4 parallel caps is going to drop ESR and ESL down pretty low.
I suspect it's a combination of 46% more capacitance and a possibly better cap.
The transformer probably has little to do with sound quality once you have clean DC available to the amplifier.
I suspect it's a combination of 46% more capacitance and a possibly better cap.
The transformer probably has little to do with sound quality once you have clean DC available to the amplifier.
Is this heatsink - 350AB length of 200 mm http://www.farnell.com/datasheets/316955.pdf good for one ch - 0.42 C / W
labjr--the difference is very slight-i can only hear a difference because i have lived with both and know which one i put in the system---if you switched them on me i probably couldn't tell you which one it was
quest of heatsinks reloaded
Hi
I wonder how effective were the heatsinks from Mr. Pass´PLH article.
Power dissipation is very close to F5.
http://www.passdiy.com/pdf/PLH_amplifier.pdf, page 8.
Thanks
Zen Mod said:
Hi
I wonder how effective were the heatsinks from Mr. Pass´PLH article.
Power dissipation is very close to F5.
http://www.passdiy.com/pdf/PLH_amplifier.pdf, page 8.
Thanks
No reason one can't get longer rectangular tubes for more dissipation. It's all about surface area.
vdi_nenna said:
I had thought about doing something like that. However, I wondered if there would be a great savings if any at all by the time you're through. Plus there's probably thermal loss from bolting it all together etc. The thing may have to be pretty large for an F5.
labjr said:
As noted in the article, the PLH dissipates 70-90W/channel - ie. more than the F5.
-j
Hello sheafer
in regards to aussie amp or greg ball power supplies there capacity is a fraction of what the f5 and aleph 3 should be .would this not effect the amps perform?the aleph 3 requires 192000uf of total capacity and the f5 120000uf. wouldn't I want to use these values?any input apprieciated!!!!thanks Peter
in regards to aussie amp or greg ball power supplies there capacity is a fraction of what the f5 and aleph 3 should be .would this not effect the amps perform?the aleph 3 requires 192000uf of total capacity and the f5 120000uf. wouldn't I want to use these values?any input apprieciated!!!!thanks Peter
You cannot compare cap-sizes between a ClassA-amp and a ClassB-amp; ClassA needs always more, the F5 being ClassAB, so something in between.
And mind you, a lot of people just exagerate grossly the cap size, some think it sounds better and others have just too much money 😀
Nelson's suggestions are always on the nice side, so don't worry about that.
Have fun, Hannes
And mind you, a lot of people just exagerate grossly the cap size, some think it sounds better and others have just too much money 😀
Nelson's suggestions are always on the nice side, so don't worry about that.
Have fun, Hannes
labjr said:
It is, up to 25 watts, into an 8 Ohm load, then goes into AB operation
above that, i.e. class A up to the bias/idle current. From what I recall reading in the F5 owners manual.
the aussie mps psu has 60,000 per rail
i dont see how this is not enough for a 25w class a amp
imho powersupply capacitance is a perfect example of the
"if some is good then more must be better" syndrome
i dont see how this is not enough for a 25w class a amp
imho powersupply capacitance is a perfect example of the
"if some is good then more must be better" syndrome
slr 5000 said:
Yeah . . .
Anyhow, the size of power supply capacitance is to be decided based on the current draw that the amp requires and the design value of ripple voltage. According to text book, the ripple voltage is directly proposional to the amp's current draw and inversely proposional to the size of psu capacitance. I think that the audio designers try to properly play themselves on the ground of this triangular relationship.
>🙂<
Hi
A simple aproach:
Someone may correct me.
C/W = (Tmax - Tamb) / Pd
Tmax = 55C (best "finger touch test")
Tamb = 27C (worse room confort, no air conditioner)
Pd = Watts of power dissipation per hestsink extrusion block
C/W = (55-27)/68 = 0,411.. Poor margin IMO. You need bigger heatsinks, lower C/Ws.
Hope this helps. Cheers
samoloko said:
A simple aproach:
Someone may correct me.
C/W = (Tmax - Tamb) / Pd
Tmax = 55C (best "finger touch test")
Tamb = 27C (worse room confort, no air conditioner)
Pd = Watts of power dissipation per hestsink extrusion block
C/W = (55-27)/68 = 0,411.. Poor margin IMO. You need bigger heatsinks, lower C/Ws.
Hope this helps. Cheers
Heat sink thermal resistance
Make sure you know, (or calculate) your heat sink's thermal resistance (C/W) the desired temperature rise above ambient (30-35 degrees C). Most are quoted at a temperature rise of 80 degrees C. Heat sinks get more efficient as they get hotter !! See attached for an illustration.
A stereo F5 or bridged mono F5 must dissipate ~ 35 watts PER FET. This means a total thermal resistance of 0.25 for a stereo amp at a 35 degree C delta T.
I am waiting for barrredboss's new extrusions, before I select a sink for my bridged F5's The more I've looked into this, the more I agree with Pass.....if you want your amps to last until you're deaf, dead or the sun goes out, (whichever comes first), you can't choose sinks with too low a thermal resistance !!
Make sure you know, (or calculate) your heat sink's thermal resistance (C/W) the desired temperature rise above ambient (30-35 degrees C). Most are quoted at a temperature rise of 80 degrees C. Heat sinks get more efficient as they get hotter !! See attached for an illustration.
A stereo F5 or bridged mono F5 must dissipate ~ 35 watts PER FET. This means a total thermal resistance of 0.25 for a stereo amp at a 35 degree C delta T.
I am waiting for barrredboss's new extrusions, before I select a sink for my bridged F5's The more I've looked into this, the more I agree with Pass.....if you want your amps to last until you're deaf, dead or the sun goes out, (whichever comes first), you can't choose sinks with too low a thermal resistance !!
Attachments
Re: Heat sink thermal resistance
One thing I have had trouble with when coming up with a good heatsink is not just the Tsink-avg, but the Tjunction. The Tj is probably your most critical parameter for longevity. The TO-247 Device max operating Tj is 150Deg C. Nominal operating temperature is 74Deg C. I think Mr pass has indicated a good comprimise is about 100Deg C.
It takes a lot of effort to get a TO-247 dissipating 32 Watts to operate under 100Deg C. Bodging the device to the heatsink will not work for a device operating in this range.
Here is some advice I recently recieved from a Thermal Engineer at R-Theta.
"All three options will lead to the same results if the input is properly entered. I typically use W/m.K and thickness for the thermal interface. The thickness is quite critical. It could mean a difference in 10-20 degrees higher in the simulation program if you put it too thick.
Personally I don’t like using thermal impedance (C.in2/W) because the value will have to change if you change the device size. Temperatures can get out of hand pretty quickly if you don’t adjust the values.
If you have a working thermal resistance (C/W) for the interface, then it is easiest to work with"
Regards
Anthony
dcbingaman said:
One thing I have had trouble with when coming up with a good heatsink is not just the Tsink-avg, but the Tjunction. The Tj is probably your most critical parameter for longevity. The TO-247 Device max operating Tj is 150Deg C. Nominal operating temperature is 74Deg C. I think Mr pass has indicated a good comprimise is about 100Deg C.
It takes a lot of effort to get a TO-247 dissipating 32 Watts to operate under 100Deg C. Bodging the device to the heatsink will not work for a device operating in this range.
Here is some advice I recently recieved from a Thermal Engineer at R-Theta.
"All three options will lead to the same results if the input is properly entered. I typically use W/m.K and thickness for the thermal interface. The thickness is quite critical. It could mean a difference in 10-20 degrees higher in the simulation program if you put it too thick.
Personally I don’t like using thermal impedance (C.in2/W) because the value will have to change if you change the device size. Temperatures can get out of hand pretty quickly if you don’t adjust the values.
If you have a working thermal resistance (C/W) for the interface, then it is easiest to work with"
Regards
Anthony