@Andrew T : thank you, now I'm sure will not use regulator.
@Mark W : during test I, after regulator I use elco 1000 microF.
Today I have time to assemble :
After spend almost a day, because chassis so small and I have no good skill of drilling, finally ready to test . Front panel still not placed yet. Voltage DC +/28 volt , but a bit surprise heat sink fast become hot.
:
Test with music. I use old Denon, parametric equalizer. Did not change setting of equalizer. Volume about 1/4 when mandarin bass draw the current , voltage drop to 22 volt. Heat sink is getting hotter.
I play my favourtite Laura Fygi, with same volume , voltage drop to 24 volt when bass is produced.
Heat sink temp is not normal , I check temp after 30 minutes:
Compare with my old "mini" Cressendo with same player and equalizer setting, at high freq. I feel more metalic, good at low, middle less open.
Generally is good I can enjoy it. It still needs few hour to get the real taste. But heat sink temp is not acceptable.
@Mark W : during test I, after regulator I use elco 1000 microF.
Today I have time to assemble :
After spend almost a day, because chassis so small and I have no good skill of drilling, finally ready to test . Front panel still not placed yet. Voltage DC +/28 volt , but a bit surprise heat sink fast become hot.
:Test with music. I use old Denon, parametric equalizer. Did not change setting of equalizer. Volume about 1/4 when mandarin bass draw the current , voltage drop to 22 volt. Heat sink is getting hotter.
I play my favourtite Laura Fygi, with same volume , voltage drop to 24 volt when bass is produced.
Heat sink temp is not normal , I check temp after 30 minutes:
Compare with my old "mini" Cressendo with same player and equalizer setting, at high freq. I feel more metalic, good at low, middle less open.
Generally is good I can enjoy it. It still needs few hour to get the real taste. But heat sink temp is not acceptable.
I checked Power Amplifier alone, using multimeter Sanwa 960 :
Volume control min and max gave same result :
DC (scale 0.1 Volt) : L and R channel comes out negative (-) a few mVolt, needle of meter move very very small, only as far as needle body.
AC (scale 10 Volt): L and R channel no detection .
The supply rail +/- 27 volt .
Before connect with Amplifier supply rail +/- 32 volt.
During standby condition the chips drinks a lot of current.
Volume control min and max gave same result :
DC (scale 0.1 Volt) : L and R channel comes out negative (-) a few mVolt, needle of meter move very very small, only as far as needle body.
AC (scale 10 Volt): L and R channel no detection .
The supply rail +/- 27 volt .
Before connect with Amplifier supply rail +/- 32 volt.
During standby condition the chips drinks a lot of current.
That's amazing!
For an Overture chip to do that is usually either overload like 4 ohm speaker or ground loop like monobloc centric PG+ PG- nonsense misapplied to a stereo amplifier.
Sure, the poor things have sonic signatures like housecats climbing chalkboards except not as good as that, but severe overcurrent is not something they do routinely, so I suspect a layout error. If it sounded really shouty AND very clear, I'd say that's normal for it; however, the overcurrent thing is not normal, so it probably needs fixed.
I think your heat sink size is enough. May be oscillate. Please check using oscilloscope.
For normal use, the heat sink temperature rise about 60C is normal. You can touch the heat sink about 30s without burning your finger.
No.
A rise of 60C degrees is far too much.
A rise from 30°C to 50°C is good, That is a rise of 20C degrees.
To rise by 30C degrees, from 30°C to 60°C, is getting a bit too hot. A bigger heatsink is needed to keep the junction temperatures down.
"housecats climbing chalkboards" :
Daniel, I think you are just plain wrong about the Overture. You have expressed those strong feelings to me before, and I'm not doubting the validity of your ears or experiences. I am however suggesting you may not have heard every implementation.
The MyRef Fremen Edition is out in circulation for audition right now, and I'm making a public challenge for you to be placed on the list as I proposed earlier.
I suspect you will hear the stealth of a leopard combined with the force of a lion.
Bob, that MyRef has the sonic signature of an LM318, and I have promoted the MyRef amplifiers many times.
Also, when the Overture chips are run inverted+buffer or simple inverted or T-net inverting, the sonic signature is workable.
Those examples are not the majority.
The majority is the non-inverting gainclone which has too much distortion for me.
Also, when the Overture chips are run inverted+buffer or simple inverted or T-net inverting, the sonic signature is workable.
Those examples are not the majority.
The majority is the non-inverting gainclone which has too much distortion for me.
If it do not violate the maximum junction temperature then it is OK. Sometime DIY'ers too paranoid using small heat sink, bigger is better
Last edited:
Most of the data given by National is for resistive loading and with Tc held rigorously to 25°C.
As soon as Tc and the internals are allowed to increase in temperature the typical data changes, sometimes very significantly.
Look for the few bits of data that National give for high temperature operation @ high output currents, into reactive loads.
As soon as Tc and the internals are allowed to increase in temperature the typical data changes, sometimes very significantly.
Look for the few bits of data that National give for high temperature operation @ high output currents, into reactive loads.
If the LM3886 runs hot at idle, it's either because of an inadequate heat sink or HF oscillation. Ground loops may cause issues with EMI, but should not cause instability. That said, the use of a proper grounding scheme, Zobel networks, good decoupling, and short wire lengths is always preferred.
~Tom
~Tom
Hi Thank's for comments and suggestion. I do recheck this wiring and everything and I think there is not Zobel network for series with load. I made L over R10 ohm and series with load for every chip. Regarding grounding I'm pretty sure no problem, eventhough I did not configure as 'star grounding'. I take one point grounding system to chassis from PSU PCB. Experience heat spreading in heatsink is not good with present configuration, so I decide to change placement of chips. To do this I must move transformer position and change PSU PCB to shorter one in order to get more space.
Amazing !! After added the L and R in series with load , idle Voltage rail is +/- 31 Volt, it is fair for 4 chips (from PSU +/- 32,5 Volt) and after 15 minutes heatsink temp just a bit warm. No AC/DC detected from output.
Amazing !! After added the L and R in series with load , idle Voltage rail is +/- 31 Volt, it is fair for 4 chips (from PSU +/- 32,5 Volt) and after 15 minutes heatsink temp just a bit warm. No AC/DC detected from output.
Alumunium is 6 mm thick. The bar shape already just it is from factory in 6 mtr length. The strength / tenssion resistant is better compare if you cut alumunium slab with same thickness.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.