Darn...... I should have done my heatsinks like David did...... gives me a place to keep my cup of coffee hot...... or maybe cook my "Hot Pockets" for breakfast......
Nice job on the construction......!
F5 Monoblocks
The horizontal heatsinks work fine and you can leave the palm of your hand on them for as long as you like (so coffee would be tepid). They are Conrad MF35 151.5 types rated at 0.21 degree C/W. As an afterthought I may drill some holes in the side plates to let some air get in and out just to keep the electros happy although the whole case acts as a heat dissipator anyway. They are 105 degree type caps.
The bridge rectifiers are GBJ2004 types, 400V @ 20A's and there is one bridge used on each rail. They are no warmer than the surrounding components.
I forgot to mention in my original post that I have a WIMA 10uf on each rail as well. They seemed to give a little more dynamics to the mids and high frequencies. Buoyed with that success I have ordered 48 PETP 22uf Russian caps and will put 12 on each rail (hopefully they will all fit). Whatever they do, I'm sure it won't be bad.
I listened to a CD last night "Bits, Bytes @ Pipes" produced by Organ Grinder Recordings. This CD features the Wurlitzer at the Organ Grinder Restaurant in Portland, Oregon. Track 13, Warsaw Concerto, has bass extension almost to DC whilst still pushing out plenty of higher pitched instruments at the same time. There was absolutely no sign of the highs being squeezed out by the low notes. My osborn speakers feature 4 x 13" Focal drivers at 92db efficiency so it goes without saying that the listening experience was awsome. Sorry to rabbit on so much but I love the F5's.
David.
The horizontal heatsinks work fine and you can leave the palm of your hand on them for as long as you like (so coffee would be tepid). They are Conrad MF35 151.5 types rated at 0.21 degree C/W. As an afterthought I may drill some holes in the side plates to let some air get in and out just to keep the electros happy although the whole case acts as a heat dissipator anyway. They are 105 degree type caps.
The bridge rectifiers are GBJ2004 types, 400V @ 20A's and there is one bridge used on each rail. They are no warmer than the surrounding components.
I forgot to mention in my original post that I have a WIMA 10uf on each rail as well. They seemed to give a little more dynamics to the mids and high frequencies. Buoyed with that success I have ordered 48 PETP 22uf Russian caps and will put 12 on each rail (hopefully they will all fit). Whatever they do, I'm sure it won't be bad.
I listened to a CD last night "Bits, Bytes @ Pipes" produced by Organ Grinder Recordings. This CD features the Wurlitzer at the Organ Grinder Restaurant in Portland, Oregon. Track 13, Warsaw Concerto, has bass extension almost to DC whilst still pushing out plenty of higher pitched instruments at the same time. There was absolutely no sign of the highs being squeezed out by the low notes. My osborn speakers feature 4 x 13" Focal drivers at 92db efficiency so it goes without saying that the listening experience was awsome. Sorry to rabbit on so much but I love the F5's.
David.
Aaahhh..... tepid--that's just how I like my coffee....!
Kidding aside, I'm running the same Conrad heatsinks in my amp, with the fins running vertically. If I have any problems, it's that my heatsinks (and therefore, probably the MOSFETs) are running too cool.
(I would, though, put some ventilation holes in the cabinet, to give the internal components a little air....)
Ken
Kidding aside, I'm running the same Conrad heatsinks in my amp, with the fins running vertically. If I have any problems, it's that my heatsinks (and therefore, probably the MOSFETs) are running too cool.
(I would, though, put some ventilation holes in the cabinet, to give the internal components a little air....)
Ken
Hi David
I am plan to build one , still collecting parts , if I use half of the heat sink ( ie MF35-75 ) can you imagine how hot it is , may be not too bad for Sydney summer , or may be I should go for MF35-100 , my space in the rack is limited
Very nice built David .
Regrads
The issue might be then: not, how hot it is in the room--but rather, how hot is it in your limited rack space?
F5 Monoblocks
As I write, the ambient in the vicinity of the F5's is 19 degrees C and the temperature in the middle of the heatsink is 41 degrees. The amps are behind each loudspeaker in a corner. They have been on for a few hours. Had the heatsinks been mounted vertically they might be a degree or two cooler.
I would expect that a heatsink half the size would get to a bit more than twice this temperature which to my mind would be verging on too hot for human safety let alone the occasional inquisitive cat, especially if your room was 30 degrees C in summer. Have a look at the Conrad site as there is usefull info on application of heatsinks. Cooling function of heatsinks is not linear either and their C/W/degree ratings are for an already elevated temperaure. You could use a smaller heatsink if you are prepared to use a fan but quiet fans are scarcer than hens' teeth.
Whatever might be the sonic benefits of running the FET's at higher temperatures I am more than happy with my sound.
As an aside, the temperature compensating elements (thermistors?) in the F5 are often seen mounted in free space where they get a sniff of the ambient in the vicinity of the FET. I chose to mount mine so that they actually touch the heatsink adjacent to where the FET's are mounted. I suppose they would track the temperature better but there may be some obscure reason why I shouldn't have done it this way.
David.
As I write, the ambient in the vicinity of the F5's is 19 degrees C and the temperature in the middle of the heatsink is 41 degrees. The amps are behind each loudspeaker in a corner. They have been on for a few hours. Had the heatsinks been mounted vertically they might be a degree or two cooler.
I would expect that a heatsink half the size would get to a bit more than twice this temperature which to my mind would be verging on too hot for human safety let alone the occasional inquisitive cat, especially if your room was 30 degrees C in summer. Have a look at the Conrad site as there is usefull info on application of heatsinks. Cooling function of heatsinks is not linear either and their C/W/degree ratings are for an already elevated temperaure. You could use a smaller heatsink if you are prepared to use a fan but quiet fans are scarcer than hens' teeth.
Whatever might be the sonic benefits of running the FET's at higher temperatures I am more than happy with my sound.
As an aside, the temperature compensating elements (thermistors?) in the F5 are often seen mounted in free space where they get a sniff of the ambient in the vicinity of the FET. I chose to mount mine so that they actually touch the heatsink adjacent to where the FET's are mounted. I suppose they would track the temperature better but there may be some obscure reason why I shouldn't have done it this way.
David.
Plus, do you control the room temp to below 30*C?
From the Conrad website, the MF35-75 has a thermal efficiency of 0.32 *C/W at the regulation measuring temp of 80*C, so you have to derate this by about 15% for the 45->50*C you probably want to run it at - so your lower efficiency is now 0.35*C/W.
And with about 70watt dissipation each channel for a standard build F5 (25Vrails, etc) this will raise the heatsink temp by about 25*C (0.35 X 70) IN FREE AIR for each channel, and if you have a room temp control at about the usual 28*C, the heatsink will be about 28 + 25 = 53*C sitting on the floor.
If not, with summer similar to Melb, 35+ is often the case which would make that heatsink pretty hot indeed, (35 + 25 = 60) but not unreasonable.
So this is where the CanAm words question is critical.
Adding 2 x 70 watts to a closed rack isn't a viable proposition without an exhaust fan, and having an open rack in a corner is better, but still...
The Conrad MF35-100 has 0.28 efficiency at 80*C, derated to 0.32*C/w - the MF35-151 has 0.21 derated to 0.24*C/w - you can do your own calculations.
Incidently, the "Elna for Audio" caps (12,000uF/63V) need to stay as cool as you can manage.
Also suggest you get some of those Keratherm fet isolation washers from the Group Buy here - makes a noticeable difference.
From the Conrad website, the MF35-75 has a thermal efficiency of 0.32 *C/W at the regulation measuring temp of 80*C, so you have to derate this by about 15% for the 45->50*C you probably want to run it at - so your lower efficiency is now 0.35*C/W.
And with about 70watt dissipation each channel for a standard build F5 (25Vrails, etc) this will raise the heatsink temp by about 25*C (0.35 X 70) IN FREE AIR for each channel, and if you have a room temp control at about the usual 28*C, the heatsink will be about 28 + 25 = 53*C sitting on the floor.
If not, with summer similar to Melb, 35+ is often the case which would make that heatsink pretty hot indeed, (35 + 25 = 60) but not unreasonable.
So this is where the CanAm words question is critical.
Adding 2 x 70 watts to a closed rack isn't a viable proposition without an exhaust fan, and having an open rack in a corner is better, but still...
The Conrad MF35-100 has 0.28 efficiency at 80*C, derated to 0.32*C/w - the MF35-151 has 0.21 derated to 0.24*C/w - you can do your own calculations.
Incidently, the "Elna for Audio" caps (12,000uF/63V) need to stay as cool as you can manage.
Also suggest you get some of those Keratherm fet isolation washers from the Group Buy here - makes a noticeable difference.
sudden drop in bias
After playing flawlessly for over a year, one of the channels has developed a problem. The voltage across the output(r11/r12) has dropped to about 250 mv and the offset is about 175mv. The other one is steady around 580mV and 0 offset. The bad channel still plays but with audible distortion.
As I get ready to open it up and troubleshoot, can someone give me some quick pointers to which component may have likely failed causing this? I am sure this information can be gleaned from somewhere in this long thread, but my searches so far has not been able to get at it specifically.
thanks
uv
After playing flawlessly for over a year, one of the channels has developed a problem. The voltage across the output(r11/r12) has dropped to about 250 mv and the offset is about 175mv. The other one is steady around 580mV and 0 offset. The bad channel still plays but with audible distortion.
As I get ready to open it up and troubleshoot, can someone give me some quick pointers to which component may have likely failed causing this? I am sure this information can be gleaned from somewhere in this long thread, but my searches so far has not been able to get at it specifically.
thanks
uv
Thanks. Will give these a try.
Could be the input devices. I had accidentally cranked up the preamp to max once when I turned it on recently. It's a tube preamp with a gain of 10 or 11db and it was hooked up to a cd player. So it must've sent quite a few volts through. Incidentally what is the input max voltage for the F5 using 2sk170/2sj74 on the input?
Could be the input devices. I had accidentally cranked up the preamp to max once when I turned it on recently. It's a tube preamp with a gain of 10 or 11db and it was hooked up to a cd player. So it must've sent quite a few volts through. Incidentally what is the input max voltage for the F5 using 2sk170/2sj74 on the input?