Aleph 4 - wonderful sound but too hot

Alois

Member
2002-06-21 4:04 pm
Zurich
Hi everybody!


Introducing myself
I am new here as an active member, but profited a lot from this nice forum! Well I was inspired by this forum and this kept me busy for the last few month. I am more or less finishing my Aleph 4 based on the research done by Mark Finnis and others (thank you!). My major problem seems to be the heat dissipation of this amp. When I turn it on it reaches 75C within 15 minutes and the thermal protection switches turn the amp off again. However the sound within this 15 minutes of possible listening seems to come from another planet! It is sooooooooo gorgeous!

I have decided to go with the X6 configuration and installed a total of 24 matched IRF240 on two heat sinks each with the following specifications: 20cm x 40cm (8 inches x 16 inches) with weight of 4 kilos or 9 pounds. My judging out of the discussions and threats was that this should be enough, but obviously its not.

The rail voltage is @ +- 46 Volts and the Voltage drop over the source resistors is around the 0.6 V indicated on the schematics. I was not able to perform other measurements. Before I start redesigning the heat sinks I’m trying to find other solutions.

My Questions:

1. How can I measure with a multimeter the bias current of the constant current source which should be around 3A for the Aleph 4?

2. Is there a possibility to reduce the heat dissipation through reduction of the bias current and to keep the sonic characters of the amplifier the same. Would I have to increase the source resistors to 2ohms or another Value?. As a Trade off I would accept around 50W of loss in power.

3. Looking at R29 on the schematic showing a value of 619ohms it is written in red “523 on right”. What does that mean? Is 619 intended for the X3 Configuration where as I would have to change R29 to 523 as X6 and the 1.5 of the source resistors are written in red as well.

4. Has anybody made good experiences with the use of fans?

5. Are there other ways to go?


Any advice is very much appreciated! Thank you!

Regards, Alois
 

cp642

Member
2001-08-17 6:00 pm
Alois,
You'll need more heatsink, your 8X16 inch of heatsink is insufficient to dissipate the 200W or more of heat per ch.

Go for 2 pieces of those per ch. ( 6 mosfets per heatsink ) and space the output devices out evenly....or consider the use of a fan

Since the music is out of the planet as it seems, you might not want to lower the bias.;)
 
I have decided to go with the X6 configuration and installed a total of 24 matched IRF240 on two heat sinks each with the following specifications: 20cm x 40cm (8 inches x 16 inches) with weight of 4 kilos or 9 pounds. My judging out of the discussions and threats was that this should be enough, but obviously its not.

I have 4 SK435'S from fischer. Which are about 21cm x 20cm x 8cm a piece. Two of them together are one channel. So I have about 42cm x 20cm x 8cm of them a channel, this is 'just' enough. How deep are your fins? How thick is your base plate to which the fins are connected? My baseplate is about 18mm and the fins are about 60mm .

I idle on about 60C ... which is about 35C above ambient. I hope the rest of the case (10mm front, 4mm back, 4mm bottom, 4mm top) are enough to lower it to about 30C above ambient.

Edwin
 
1) 3A sounds a little high for an Aleph 4 (more like an Aleph 2); should be more like 2.5A. You've got two ways to measure the current draw:
--Measure the current across each MOSFET's Source resistor and add them up
--Insert a small resistance, say .1 ohm, in the rail and measure the voltage drop across the resistor. Ohm's Law will tell you the current.
2) Yes, you can decrease the heat dissipation by decreasing the bias current, but it will effect the sound of the amp. This is one of those 'there ain't no free lunch' kind of things. Bummer, but there it is. I've played with the bias current of my Aleph 2s, and can attest that it gets more mid-fi sounding as you reduce the bias.
3) The value of R29 will very likely need to be changed anyway. There seems to be a growing consensus that it needs fine-tuning in DIY versions of the amp. Why? My experience indicates that it's thermally sensitive. Nelson knew what his thermal environment was, but out here in the wilds we get into catch-as-catch-can heatsinks and the thermal environment changes from the production units.
4) Fans are an excellent option if you can live with the noise.
5) I water-cooled my Alephs. It's extremely efficient, and absolutely silent, but it's not an option that will be practical for everyone.

Grey
 
Hello Alois,

I´ve build my Aleph5 with 4 x 0.65 °C/W heatsinks and two papst 8412ngl fans. These produce 12dBA at 12V wich is a bit too loud but if you run them at 8 or 9 volts you won´t be able to hear them at about 2m distance. I get about 60°C at 25°C ambient so 35° for 260watts. This is about 0.13°C/W for the sinks + 2 fans.

william
 

Alois

Member
2002-06-21 4:04 pm
Zurich
Thank you all for your input!:)

Below please find my reply/comments:

Cp642:
I don’t like to use a fan and will probably double my heat sink capacity, hoping that this will be sufficient. That’s the price I have to pay for underestimating class A operation.

Edwin Dorre:
My fins are only 3cm deep and the base plate is 10mm, the rest of the case is 2mm only. How long does it take until you reach the mentioned 60C?

Grollins:
1) Thank you for the lesson. I followed your advice and measured 2.40 A for the left and 2.41 A for the right channel. Is there a problem since they are below your mentioned 2.5A? Does it matter if left and right channel is not exactly the same?

3) What do you mean when you say the value of R29 will very likely need to be changed anyway. I understand that it is thermally sensitive. To witch value should I change. Or do you suggest a certain procedure to find the best value?

4) and 5) I don’t like the idea of using a fan and water cooling seems not to be manageable by me. I think that I will have to double the heat sink capacity, hoping that it will be sufficient at the end.

How can I measure the DC off-set mentioned in different other threats?

I mentioned that the voltage drop over the source resistors are around 0.6V. I have remeasured and had a range from 5.5 to 5.85 volts. Should I be worried and look for other source resistors of better quality. I bought cheap ones and did not matching them.

Wuffwaff:
I will double my heat sink capacity. However if I will not end up with 60C or I will consider your solution. Did you measure the impact of your 2 fans?

Thank you all very much for your help or your reference to any threads dealing with my beginner questions.

Alois
 
Edwin Dorre:
My fins are only 3cm deep and the base plate is 10mm, the rest of the case is 2mm only. How long does it take until you reach the mentioned 60C?

I mentioned that the voltage drop over the source resistors are around 0.6V. I have remeasured and had a range from 5.5 to 5.85 volts. Should I be worried and look for other source resistors of better quality. I bought cheap ones and did not matching them.


Hello,

It takes about 30 to 45 minutes to reach this from cold.

I also have about 5.65 to 5.87 or something like this difference, with 2% 4W 1.5 ohm, and 1% 1/8W 221 ohm. Matched mosfets (0.01) and not matched resistors.

Edwin
 
Hello Alois,

without the two fans the temp rises over 70° (didn´t wait long enough to see where it stops). The diff between 12v and 8v operation is about 4°C.

The 0,6V is only an indication and varies with the bias current. I use MCP71 resistors (10%) and the voltage drop also varies between the resistors.

The effect of R29 can only be measured and adjusted for a current gain of 50%.

william