I don't know whether it would qualify as a "solution," but it's something I thought about trying. Still haven't gotten around to doing it, as I'm not set up to machine metal, just wood. Aluminum, copper, silver, gold, whatever you've got sitting around going to waste.
To date, all my Aleph-X variants have either had free-standing MOSFETs or back-to-back. Yes, the free-standing ones tend to drift. The problem can be reduced if you mount the MOSFETs against the circuit board.
I have at least two more variations on the Aleph-X theme that I intend to try, maybe three. It'll all depend on time and money.
(Would you believe that the heat pump problem is still dragging on? I've now gone to the BBB, and intend to start in with the Department of Consumer Affairs if someone doesn't say something pretty astounding by next Monday. I would advise against going with Trane heat pumps...)
Grey
To date, all my Aleph-X variants have either had free-standing MOSFETs or back-to-back. Yes, the free-standing ones tend to drift. The problem can be reduced if you mount the MOSFETs against the circuit board.
I have at least two more variations on the Aleph-X theme that I intend to try, maybe three. It'll all depend on time and money.
(Would you believe that the heat pump problem is still dragging on? I've now gone to the BBB, and intend to start in with the Department of Consumer Affairs if someone doesn't say something pretty astounding by next Monday. I would advise against going with Trane heat pumps...)
Grey
Thanks for the reply Mr Rollins
My guess is that you want to avoid a heatsink with cooling fins.....I would imagine a simple 4 or 6mm alloy shim more less the size of transistor would suffice...................
When ready I will try this and report back!!
Meanwhile I will check out other peoples findings ... cheers
My guess is that you want to avoid a heatsink with cooling fins.....I would imagine a simple 4 or 6mm alloy shim more less the size of transistor would suffice...................
When ready I will try this and report back!!
Meanwhile I will check out other peoples findings ... cheers
You could conceivably get to the point where you'd need a heatsink if you raise the rails enough. Nelson likes to run things hot; he probably doesn't worry until the MOSFETs glow cherry red. I tend to start thinking in terms of heatsinks when I see a TO-220 package at .5 to 1W Pd. It's just two views of the same thing. You can choose either endpoint, or somewhere in between.
Given that the rails for an Aleph-X are about half that of an equivalent Aleph for the same power, you're not likely to reach the point where the MOSFETs get really warm unless you're building a pretty large amp.
Grey
Given that the rails for an Aleph-X are about half that of an equivalent Aleph for the same power, you're not likely to reach the point where the MOSFETs get really warm unless you're building a pretty large amp.
Grey
My proposed amp design!!!
Hi....hope you can take a look at this......my amp will be running with 8a bias and 23v rails and 16 fets(current proposal) this will push out nearly 400w of heat!!!!
I have done some calcs based on Conrad Heatsink MF35-151
this is 350x151x40mm rated at 0.21C/W..........each side of amp will need to dissapate 200w........I am proposing to increase cooling by integrating the heatsink into the prerimeter casework increasing emmitance area....by at least 30%. The idea is to make the amp thermally stable at around say 65degC (total asuming 25deg room temp)......this is to ensure the thing does not go splat. To make matters friendly for all the components...I am proposing very mild air cooling using 4 high quality brush less 125mm fans running at around 5v............this should bring temps down to arround 45-50deg....please take a carefull look at the design.....here!!!! I would like as many people to chip in and advise and crit this compact design (wife friendly!)
nb here are my calcs............
R(80degC)=0.21C/W
Temp orrection factor for 25deg K(25degC)=1.42
R25(degC)=1.42x0.21=.2982C/W
Power Dissapation @ 40degC rise above ambient P(40degC)=40/.2982=134.13Watts
Power Dissapation to allow for 200w = 200x0.2982 = 59.64deg above ambient!
Therefore bare sink temp = 59.64+25 = 84.64 Say 85 deg.
Mild air assistance will give 0.17C/W (very conservative) gives 0.2414
0.2414x200 = 48.28 = final sink temp of 73.28
However my design increases thermal radiating area by at least 30% (conservative)
Therefore 30% off 0.17C/W = 0.119
0.119 x200 = 23.8 above ambient giving a final figure of 48.8deg (bingo)
Hi....hope you can take a look at this......my amp will be running with 8a bias and 23v rails and 16 fets(current proposal) this will push out nearly 400w of heat!!!!
I have done some calcs based on Conrad Heatsink MF35-151
this is 350x151x40mm rated at 0.21C/W..........each side of amp will need to dissapate 200w........I am proposing to increase cooling by integrating the heatsink into the prerimeter casework increasing emmitance area....by at least 30%. The idea is to make the amp thermally stable at around say 65degC (total asuming 25deg room temp)......this is to ensure the thing does not go splat. To make matters friendly for all the components...I am proposing very mild air cooling using 4 high quality brush less 125mm fans running at around 5v............this should bring temps down to arround 45-50deg....please take a carefull look at the design.....here!!!! I would like as many people to chip in and advise and crit this compact design (wife friendly!)
nb here are my calcs............
R(80degC)=0.21C/W
Temp orrection factor for 25deg K(25degC)=1.42
R25(degC)=1.42x0.21=.2982C/W
Power Dissapation @ 40degC rise above ambient P(40degC)=40/.2982=134.13Watts
Power Dissapation to allow for 200w = 200x0.2982 = 59.64deg above ambient!
Therefore bare sink temp = 59.64+25 = 84.64 Say 85 deg.
Mild air assistance will give 0.17C/W (very conservative) gives 0.2414
0.2414x200 = 48.28 = final sink temp of 73.28
However my design increases thermal radiating area by at least 30% (conservative)
Therefore 30% off 0.17C/W = 0.119
0.119 x200 = 23.8 above ambient giving a final figure of 48.8deg (bingo)
I honestly don't understand how you calculated that your case will increase the dissipative area of your heat sink by 30%. I will address that later. I had a look at your specs and this is what I come up with. Each heat sink will dissipate one half of the total dissipation of 352 watts as taken from the AXE-1 spreadsheet for your proposed 23 volt rail and 8 amp bias. if you divide that in two each heatsink will need to dissipate 176 watts. With your proposed heatsinks that have a thermal resistance of .21C/watt you can expect a rise above ambient of 36.96C. If I am correct your heatsinks will run at about 57C or for those here in the States 123F. It appears from your drawing as if you are going to sandwich your chassis side panels between the heat sinks and your transistors in hope that you get more surface area to dissapate heat. If this is your intent (I can't really tell from your drawing), I think you will have some problems in that you will rob your transistors of good thermal contact with the component that will dissipate the most heat, the heatsinks. Your chassis will be dissipating most of the heat. I do however like your idea of using copper as a heat spreader. What I would do, because you plan on using some fans anyway, is to get the transistors on a piece of copper plate much larger than you propose to spread the heat over a much wider area and bolt that directly to the heatsink with some thermal goo between the copper and heatsink. With the fans and that setup, I think you will doing just fine. BTW, I really like your chassis design. It is very professional and extremely well thought out. Just my suggestion, I hope it helps.
My amp changes!!!!
I have been looking at the heat dissapation figures for my amp....
I am going to change the design................
I going to double the heatsinks..............this will keep things to a max of around 55-60degc ........... without fan assistance.
I suggest that other people in the forum.......check out their dissapation ... taking into account the Trafo/rectifier and any resistors....if you have a CRC setup.....
I have been informed that running computer fans at low speed can be very unreliable............is this the case......or have others in this forum found otherwise?????
Because I am changing the design.......I am ordering 4 more Conrad Heatsinks in about a months time...........is anybody else interested in extra heatsinking from Conrad???
I am also going to be using copper bus bars to mount the fets directly.........I have been informed that this assists in reducing the all critical fet junction temp......again I can let you know who my supplier is.....they work out at approx £10 per 250x50x12mm
bus bar....
I have been looking at the heat dissapation figures for my amp....
I am going to change the design................
I going to double the heatsinks..............this will keep things to a max of around 55-60degc ........... without fan assistance.
I suggest that other people in the forum.......check out their dissapation ... taking into account the Trafo/rectifier and any resistors....if you have a CRC setup.....
I have been informed that running computer fans at low speed can be very unreliable............is this the case......or have others in this forum found otherwise?????
Because I am changing the design.......I am ordering 4 more Conrad Heatsinks in about a months time...........is anybody else interested in extra heatsinking from Conrad???
I am also going to be using copper bus bars to mount the fets directly.........I have been informed that this assists in reducing the all critical fet junction temp......again I can let you know who my supplier is.....they work out at approx £10 per 250x50x12mm
bus bar....
Fans for Aleph
Hi I was recommended to use panaflo fans which are very quiet.....I have attached a link....hope it works.....
http://homepage.ntlworld.com/dorothy.bradbury/panaflo.htm
Hi I was recommended to use panaflo fans which are very quiet.....I have attached a link....hope it works.....
http://homepage.ntlworld.com/dorothy.bradbury/panaflo.htm
high powered Aleph X
This maybe a little redundant but has any one built the Aleph X high power version, oh 130 watt or more. I am looking to start my project but have not been keeping up with the threads. It seems that some values with the current sources should be changed and of course the biasing circuit resistors. So does any one have an high powered X that is up and working and info to share.
Thanks
Jim
This maybe a little redundant but has any one built the Aleph X high power version, oh 130 watt or more. I am looking to start my project but have not been keeping up with the threads. It seems that some values with the current sources should be changed and of course the biasing circuit resistors. So does any one have an high powered X that is up and working and info to share.
Thanks
Jim
Fans
Yes you are right in theory a large fan.........at very very low speed could be used..........ie a house fan ..........but they are BIG.....so these panaflo fans are really quiet......15 dB is very quiet.......and when slowed down they would probably be as low as 10 dB............now thats quiet........our pc's have these fans fitted and they are almost silent...................
My plan is to use 4 of these fans.........remember if you are relying on a fan and it is critical to your cooing .... failure could be a disaster...........so a number of small low speed fans will provide safety.............. + they are quiet cheep and excellent quality....using fluid bearings.............
Yes you are right in theory a large fan.........at very very low speed could be used..........ie a house fan ..........but they are BIG.....so these panaflo fans are really quiet......15 dB is very quiet.......and when slowed down they would probably be as low as 10 dB............now thats quiet........our pc's have these fans fitted and they are almost silent...................
My plan is to use 4 of these fans.........remember if you are relying on a fan and it is critical to your cooing .... failure could be a disaster...........so a number of small low speed fans will provide safety.............. + they are quiet cheep and excellent quality....using fluid bearings.............
Over temerature alarm on fan failure?
Suggest that you use a thermal switch mounted to one (or both) of the heatsinks - thes come in TO220 packages and are easy to mount. The switches could be used to activate a sounder is the heatsinks reach the trigger threshold of the switch (say 75 degrees or thereabouts) - might save you from frying your amp if one of the fans fails and will only cost a fiver to add.
Suggest that you use a thermal switch mounted to one (or both) of the heatsinks - thes come in TO220 packages and are easy to mount. The switches could be used to activate a sounder is the heatsinks reach the trigger threshold of the switch (say 75 degrees or thereabouts) - might save you from frying your amp if one of the fans fails and will only cost a fiver to add.
Faaaaan!
I like fans! But what about injected electrical noise into the signal path. Moreover, is it wise to use DC fans or AC fans (electrical) noise wise? I am asking this as I designed something similar for a year now but only since I saw your design I gain the remotivation I needed to get it going again. Nice idea!
I like fans! But what about injected electrical noise into the signal path. Moreover, is it wise to use DC fans or AC fans (electrical) noise wise? I am asking this as I designed something similar for a year now but only since I saw your design I gain the remotivation I needed to get it going again. Nice idea!
Fans....some more thoughts
Well I thought about this.........I don't like AC near signal paths.
I also did not want to draw power from the 'Audio power supply'....I did'nt want the risk of injecting noise etc..........into system. I was going to buy a small 12v .5amp regulated powersupply...on its own circuitboard (chip type)...and use a very small transformer to run it...........then use a pot...to drop the fan speed........ so I would use dc fans....
I think its important to have a good smooth rectified mini supply for the fans...........other wise they will pack up.
Another option is to use the above but via a nicad........and use a timer circuit so the fans continue running when the amp is switched off...............say 10minute overun.........this would be good for your valuable components.
Well I thought about this.........I don't like AC near signal paths.
I also did not want to draw power from the 'Audio power supply'....I did'nt want the risk of injecting noise etc..........into system. I was going to buy a small 12v .5amp regulated powersupply...on its own circuitboard (chip type)...and use a very small transformer to run it...........then use a pot...to drop the fan speed........ so I would use dc fans....
I think its important to have a good smooth rectified mini supply for the fans...........other wise they will pack up.
Another option is to use the above but via a nicad........and use a timer circuit so the fans continue running when the amp is switched off...............say 10minute overun.........this would be good for your valuable components.
Thermal switches
I bought some a long time back from RS Compnents to run a fan for a JLH '96 amplifier that cooked sometimes (heatsink too small).
May i make the following suggestion? - Use a "plug-in-brick" DC 12v power supply and have it power the fans in each amplifier (just use a power jack connector in the back of each amplifier for the Fan DC power input. Then use 2 thermal switches rated at different temperatures - say 65 degrees and 75 degrees. The fan supply comes from the FAN DC socket to the fan via a chain of diodes (say 10 diodes - each diode will drop 0.6 to 0.7v off or use a variable resistor, whatever). the 65 degree fan switch is connected across the diodes so as to short them out if the heatsink gets to 65 degrees - the fans will go at full tilt to cool the heatsink if the switch trips. The 75 degree switch is used to power a sounder (also powered off the FAN DC socket).
The fans will also run-on after the amp is switched off via the seperate power supply
I bought some a long time back from RS Compnents to run a fan for a JLH '96 amplifier that cooked sometimes (heatsink too small).
May i make the following suggestion? - Use a "plug-in-brick" DC 12v power supply and have it power the fans in each amplifier (just use a power jack connector in the back of each amplifier for the Fan DC power input. Then use 2 thermal switches rated at different temperatures - say 65 degrees and 75 degrees. The fan supply comes from the FAN DC socket to the fan via a chain of diodes (say 10 diodes - each diode will drop 0.6 to 0.7v off or use a variable resistor, whatever). the 65 degree fan switch is connected across the diodes so as to short them out if the heatsink gets to 65 degrees - the fans will go at full tilt to cool the heatsink if the switch trips. The 75 degree switch is used to power a sounder (also powered off the FAN DC socket).
The fans will also run-on after the amp is switched off via the seperate power supply
Hi, it looks like it should still work (as long as you add the AC at one end of each run of capacitors and resistors and take the opposite end for DC you should be fine), although, I'm not too sure why you needed to swap this round on the scheematic? If you are intending to do your physical layout in this way it should still be fine though, and should be intresting to see when you have things put together.
Supply
Thanks BP......I thought it looked OK......the reason I wanted this layout...was an attempt to make the internal wiring much neater.....I am basically having the + fets on one side of the amp and - fets on the other....I will post my Mk2 CAD layout shortly.......
I am also getting some quotes for some serious aluminium...I am proposing to use an 8mm thick aluminium chassis.....!!! (to assist in heat sinking)
I am also going to be ordering ..... (in approx 8 weeks) more conrad heat sinks....I will contact you then.......before I order
Thanks BP......I thought it looked OK......the reason I wanted this layout...was an attempt to make the internal wiring much neater.....I am basically having the + fets on one side of the amp and - fets on the other....I will post my Mk2 CAD layout shortly.......
I am also getting some quotes for some serious aluminium...I am proposing to use an 8mm thick aluminium chassis.....!!! (to assist in heat sinking)
I am also going to be ordering ..... (in approx 8 weeks) more conrad heat sinks....I will contact you then.......before I order
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.