Matthewong said:
Yes, probably.
You said the temp was 43 degrees without any signal. If the temperature only went up by 22 degrees (more) at high power vs no signal, then you must have a good heatsink, so there is no way the temperature should reach 43 degrees without any signal.
I'll put forth a more scientific arguement: lets assume you have a 25 degree ambient temperature. So the heatsink rose 40 degrees at high output. Looking at the power dissipation vs. output power graphs, and interpolating for +-28 V supplies, you probably get around 25 to 30 Watt of dissipation with 8 ohms, at moderate to high output levels. So your heatsink is at best, (40 deg / 30 W) = 1.33 degree/W. Now since we know the chip should dissipate about 3 to 4 W when idle, you should get a maximum temp increase of about 1.33 deg/W * 4 W = 5.3 degrees when idle, or a temperature (assuming 25 degrees room temperature) of about 30 degrees. You said 43 degrees. The extra heat is quite likely caused by oscillations.
Of course if the ambient room temperature is 39 degrees C (102 degrees F) then the numbers actually work out and you probably don't have oscillations. But I am betting that the ambient temperature is much lower than that.
I've checked everything and most probably, the temperature reader has some problem. I've borrowed a more advance temperature reader from my friend and the LM3886's heatsink temperature was down to around 38.6C and the room temperature is around 32.5C. Tomorrow I'll check the current for both + and - DC voltage and see how many ampere it draws from from the power supply with and without the input signal.
Take a look at Figure 2 on page 6 of the LM3886 datasheet. Ignore the stuff having to do with single supply operations.
Add the following to your project.
Add L & R. Create L by 10 turns of #24 magnet wire (available at Ratshak) around a 10ohm/1W resistor.
Add Rsn and Csn, use a 1W resistor, values in schematic are fine.
Add Cc between in+ and in-
Add Cf and Rf2 in paralleled with Rf1 to create a high frequecy rolloff.
Make sure all metal parts (heatsinks, standoffs, potentiometer case, metal base of rectifier bridge, chassis, etc.) are connected to the safety ground. Use shielded input cables to reduce HF noise. Connect shields to safety ground.
Add the following to your project.
Add L & R. Create L by 10 turns of #24 magnet wire (available at Ratshak) around a 10ohm/1W resistor.
Add Rsn and Csn, use a 1W resistor, values in schematic are fine.
Add Cc between in+ and in-
Add Cf and Rf2 in paralleled with Rf1 to create a high frequecy rolloff.
Make sure all metal parts (heatsinks, standoffs, potentiometer case, metal base of rectifier bridge, chassis, etc.) are connected to the safety ground. Use shielded input cables to reduce HF noise. Connect shields to safety ground.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.