Too hot or not too hot, that's the question

[destroyer X]

What is thermal, continuous, stabilized heatsink temperature on the chip amplifiers from National Semiconductor series?... the small ones, alike LM4701T.

Do they use to operate hot?

I perceive temperature increasing.... i felt some fear and switched the unit off.... i have not wait the automatic temperature protection to enter (i felt it could be melted down together the Chip).

The amplifiers i have assembled (discrete), use to operate around 45 degrées celsius into the heatsinks.... more than that is not very common to AB class amplifiers that makes use of good sized heatsinks.

Datasheet talks that protection enters into more than 100 degrées celsius... this seems to me this is a "possible temperature".... seems also the unit may operate hotter than usual.

I am using a 50 watts heatsink.... able to dissipate more than that.... but guaranteed to 50 watts, and i felt the unit much more hot than the normal temperature i use to perceive into electronics....much more than the temperature i use to call as "human hot fever", and this grade of temperature is very common into AB class amplifiers (with the use of good sized heatsinks).

Class A is another story.... those use to work alike heaters...home ambience warmers.

Can you please folks, inform me if "chips works hot" or not?

If possible, inform in Celsius, will be more understandable and easier to me.

regards,

Carlos

[AndrewT]

Hi,
there is no need to operate them hot.
You should be able to hold your finger on the chip comfortably while operating.
If the chipamp has gone into oscillation then it will run much hotter.
If the heatsink is too small it will run hotter than necessary and this will reduce performance.
Can you check the quiescent current flowing in the two supply rails when the input is shorted and the output has only the Zobel connected?

[Dxvideo]

I made more than 10 chipamps and about 5 discrete and semi-discrete solid states..
As my observations; chipamps run warmer than the discrete ones. I think its because of the smaller thermal contact surface.
For example,
I have a LM4780 GC working with +/-30v and drives a 6R speaker pair.. It works at 70-75°C with a 1°C/W heatsink while,
The LM4702 + Lateral MOSFETs working with +/-55v drives the same speaker set It has 115mA per channel BIAS and its surface temperature is 60° with 0.5°C/W heatsinks..
So this is a comparison;
Total 120W output power and 75°C temperature with 1°C/W HS into 6R vs.
Total 350W output power and 60°C temperature with 0.5°C/W HS into 6R..
If a temperature corelation would be calculated with the same output power levels then that would means LM4702 is more efficient than LM4780..

PS: These comparisons made with the same listening level, not the maximum power level!

[Dxvideo]

Ok...
Never mind! This is bulls hit.
If they were all made in the same power levels then LM4780 works cooler than the LM4702.

[destroyer X]

Those were my suspections.... delay to enter the "stable point of operation"... the point were the internal VBE multiplier will sense and reduce to the correct ammount to operate hot.

Hotter than discrete amplifiers because of the case size.....fine

I loved that "bull hit"...good idea.... using some clever ways and elegance, some class, we can use the superlatives we need to give the needed strength to our affirmatives.

thank you,

regards,

Carlos

[Dxvideo]

using some clever ways and elegance, some class, we can use the superlatives we need to give the needed strength to our affirmatives.

Sounds like from a philosopy book! The meaning of life etc..

I dont understand that, men.. What do you mean? I am a poor Turkish guy, not from London College..

[destroyer X]

downtown.... were you have hundreds of Turkish folks and descendents... they are considered specialy clever.... they have shops, including one that sells LM4701T (Grrrrrrr.)

Interesting, one side of the street is Turkish, the other street side is Jewish...they use to play chess in a very friendly way.... the problems you see around the world between those folks does not exists here.

Turkish are famous into the Commerce (business, shops, fabrics, leather and plastics) in Brazil.

You are clever too man!

The amplifier is oscilating.... that crazy thing.

Carlos

[pacificblue]

Originally posted by destroyer X
Datasheet talks that protection enters into more than 100 degrées celsius... this seems to me this is a "possible temperature".... seems also the unit may operate hotter than usual.

The forum is full of posts around AN-1192 and the BPA-200 configuration described in it. Pages 3 and 4 explain thermal considerations about the LM3886, a typical representative of the Overture series.

In short, the maximum power dissipation capability without fan for the isolated package is 30 W, for the non-isolated package 40 W. With a fan that can be increased to 45 W for the isolated package and 60 W for the non-isolated package. Bigger heatsinks have no advantage, because the IC cannot make use of the additional cooling capacity.

With the above values the IC is supposed to run at the highest temperature, where SPiKe protection does not yet kick in. That would be well below 150 °C and well above a temperature that is nice to touch.

Lower temperatures give more headroom and increase the IC's lifetime, but are not necessarily better for the amplifier's sonic performance. Many people swear that their amplifiers sound better when they are warm than when they are cold. Some people even go as far as to never switch their amplifiers off, so they never have to cope with the (for their ears) bad sound of a cold amplifier.

[danielwritesbac]

I'm with AndrewT on this one.
So one more vote for: "There's no need to operate them hot."

If its doing a lot of unnecessary work at the detriment to audio, the opamp runs hotter--just like a car with fifteen unnecessary anvils in the trunk.

Reverb, equalizer, and other concert prosound standards may be good reasons to run hotter.

I'd like to hear of other good reasons to run hotter?

Its because, if there's not a good reason, then why not remove the anvils (noises)?