No Tonza,
Not completed yet. I'll let you know!
Calculate the dissipation in the outputs, then add, in watts from LTSPice.
For the ALPHA its around 98W.
If the heatsink is 0.31C/W, the rise over ambient is 0.31x98=30.4C.
If the ambient is 22C, a normal room temp, the heatsink, IF the heat is spread well on the heatsink, will be 22+30.4=52.4W.
As the you turn up the level, more power goes to the speaker, reducing the power on the heatsink. For 5W listening level (LOUD) the heatsink will cool to 50.8W, not much, but indicates at high ambient temperature you should have LOUD music and a COLD icebox for your beer..........
Hugh
Not completed yet. I'll let you know!
Calculate the dissipation in the outputs, then add, in watts from LTSPice.
For the ALPHA its around 98W.
If the heatsink is 0.31C/W, the rise over ambient is 0.31x98=30.4C.
If the ambient is 22C, a normal room temp, the heatsink, IF the heat is spread well on the heatsink, will be 22+30.4=52.4W.
As the you turn up the level, more power goes to the speaker, reducing the power on the heatsink. For 5W listening level (LOUD) the heatsink will cool to 50.8W, not much, but indicates at high ambient temperature you should have LOUD music and a COLD icebox for your beer..........
Hugh
I think Hugh means C not W for temperature calcs above.
For lower bias current Alpha you can also calculate heat output with P=IV. Bias if 1.35amps and rails are +/-25v so 50v total so P=1.35amp x 50v=68W.
68W x 0.31C/W heatsink plus 22C ambient is 43C final heatsink temp.
Thats a 4U Dissipante 300mm deep heatsink from DIYA store. I measure about 42C in mine.
For lower bias current Alpha you can also calculate heat output with P=IV. Bias if 1.35amps and rails are +/-25v so 50v total so P=1.35amp x 50v=68W.
68W x 0.31C/W heatsink plus 22C ambient is 43C final heatsink temp.
Thats a 4U Dissipante 300mm deep heatsink from DIYA store. I measure about 42C in mine.
Heat transfer for natural convection fins, to first approximation, is proportional to characteristic length^3, not mass. Look at Rayleigh number for natural convection near vertical plate. The mass only affects the transient capability to absorb heat before it starts to transition to fin limiting heat flux.
Sure , very usefull for diy .
With At present, numerous practical applications involve
high-Rayleigh-number turbulent natural-convection
flows. Such flows may be realized both under conditions
of heat input through the wall and as a result of
internal (bulk) heat release. The numerical simulation
of high-Rayleigh-number natural-convection heat
transfer is a fairly complex problem, which is primarily
due to the formation of very thin wall layers. Therefore,
the use of methods of direct numerical simulation (rapidly
developing in recent years) based on the solution of
unsteady-state three-dimensional Navier-Stokes and
energy equations involves the need to employ very
detailed nets for resolving small-scale vortex structures
and, consequently, is very time-consuming in the case
of even the fastest computers. No doubt, the direct
numerical simulation appears to be of great use from
the standpoint of testing models being developed; however,
it will hardly hold any real promise in the foreseeable
future as regards the solution of problems of practical
interest. Another approach to simulation of turbulent
natural-convection flows is based on the use of
traditional models of turbulence (for example, the k-e
model) developed for forced convection. Such models,
however, may lead to major errors, because they fail to
include the buoyancy forces (do not describe the phenomena
of thermogravitational generation of turbulence,
stable and unstable stratification, negative turbulent
viscosity, etc.). For correct utilization of conventional
models of turbulence, they need to be modified to
include the effect of the buoyancy forces on the turbulent
transfer coefficients.
Thanks xrk ,
Experimental study of high-Rayleigh-number convection in a horizontal cavity with different end temperatures.
The paper summarizes results from an experimental study of buoyancy-induced motion and heat transfer in a horizontal rectangular cavity with the two vertical ends at different temperatures and the long horizontal walls adiabatic. The cavity height/length ratio is A = 0·0625. The high-Rayleigh-number range reported on in this paper, 2 × 108 < Ra < 2 × 109, has not been studied before. It is shown that, contrary to lower-Rayleigh-number behaviour known previously, the core flow structure is non-parallel and is dominated by horizontal intrusions flowing along each of the two insulated horizontal walls of the enclosure. The fluid embraced by the two horizontal jets is practically stagnant and thermally stratified. Flow visualization experiments suggest that adjacent to the two horizontal jets two secondary flat cells are formed by the baroclinic pressure field in an analogous way to what is observed in intrusions in a stratified fluid. Nusselt-number-Rayleigh-number results for the overall end-to-end heat transfer in the horizontal direction are reported and compared with previous experimental and theoretical results available for lower Rayleigh numbers. It is shown also that the transition from a parallel core structure to one dominated by intrusion layers is governed by the parameter Ra½A, with Ra¼A < 1 as necessary condition for a parallel core flow.
I have to read more
No wonder the temp was a little higher on diyaudio heatsink very cost effective vs FISCHER ELEKTRONIK and aluprofile 20mm x 40mm
With At present, numerous practical applications involve
high-Rayleigh-number turbulent natural-convection
flows. Such flows may be realized both under conditions
of heat input through the wall and as a result of
internal (bulk) heat release. The numerical simulation
of high-Rayleigh-number natural-convection heat
transfer is a fairly complex problem, which is primarily
due to the formation of very thin wall layers. Therefore,
the use of methods of direct numerical simulation (rapidly
developing in recent years) based on the solution of
unsteady-state three-dimensional Navier-Stokes and
energy equations involves the need to employ very
detailed nets for resolving small-scale vortex structures
and, consequently, is very time-consuming in the case
of even the fastest computers. No doubt, the direct
numerical simulation appears to be of great use from
the standpoint of testing models being developed; however,
it will hardly hold any real promise in the foreseeable
future as regards the solution of problems of practical
interest. Another approach to simulation of turbulent
natural-convection flows is based on the use of
traditional models of turbulence (for example, the k-e
model) developed for forced convection. Such models,
however, may lead to major errors, because they fail to
include the buoyancy forces (do not describe the phenomena
of thermogravitational generation of turbulence,
stable and unstable stratification, negative turbulent
viscosity, etc.). For correct utilization of conventional
models of turbulence, they need to be modified to
include the effect of the buoyancy forces on the turbulent
transfer coefficients.
Thanks xrk ,
Experimental study of high-Rayleigh-number convection in a horizontal cavity with different end temperatures.
The paper summarizes results from an experimental study of buoyancy-induced motion and heat transfer in a horizontal rectangular cavity with the two vertical ends at different temperatures and the long horizontal walls adiabatic. The cavity height/length ratio is A = 0·0625. The high-Rayleigh-number range reported on in this paper, 2 × 108 < Ra < 2 × 109, has not been studied before. It is shown that, contrary to lower-Rayleigh-number behaviour known previously, the core flow structure is non-parallel and is dominated by horizontal intrusions flowing along each of the two insulated horizontal walls of the enclosure. The fluid embraced by the two horizontal jets is practically stagnant and thermally stratified. Flow visualization experiments suggest that adjacent to the two horizontal jets two secondary flat cells are formed by the baroclinic pressure field in an analogous way to what is observed in intrusions in a stratified fluid. Nusselt-number-Rayleigh-number results for the overall end-to-end heat transfer in the horizontal direction are reported and compared with previous experimental and theoretical results available for lower Rayleigh numbers. It is shown also that the transition from a parallel core structure to one dominated by intrusion layers is governed by the parameter Ra½A, with Ra¼A < 1 as necessary condition for a parallel core flow.
I have to read more
No wonder the temp was a little higher on diyaudio heatsink very cost effective vs FISCHER ELEKTRONIK and aluprofile 20mm x 40mm
For all practical purposes, natural convection over the small distances of a 4U tall (7in) rise heatsink is all laminar flow. That’s why natural convection is so inefficient vs forced convection (with a fan). The turbulence and thinner boundary layers of forced convection increase heat transfer by 100x. If you use a bright collimated light source to shine across the thermal gradients at the top of the fins and project onto a screen, the shadows will show a laminar flow. No CFD or DNS needed, the laminar nature allows tabulated engineering solutions to the cooling capability of the fins.
Practical tools
Let's keep it practical for diyer sake.
Power at given voltage and ampere . Bias for alpha or B.B
I put this link it's easy to read.
https://www.infineon.com/dgdl/smdpack.pdf?fileId=db3a304330f6860601311905ea1d4599
Or if your calculus if out of date. Apps on android for example electodroid .
I can highly recommed such app for practical purposes in electronics
1. It's in your language
2. Basics of electronics well covered.
3. Ease of use.
4. Plugins
Bye.
Let's keep it practical for diyer sake.
Power at given voltage and ampere . Bias for alpha or B.B
I put this link it's easy to read.
https://www.infineon.com/dgdl/smdpack.pdf?fileId=db3a304330f6860601311905ea1d4599
Or if your calculus if out of date. Apps on android for example electodroid .
I can highly recommed such app for practical purposes in electronics
1. It's in your language
2. Basics of electronics well covered.
3. Ease of use.
4. Plugins
Bye.
Both channels, fans, and speaker delay/protect board are up and running. Right channel DC offset is 5mV and the left is 8mV. Not super low but I could be perfectly happy with that. DC delay/protect is doing its job. No more cone movement at switch on. And...... It sounds fantastic!!! Playing some Oscar Peterson jazz piano through my makeshift WAW, Mark Audio Pluvia Sevens on 18"x18" 1/2" foam core open baffles with Visaton WS25E-8's for low frequency support. Signal is running through my new AKSA Lender Preamp to minidsp 2x4hd to Marantz 1060 for LF and ALPHA 20w for the HF. Having so much fun playing around with everything.
Thanks Vunce and Zman!
I saw that X! Exciting stuff and glad he got the Harsch XO working. I'm going to try my hand at that XO as well. As soon as I get my Visaton's into a u-frame. I just have them playing up into the air behind my Pluvia OB's right now and the coherence isn't very good at all.
I saw that X! Exciting stuff and glad he got the Harsch XO working. I'm going to try my hand at that XO as well. As soon as I get my Visaton's into a u-frame. I just have them playing up into the air behind my Pluvia OB's right now and the coherence isn't very good at all.
Last edited:
Nice work JW!
I did some testing with fans I have in hands. I think that I won't even use PWM-controlled fans on B.B. I'm going rather with 4 x of these: NF-A14 ULN and spin them with constant RPM.
With U.L.N.A - adaptor (I guess it's only a resistor) it's really, I mean really, quiet. 650rpm, 66,4m3/h and 0,42mm H2o and only 9,1db(A). If that's not enough cooling, you can rotate it 800rpm only with 11,9db(A) and get better performance. Should be enough, what do you think?
Theres also couple of orders on it's way; 20 x 15000uF caps and parts to populate pcb's. Casework for the first monoblock starts probably tomorrow. To make myself busy, there's also M2x pcb's coming but that is another story
Edit: Thanks wtl for tipping out those Kemet caps, I bought those also
I did some testing with fans I have in hands. I think that I won't even use PWM-controlled fans on B.B. I'm going rather with 4 x of these: NF-A14 ULN and spin them with constant RPM.
With U.L.N.A - adaptor (I guess it's only a resistor) it's really, I mean really, quiet. 650rpm, 66,4m3/h and 0,42mm H2o and only 9,1db(A). If that's not enough cooling, you can rotate it 800rpm only with 11,9db(A) and get better performance. Should be enough, what do you think?
Theres also couple of orders on it's way; 20 x 15000uF caps and parts to populate pcb's. Casework for the first monoblock starts probably tomorrow. To make myself busy, there's also M2x pcb's coming but that is another story
Edit: Thanks wtl for tipping out those Kemet caps, I bought those also
Last edited:
Question was; is it enough to cool B.B.?
There are many other factors that counts also: 6y warranty, reliable and quiet. It's hard to measure, but what I have read is that many other manufacturers give waaaay too optimistic specs what they really are.
So, if there is provenly more efficient and atleast as quiet or quieter (and that's really important part here) fans out there on 120-140mm class, please let me know now before I go and buy more of these..
There are many other factors that counts also: 6y warranty, reliable and quiet. It's hard to measure, but what I have read is that many other manufacturers give waaaay too optimistic specs what they really are.
So, if there is provenly more efficient and atleast as quiet or quieter (and that's really important part here) fans out there on 120-140mm class, please let me know now
before I go and buy more of these..
Juntuin, resistor based speed control works and can be very quiet on non PWM fans. Be careful of motor stall. The PWM can be slow and not stall. Also the controller senses temp and kicks up speed if something happens plus audible beep for fan stall.
But you should be fine. The use of CPU heat piped and fans is like 100x more efficient than passive radiator so almost never an issue of sufficient cooling if CPU cooler rates for 100w+ cpu’s.
But you should be fine. The use of CPU heat piped and fans is like 100x more efficient than passive radiator so almost never an issue of sufficient cooling if CPU cooler rates for 100w+ cpu’s.
It is Noctuas own piece of wire, that probs include a resistor inside. So I'm not worried about stall while using that
reason I wan't to avoid PWM is simply just keeping things and wirings as simple as possible. And big plus is that I can't even hear it running on steady speed until my ear is on the fan. I have ordered PWM-controllers though, just in case. They were so cheap also.