You're in the wrong section.
Chipamps are limited to ~70W maximum
They are also limited in peak current capability.
Don't expect good quality and high power into a 4ohms speaker.
Switching amps are often recommended for bass duty, but I would recommend a discrete or an lme498xx based amplifier for good bass performance.
Chipamps are limited to ~70W maximum
They are also limited in peak current capability.
Don't expect good quality and high power into a 4ohms speaker.
Switching amps are often recommended for bass duty, but I would recommend a discrete or an lme498xx based amplifier for good bass performance.
I'll second that. (It was going to be my suggestion too.)
Use a 25V transformer to give yourself about 39V on each rail. Be careful that you send the same signal to each amp and match up the gain resistors.
I found project like this 100W LM3886 Parallel Stereo Power Amplifier
Its 50w @ 8 ohm
and 100w @ 4 ohm
I think that will do .
But the most expencive thing is transformer.
Only thing the regulator schematic is missing, but the good thing is the parts are listed .
Tekko
I checked out the TDA7293, if u put that to 4 ohm load it will be only like 70-80w real power .
Its 50w @ 8 ohm
and 100w @ 4 ohm
I think that will do .
But the most expencive thing is transformer.
Only thing the regulator schematic is missing, but the good thing is the parts are listed .
Tekko
I checked out the TDA7293, if u put that to 4 ohm load it will be only like 70-80w real power .
Last edited:
Actually it's better than 100W. Use 39V rails and 0R1 output resistors. Put a Zobel on the o/p, definitely. You can do without an L//R.
Your sub may be 4 ohms nominal but it will have a DCR of about 3.3. That gives you about 200W, depending on where you choose to pick the point vis distortion.
It will be plenty. Excursion will likely be your limiting factor, not power (though of course that depends on the unit and how low in frequency you are going to go.)
Your sub may be 4 ohms nominal but it will have a DCR of about 3.3. That gives you about 200W, depending on where you choose to pick the point vis distortion.
It will be plenty. Excursion will likely be your limiting factor, not power (though of course that depends on the unit and how low in frequency you are going to go.)
Thats per chip, but two in parallel will do 140 watts in 4 ohms and around 90W in 8 ohms.
Unless the site i saw these ratings on was false.
Anyways for subwoofer use, i'd look at a discretely built amp rated at 200 watts minimum. I base this on personal experience, i was watching the movie Tron Legacy and noticed that my 220 watt subamp quite often clipped hard while the sub wasent even loud.
Last edited:
Where did that come from?
Let's give you a chance of being near the truth.
200W into 3.3ohms impedance requires an output of 36.3Vpk and 11Apk.
How can that output come from a supply that holds a quiescent voltage of +-39Vdc?
If we were to use 4ohms as the load impedance, then we find that the required outputs for 200W become 40Vpk and 10Apk. That is even more impossible to achieve with a +-39Vdc supply.
Now look at a 3886 datasheet.
A dual pair in parallel passing 11Apk to the output will each have to pass 5.5Apk. The Vdrop through the 3886 will typically be ~4.5V. The Vdrop through the balancing 0r1 resistor will be 0.55V.
This combined with the sag (assumed to be ~5V) in the supply voltage when delivering full power will reduce the peak output voltage to ~ 39 - 4.5 - 0.55 - 5 ~ 29Vpk.
That equates to ~105W into a 4ohm load.
If one were to design the build to maximise the power delivery, then expect upto 120W into 4r0, not 200W.
About now I could introduce the difficulties the paralleled 3886s will face when trying to drive a reactive speaker load, but I won't bore you with repeating my oft quoted caution.
Last edited:
The 4.5V drop is across both supplies, I believe. The peak sag on the rails in practice is never greater than about 2V in practice (more often 1V on scope with 300VA) and 3.3 ohms is probably higher than the DCR is likely to be. Most of the Peerless ones are 2.7 ohms, so we are back in at least the same ballpark. The reactive element is really only at resonance and current falls through that. The inductance doesn't get to kick in much because you are filtered so low down.
Believe me, I've done it and it's about 6 feet from where I'm typing on a credit card sized footprint. And it is also the coolest running amp I have ever made. I have not managed to get it above skin temperature so far because excursion sets the limit (and I don't have a load that will take that sort of power). Incidentally, it might be worth going up to 0R15 to be cautious.
A PA150 with 3 paralled LM3886's running on a 28V-0-28V transformer would most likely suit your requirements or even a PA200 with 4 of them to reduce the current demands for each chip could be even better.
FWIW
jer
Well, that would be the ultra conservative approach. Layout gets difficult, though. But since he's not augmenting the LF with a Linkwitz circuit, he could probably manage on a single chip. The only reason for high power demands in subs is if you are extending the LF response of the drivers. Otherwise there is no difference from just driving a big speaker assuming the efficiencies are similar.
Given that, two chips would seem to be ample and then all he's really doing is spreading the dissipation across two chips, which is no bad thing.
I cannot confirm from any of my experiments that any of what you say/claim in this paragraph to be true.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.