, became 'thicker' or perhaps slower and lost some detail which is one of this amp's better qualities and I set it at about the high 50*C point and that's where it now stays.
Curiously, I built a slightly different circuit of Patrick's (EUVL) with the same/similar components (the DAO headamp) and it runs about 350mA but have the heatsinks sizzling about 70*C all the time (estimate junction temp over 110*C) and this does make quite a difference in the sound via the AKG k701 heaphones (this is added just to confuse!!)- I do think I'm pushing things a bit here, but ....
Sorry this meandering chatter is so imprecise and lacking in measurable details, test results, etc - [it's Sat night here and my mind is on holiday!]
I built this second F3 with Peter D's boards and they're more compact than the factory amp and have different resistors, the original design electro signal caps (for active current source) are now big Propylene caps and some other component differences have made it a much "faster" amp (sorry about the vague description)
more current means modulated current is lesser percentage of steady one
more heat means heat modulation is lesser percentage of steady one
go figure
left leg usually goes combined with right one
especially if you have two pairs of them
legs
term combined is where one needs to point peripheral vision
more heat means heat modulation is lesser percentage of steady one
go figure
left leg usually goes combined with right one
especially if you have two pairs of them
legs
term combined is where one needs to point peripheral vision
Last edited:
Hmm, maybe I should put it back on 'the list' -
I have 'difficult' speakers (Coral beta8, > 125Hz) and the balanced F3 might just 'do the trick' with the Aikido preamp and the 'tone controls'!
I got a bit creative with some of the components and the power supply for the standard design F3 and it definitely made a big difference - it's 'sort of' yesterdays amp, but it just sounds good, IMO!
I have 'difficult' speakers (Coral beta8, > 125Hz) and the balanced F3 might just 'do the trick' with the Aikido preamp and the 'tone controls'!
I got a bit creative with some of the components and the power supply for the standard design F3 and it definitely made a big difference - it's 'sort of' yesterdays amp, but it just sounds good, IMO!
