If you refer to the Thermal plots I posted in the first few pages of this thread from a very similar profile, I think it is reasonable to expect one 10" section of 66449 to stay under 55 Deg C. even when dissipating up to 50 watts.
Refer to Posts #20 and #21
BTW I have added a 5" profile to the wiki.
Regards
Anthony
Refer to Posts #20 and #21
BTW I have added a 5" profile to the wiki.
Regards
Anthony
Coulomb said:
Exactly my point... If a pair of Conrads is proving to be more than sufficient for an F5, why do I need 3 times the mass and twice the surface area? Shouldn't just two 10" be enough? I mean if I go Conrad I use 4 sinks for a pair of F4s. If I go R-theta, I need 8? Again, I am basing this on the Conrad thread where all seem happy with the dissipation and temps. Unless Conrad is being dishonest, they rate the MF35 at .32C/watt at 80degree rise for a 3" length.
I have added a 5" profile to the wiki for $32 CAD with mounting holes Drilled and tapped. 4 of these will give the same benefit as two of the Conrads if not a little more than the Conrads.
Although this is not $100 USD delivered it is pretty close to $100USD before delivery. The added benefit of more thermal capacity and the drilling & Tapping make it clearly a better deal than the Conrad Heatsinks.
Anthony
The idea that just because these weigh more than the Conrads and therefore disperse more heat is bogus. If that was the case, I would by some blocks of aluminum that weighed 15lbs each and avoid the extrusion altogether. There are a few more profiles from R-Theta themselves that would be more efficient per weight. The 9017 and 9022 are two good examples. The 9017(0.5C/watt) may even be cheaper because it uses 20% less aluminum and the 9022(0.8C/watt) is rated the same as the 9013 with nearly 1/2 the weight. Same cooling and less weight seem to be a good idea for shipping and racks.
I am not trying to throw stones here, just want the best value and performance. Seems to me that a 5" length of 9017 would do the same job as a 10" 9013. Unless of course the only goal is to have the biggest, heaviest, least efficient case possible? I mean a perfect heatsink would have a fairly healthy spreader plate and as many and as thin of fins as possible while still allowing some room for air.
I am not trying to throw stones here, just want the best value and performance. Seems to me that a 5" length of 9017 would do the same job as a 10" 9013. Unless of course the only goal is to have the biggest, heaviest, least efficient case possible? I mean a perfect heatsink would have a fairly healthy spreader plate and as many and as thin of fins as possible while still allowing some room for air.
I have started collecting actual numbers instead of just votes on the wiki. Could the interested members please convert thier votes of interest into actual numbers in the appropriate columns. Please be sure to check your entries carefully. Do not worry about editing the totals on the bottom, I will take care of that.
Anthony
Anthony
Wiki updated.
One question to Anthony, is that true that I will need to opt for the FPE option if I want the back panel cut for the part kit?
Thanks!!
One question to Anthony, is that true that I will need to opt for the FPE option if I want the back panel cut for the part kit?
Thanks!!
pchw said:
Well you can buy the parts kit without the FPE back panel, you can buy the FPE back panel without the parts kit, but if you want the back panel with the cutouts for the parts kit then yes you need to buy both.
Does this answer your question?
Anthony
ALL heatsinks MUST be rated for the temperature differential (DeltaT = Ta-Ts) at which they will operate.khundude said:
ALL semiconductors MUST be rated for the temperature that the case (Tc) will operate at.
A heatsink with double the surface area is a good indicator that it can dissipate more heat than the reference IF the average temperature of the whole sink is comparable to the reference sink.
A heavy heatsink that has double the surface area indicates that material thickness is likely to be adequate in dispersing the heat over the whole sink. This is equivalent to saying the average temeperature of this sink will be higher than the reference heatsink.
Coulomb's choice of three times the weight and double the surface area indicates that he knows how to identify an efficient heatsink that will perform better than the lighter and smaller Conrad.
updated Wiki
also, I did not mess with the column totals as you stated. This will really be cool if it goes through!! I really feel like part of a major project🙂 🙂 🙂
also, I did not mess with the column totals as you stated. This will really be cool if it goes through!! I really feel like part of a major project🙂 🙂 🙂
it's unfortunate that in this day and age we need to be trying to dissipate so much wasted heat...I would bet that have of the people who run hot-as-hell class A amplifiers also run their air conditioning all summer. It really makes me think that maybe we should be investigating making 'class d' sound better. 🙂
b
b
I am reposting some old Data and some new Data. This is hard Data generated from R-Theta specifically for the 9013 profile. This is gospel people please do even try arguing with it, it is factual Data.
I can generate this Data for pretty much any device configuration if I am provided enough Data on the Devices used and how many you are using per channel.
10" Profile, 2 x IRF240N, 22 Watts per device
I can generate this Data for pretty much any device configuration if I am provided enough Data on the Devices used and how many you are using per channel.
10" Profile, 2 x IRF240N, 22 Watts per device
Finally here is the type of Data I need if you want me to run a specific configuration for your application.
When reviewing these reports pay careful attention to Tjunction figures as most FET operate nominally at 75 Deg. C.
Secondly look at Tsink-avg, this is the number we want to see below 55 DEG C, ideally 50 or less.
Reducing the dissapated power per device will reduce both numbers, increasing the number of FETs will reduce power dissipated per device and lower Tjunction.
See the 10" 52 Watt report and compare Data with the other reports.
Regards
Anthony
When reviewing these reports pay careful attention to Tjunction figures as most FET operate nominally at 75 Deg. C.
Secondly look at Tsink-avg, this is the number we want to see below 55 DEG C, ideally 50 or less.
Reducing the dissapated power per device will reduce both numbers, increasing the number of FETs will reduce power dissipated per device and lower Tjunction.
See the 10" 52 Watt report and compare Data with the other reports.
Regards
Anthony
AndrewT said:
That is all fine and dandy and thus furthur illustrates my point. If the Conrads are sufficient for all FirstWatt amps, why the hell do I need 4 of these super fantastic man sized ones? If the "data sheets" say and your calculations show that the R-Theta should destroy the Conrads, why do I need twice as many to dissipate the heat? I mean, if the R-Theta perform that much better, shouln't two 8" or two 10" at the most be sufficient? Either that or all of the guys on the Conrad buys are full of it and telling tales about how good their purchase was.
I think you guys miss the point, I just don't see how they can be viewed as all that and a bag of chips and still need two to three times as much sink to achieve the same temps. I was told earlier in this thread that I would need at least 4, 10" heatsinks for an F4 and even that might not be enough. Now, a 5" kit comes up that is good for an F5? Hugh? I understand that the F5 needs to dissipate less watts than the F4, but half? Again, I am patient and I am sure someone will add some clarification to the FW amp needs, but I don't appreciate the ridicule when those doing so can't agree on the facts. Also, I can't figure out why a more efficient profile was not used? Why 0.8C/watt insteat of one of the 0.5C/watt which would require half the sink? If you call asking frugal questions about uncertain information and specs crazy? Then I am crazy and happy to be so.
Hi khundude,
If I may interject, I'm on the wiki as well. My understanding is that this GB wasn't created specifically for Pass Fx and previous comments in the thread indicated that this chassis would go well beyond the requirements for my F5 build.
For me personally, I didn't want to have to mess with the sourcing, drilling, tapping, etc. of components. I guess also it helps me tremendously the current value of the US $ to CAN $.
And of course, if Mr. Pass comes out with a design that requires more heatsinking, this chassis will probably do it for me. I originally went for the smaller size chassis because it would have been adequate, but I upgraded to the larger size so I can play with the F5 bias and/or scrap it for another in the Fx series.
Just another's perspective for this GB.
Cheers.
If I may interject, I'm on the wiki as well. My understanding is that this GB wasn't created specifically for Pass Fx and previous comments in the thread indicated that this chassis would go well beyond the requirements for my F5 build.
For me personally, I didn't want to have to mess with the sourcing, drilling, tapping, etc. of components. I guess also it helps me tremendously the current value of the US $ to CAN $.
And of course, if Mr. Pass comes out with a design that requires more heatsinking, this chassis will probably do it for me. I originally went for the smaller size chassis because it would have been adequate, but I upgraded to the larger size so I can play with the F5 bias and/or scrap it for another in the Fx series.
Just another's perspective for this GB.
Cheers.
I understand that a large chassis may be too much for an F5, but I was led to believe that I would need a large with 10" sinks for an F4. I am thinking that two 5"/ch should be enough. Also I would like to point out that a 10" sink will only be good for 50% increase over a 5", not double. That means one 5" does 67% and the 8" is good 90% of the heat dissipation of the 10". This shows why the Conrad is more efficient. Double the width equals double the cooling, double the height equals 40-50% more cooling.
From Conrad...
As a guide, thermal performance for heatsinks used with natural convection varies:
- in direct proportion to the width (double the width, double the heat dissipation);
- in proportion to the square root of the length (double the length, 40-50% increased heat dissipation).
As a result, width is thermally more effective than length. Comparing two heatsinks of similar thermal performance as shown below, the wider heatsink on the right (MF30-50) gives 45% more effective power dissipation per unit volume and weighs 28% less than the heatsink on the left (MF15-151.5). Hence, the inclusion of relatively broad heatsinks into the Conrad range.
For similar performance, the MF30-50 (right) occupies 30% less volume
and weighs 28% less than the MF15-151.5 (left).
This is why the lighter and wider R-Theta are rated around 0.5C/watt.
From Conrad...
As a guide, thermal performance for heatsinks used with natural convection varies:
- in direct proportion to the width (double the width, double the heat dissipation);
- in proportion to the square root of the length (double the length, 40-50% increased heat dissipation).
As a result, width is thermally more effective than length. Comparing two heatsinks of similar thermal performance as shown below, the wider heatsink on the right (MF30-50) gives 45% more effective power dissipation per unit volume and weighs 28% less than the heatsink on the left (MF15-151.5). Hence, the inclusion of relatively broad heatsinks into the Conrad range.
For similar performance, the MF30-50 (right) occupies 30% less volume
and weighs 28% less than the MF15-151.5 (left).
This is why the lighter and wider R-Theta are rated around 0.5C/watt.
- Status
- Not open for further replies.