How many 3886 chips required to drive these woofers?

[rhythmdiy]

How many LM3886 chips are required to drive this woofer very tightly...

http://www.riviera-acoustics.com/cat...roducts_id=273

I want to use a 5 ohm version of this 10 inchwoofer to drive using 3886 chips...

Im thinking to use 4 in parallel to get enough current to drive the woofer with well controlled and tight bass..

Im thinking of the following...

using a 400VA toroidal transformer

and parallel the 4 chips to get enough current. Continuous output current rating is arround 2AMPs per chip and hence i think i need atleast 8 amps to drive this woofer...

what about Scanspeak how many chips are required for this.. woofer ...
http://www.riviera-acoustics.com/cat...products_id=76


Im getting very much confused with the true current ratings of this chip...ofcourse its in the datasheet as 7 amps but in reality it gives only 2+ amps...output current...

please guide me for the right number of chips to drive these babies....

thanks,
ken

[AndrewT]

Hi,
transient peak current ability is the critical factor.
The 3886 is specified as >=7A and typical =11.5A.

The minimum impedance of an 8ohm speaker can be as low as 35% of the nominal impedance when driven with a fast starting transient signal (and also a fast stopping signal). I would expect 35% for a 5ohm speaker as well.

Using these and assuming +-40V supplies, allowing a maximum output voltage of 36Vpk, your worst case transient speaker current could be 36/5/0.35=20.6Apk.
Three parallel 3886 would have a minimum peak output current of 21Apk and could be as high as 30 to 34Apk if all the chipamps managed the typical figure. The worst of the combination will set the limit for Spike operating.

However, what kind of music are you intending to play? How hot will the heatsink and chipamps get?
Spike protection will limit earlier as Tc rises. The 7Apk to 11.5Apk applies at Tc=25degC. You will get away with three 3386 for normal domestic duty, but if you are a party animal that likes lots of bass and/or approaching PA duty then even four 3886 might struggle if Tc rises significantly. Some builders seem to use National's suggested heatsink examples where Tc is allowed to rise to above 90degC. National do not specify how the maximum current that Spike lets pass when temperatures are up around Tc=100degC. I think that National's suggested sink capacity is quite misleading for the unwary that do not read the operating conditions. Double their suggested sink size if you want to stretch the chipamps to near maximum outputs.

[DJ Exprice]

Originally posted by AndrewT
...Using these and assuming +-40V supplies, allowing a maximum output voltage of 36Vpk, your worst case transient speaker current could be 36/5/0.35=20.6Apk.
Three parallel 3886 would have a minimum peak output current of 21Apk and could be as high as 30 to 34Apk if all the chipamps managed the typical figure. The worst of the combination will set the limit for Spike operating.
...

I'd say that three's a charm. Go for it.

3 in Parallel will sound great, too

[BWRX]

I don't have anything against all these parallel chip amp designs, but I fail to see the practicality of using more than two chip amps in parallel when you could just use the LM4702 or LME49810 to drive a discrete power stage that can handle higher current.

[AndrewT]

Originally posted by BWRX
I don't have anything against all these parallel chip amp designs, but I fail to see the practicality of using more than two chip amps in parallel when you could just use the LM4702 or LME49810 to drive a discrete power stage that can handle higher current.

and do a better job, not least due to keeping the output stage heat out of the front end amp section.
But some folk like Chipamps no matter how complex they become.

[rhythmdiy]

Originally posted by AndrewT
Hi,
transient peak current ability is the critical factor.
The 3886 is specified as >=7A and typical =11.5A.

The minimum impedance of an 8ohm speaker can be as low as 35% of the nominal impedance when driven with a fast starting transient signal (and also a fast stopping signal). I would expect 35% for a 5ohm speaker as well.

Using these and assuming +-40V supplies, allowing a maximum output voltage of 36Vpk, your worst case transient speaker current could be 36/5/0.35=20.6Apk.
Three parallel 3886 would have a minimum peak output current of 21Apk and could be as high as 30 to 34Apk if all the chipamps managed the typical figure. The worst of the combination will set the limit for Spike operating.

However, what kind of music are you intending to play? How hot will the heatsink and chipamps get?
Spike protection will limit earlier as Tc rises. The 7Apk to 11.5Apk applies at Tc=25degC. You will get away with three 3386 for normal domestic duty, but if you are a party animal that likes lots of bass and/or approaching PA duty then even four 3886 might struggle if Tc rises significantly. Some builders seem to use National's suggested heatsink examples where Tc is allowed to rise to above 90degC. National do not specify how the maximum current that Spike lets pass when temperatures are up around Tc=100degC. I think that National's suggested sink capacity is quite misleading for the unwary that do not read the operating conditions. Double their suggested sink size if you want to stretch the chipamps to near maximum outputs.

thanks alot for the reply...
and what the stuff is that no probs i can addup a massive heatsink... but all i need to know is that these chips can make the magic.

I belive it should do the job... lets see...

but what ive heard is that a magnet of size 155mm and 25mm thickness needs alteast 5amps of current to drive it properly...

thanks,
ken

[carlmart]

Originally posted by AndrewT
Using these and assuming +-40V supplies

For 4 chips in parallel I think I'd go for lower voltage. I think there was a table from National, where you could specify what you were doing and get a result.

[AndrewT]

Originally posted by rhythmdiy
ive heard is that a magnet of size 155mm and 25mm thickness needs alteast 5amps of current to drive it properly...

magnet size has nothing to do with current demand.

four chipamps in parallel give an equivalent loading of 20ohm for a 5ohm speaker.

National's spreadsheet does not allow impedance values like that.

But +-40V for a 20r load sounds quite acceptable to me.

[kuldeepsingh]

if u want to use all 4 ics, BPA 200 (bridged and paralleled) will be the best configuration and it can handle 5 ohms easily. And for a tighter bass, put at least 20,000 muF Capacitance per channel. I am sure this is enough.

[AndrewT]

Originally posted by DJ Exprice



I'd say that three's a charm. Go for it.

3 in Parallel will sound great, too

Originally posted by kuldeepsingh
if u want to use all 4 ics, BPA 200 (bridged and paralleled) will be the best configuration and it can handle 5 ohms easily. And for a tighter bass, put at least 20,000 muF Capacitance per channel. I am sure this is enough.

Hi,
a parallel/bridged arrangement will have EACH of the four amplifiers seeing the equivalent of the load impedance on EACH output.
Load a BPA with 5ohms and EACH amplifier thinks it is driving a 5ohm load.
Based on the above discussion EACH amplifier will need to be a parallel 3chip or parallel 4chip.
That adds up to a minimum of 12chips to do a BPA into 5ohms.
Is that what you meant when you said

the best configuration and it can handle 5 ohms easily.

?