I've been looking at the LM3886 datasheet and it seems that I could get a lot more power using this chip as the current won't be limited if I want to drive it at a higher voltage. Instead of using my 2x18Vac toroidal, I could use a 2x25Vac (35V dc).
I've also read that the sound quality of the LM3886 is almost identical to the LM3875? Can anyone comment on this?
I've also read that the sound quality of the LM3886 is almost identical to the LM3875? Can anyone comment on this?
Have you looked at the TDA7293 ? which is specifically designed for "parallel" operation ?
Even on it's own it's capable of very high outputs.
http://www.datasheetcatalog.org/datasheet2/9/0ozxelyhtgfl5jsx9108fj3sqq3y.pdf
Even on it's own it's capable of very high outputs.
http://www.datasheetcatalog.org/datasheet2/9/0ozxelyhtgfl5jsx9108fj3sqq3y.pdf
have a look at the minimum (almost guaranteed) output current lower limit for pulsed operation with Tc=25degC.
That (7Apk) is the best possible that this chipamp can be relied on to achieve.
This is not nearly comparable to what discrete output stages can supply in similar drive situations when designed/built properly.
there are some paralleled 3chip lm3886's on ebay. I picked up a couple of them for less than $20 each. A price I could not refuse. They worked fine. I did change the caps on the board from 35volt to 50 volts and ran the chips at near their maximum rating of +/- 40 volts. Maximum power at 8ohms was around 70 watts, 120 watts at 4 ohms and yes they did drive a 2.5 ohm load but things did get a little hot, real hot actually. I had to change a cap that was in parallel with one of the feed back resistors, I think I changed it to 35pf from 150 or something like that. If I didn't the top end started rolling off at 12Khz.
I've been trying to work out why my amp cuts the music (clips maybe) when turned up very loud. I used the Overture Design Guide 15.xls and entered my values. I have +/- 25V dc, LM3875 and a feedback resistor of 22KOhm.
Now, vPeak at load = 20.65
Gain (Not BTL) = 23 V/V
Does the above mean that if my input had a peak of 2V, then when my amp is turned up full I would be trying to put out 46V, which is obviously larger than my supply so it clips?
Also, I would like to know which cell is actually telling me the power of my amp. One cell says: "1% THD output : 35.54 Watts", another says "Max Po at POUT : 21.11 Watts" and another says "Total PO/IC : 21.86 Watts"?
Sorry if these questions stupid, but this is my first amp.
I must say, I just added a snubber circuit to my PSU, the same design as BrianGT from chipamp.com, and the music is noticeably better, much punchier and controlled bass.
Now, vPeak at load = 20.65
Gain (Not BTL) = 23 V/V
Does the above mean that if my input had a peak of 2V, then when my amp is turned up full I would be trying to put out 46V, which is obviously larger than my supply so it clips?
Also, I would like to know which cell is actually telling me the power of my amp. One cell says: "1% THD output : 35.54 Watts", another says "Max Po at POUT : 21.11 Watts" and another says "Total PO/IC : 21.86 Watts"?
Sorry if these questions stupid, but this is my first amp.
I must say, I just added a snubber circuit to my PSU, the same design as BrianGT from chipamp.com, and the music is noticeably better, much punchier and controlled bass.
Yep, you've got it. A typical CD player will have a peak output voltage around 3V, just for reference.
The max power output depends how distorted the output becomes. If you let the amp clip hard you get more power as then the output is close to a squarewave. You might be seeing output power measured at 0.1% or 1% or 10% distortion. I haven't played with the spreadsheet myself but I'll go have a look...
Edit: Those cells (N14,N15) are power dissipation for the IC, not in the load.
Seem eminently sensible questions to me.
Last edited:
That is one possible reason. Vpeak at load 20,65 V means you get an unclipped sine wave output of ~14,6 V. With a gain of 23 times you only need 0,634 V at the input to drive the amp into clipping. In reality even a little less, because the supply voltage will sag under load.
Another possibility is that the speakers demand too much current and the current limiter sets in. That could be the case, if you have speakers
- with low impedance (e. g. nominal 4 Ohm)
- with high diaphragm mass
- in a small CB
and in general when you try to achieve high SPL at very low frequencies.
1% THD output is the approximate value you are looking for.
Pd at Pout tells you at which output power you get the highest power dissipation and Total Pd/IC tells you how much power dissipation that is so you can calculate the necessary heatsink size.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.