MUR820 or HFA08TB60?

[Konsti]

Hi Folks!

In my main Job I am doing some stuff with micro fine pitched BGA and sorts of microprocessor Stuff, so I have not much experience, when it comes to high current/high Voltage applications such as a Aleph-J clone :-)

Despite of that I participated in a group buy and building an Aleph-J clone and mine runs very fine and I am astonished experiencing the difference of good parts and excellent parts building into it, while it is a prototype in my rack and waiting for its final housing.

The issue is, for each lane of power supply there was build a bridge of four diodes in TO-220 case to rectify a 250VA/18V toroid into an RCR of 22mF and 0.2Ohm. This works well but by accident the MUR820 are getting bleeding hot (wo heatsink), around 108Grad Celcius case Temperature. One Member has build in HFA08TB60 and tells, his are not even getting any hotter than ambient. Is that possible? The latter Diodes have roughly doubled forward Voltage (1.4V) than the MURs so their energy wasted should be doubled. Is that possible?

This is my attention of the post (my lack of experience here): What others parameters should be concerned regarding the power dissipation in this place of application?

Did anybody compared the quality regarding the sound in the poweramplifier of one or the other of these parts? I read som good stuff about the HFA and I feel its not recommended to usr MUR820 in this place unless you really need their capabilities...
Is it complete mad voodoo to check the sound or result on sound of a part in this place?

Regards, Konsti

[jacco vermeulen]

Flugzeug-Bastler Konstantino,

why do all commercial Class A power amps have very sturdy (bridge) rectifiers ?

The dissipation of a rectifier diode is the sum of forward and reverse-recovery losses. (ref: Fred Dieckmann on fast diodes-101)
The level of the Forward (Instint.) Voltage depends on If, but dissipation will always be larger than the sum of If*Vf due to the losses during reverse recovery.
I fail to see how the dissipation of an MUR820 full bridge rectifier can be less than 3.6W-4W at a 2.5A level.

Thermal resistance of case to ambient Rca=Rja-Rjc.
For the MUR820 : Rca = 73 - 3 = 70C/W.
At 25C ambient => a diode-case temperature of 88-95 C.

Your A-J is not oscillating, means :
A - your MURs Zuck
B - your amps are biased higher than 2.5A.

(me use metal bricks like paps. A: i've also got a ton of them. B:me only has 55mm height inside the amp cases and neu reum for 8 rect. diode heatsinks)

[tinitus]

I thought that MUR820/860 were best suited fore preamps or amp input/driver stage
Personally I use IR diodes and those similar to your MUR I use fore input/driver stage...fore output I use more heavy ones

Each mono amp takes a total of 16 diodes

[jacco vermeulen]

Aerosmith shouts: Like THIS

[Konsti]

Originally posted by jacco vermeulen
[B]Flugzeug-Bastler Konstantino,

why do all commercial Class A power amps have very sturdy (bridge) rectifiers ?

This is my question, why? Okay, they are cheap and indeed not so bad.

The dissipation of a rectifier diode is the sum of forward and reverse-recovery losses. (ref: Fred Dieckmann on fast diodes-101)

And my second question is, how do I estimate these reverse recovery losses...

I fail to see how the dissipation of an MUR820 full bridge rectifier can be less than 3.6W-4W at a 2.5A level.

Do I understand you correct here, that power loss of more that 3W is totally obvious to you?

Thermal resistance of case to ambient Rca=Rja-Rjc.
For the MUR820 : Rca = 73 - 3 = 70C/W.
At 25C ambient => a diode-case temperature of 88-95 C.

This way the power loss is nominal around 1W if I understood correct. But I have the feeling it is more than this.
V_f is around 0.7 Volt and current is 2.25 average = 1.575W... Hmmm..
Why can HFAx with 1.4V V_f stay so much cooler if it is mainly forward voltage losses?

Your A-J is not oscillating, means :

I measured the voltage on A-C of one Diode and the pics of oscilloskope are here:
http://www.ludenkalle.de/Aleph-J/oszi/
So I suppose not, no...
Physical Construction is here:
http://www.ludenkalle.de/Aleph-J/Dioden/

The copper application are a new idea, this way the copper gets 60 degree celciaus and the centre of four diodes goes up to 70 degrees.

A - your MURs Zuck

Yes, but why? Why I am so surprised these geting so fringing hot?
Or is it obvious and normal due to 1W Heat dissipation?
Or is it more power loss due to reverse losses, how much if?

B - your amps are biased higher than 2.5A.

No, it sucks exactly 2.25A from each line ie. 2.5A from each 4-Pack of Diodes.

[jacco vermeulen]

Originally posted by Konsti
power loss of more that 3W is totally obvious to you?

Yep.
Why the IRF HFA's remain cool without proper heatsinks is beyond me.

"Sturdy rectifier bridge" means both 35A or higher 4-pin metal case a-la KBPC or a package of 4 single heatsinked heavy duty Fast/SR rectifier diodes as used in the biggy Pass amps.
It should be obvious that a Class A amplifier requires a thicker rect. bridge than an AB version.

You can measure it with a DSO, make an estimation with help of the datasheet, or calculate it by measuring the temperature.

(your PS-boards now look like little aeroplanes)

[Konsti]

Originally posted by jacco vermeulen
Why the IRF HFA's remain cool without proper heatsinks is beyond me.

This is beyond me also. Actually I wonder if this is really the case since this happened to a friend of a member in the forum so I am very sceptical about this fact.
If the HFAx gets at least the temperature the MUR820 are getting I am convinced to search for another solution, I will not try the HFAs and will trust to the calculation and physics (reading my writing the decision is amde for that already...).

(your PS-boards now look like little aeroplanes)

Yes, they do :-) How do you think, will the over all temperatur be significantly lower if the Copper is painted or not? Actually it is painted by me with acryl-transparent and I consider this as a bad idea but can not estimate how much potential of the copper is wasted.

[jacco vermeulen]

Considered moving the second row of 820s to the bottom of the amp case ?
That should give some space to use a somewhat better heatsink for the first 4 diodes.

Acrylic varnish should be comparable to sticking the heatsink strips in a plastic sandwich bag, >25% reduction wouldn't surprise me.

[Konsti]

Temperature is nearly the same. I disassembled the whole stuff and removed varnish with aceton. The temperature between the inner two TO220 Fins reaches nearly 74 degrees. There probably are measurement tolerances of course so lets stick between -0 to -5 degrees.
May be heat dissipation due to radiaton comes into the game, when the copper oxydes :-) (factor 0.07 to 0.7 as I read about this constant).

When reassembling the whole stuff, I understood the "Flugzeug-Bastler" joke, oh my!

I consider to turn all diodes around (also the elkos of course) and mount the inner row onto the case. But not today nor tomorrow...
I thought about that but I get ideas for this while you mention this...