Bridge or Parallel?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
Roushon

Roushon

Member
Joined 2004
Thanks

juma said:
The schematics above was not the subject of discussion.
That drawing shows a combination of bridged and paralleled topology and it will deliver more power because bridging doubles the voltage on the load. This will result in more current flowing through the load.
So, the previous conclusions stay valid: there will be no increase in power without increasing of voltage or reducing of the load's resistance.

Cheers!
Click to expand...

Thank you Juma. You proved again that this is physics which follows laws of science. Wrong words or pictures do not work, not at least in this forum.

Regards
Roushon.
 
juma

juma

Member
Joined 2005
Hi tiltedhalo
please don't feel that way about our discussion. It's not a question of who's got the uper hand at anything. We are here to exchange opinions and knowledges. Nobody is a all-knowing-deity, everyone has something to learn at some point. Your design proves that you are a contributing member of community, there is no reason to feel upset.
I'm sorry if I made you feel that way.

Best regards !

Cheers!
 
G

gmphadte

Member
Joined 2003
The pcb...
The pairs of capacitors are not needed as this is not dual mono design. U can put one double value cap instead. However two caps have smaller ESR.

U have to make one point as a main star at one cap lead, for all the supply (+/-/gnd) connections. Rest all will start from here, even the other parallel caps.

Tracks that run below the ICs (on component side) should preferably be gnds.

Will come later, too busy now.

Gajanan Phadte
 
Sangram

Sangram

Moderator
Joined 2002
Paid Member
tiltedhalo said:
would a moderator pull my post and this thread and delete my posts and user account, Thank you
Ive had to many things miss-represented and ideas taken that I no longer want anything to do with this forum.
Thanks again
Click to expand...


juma said:
Hi tiltedhalo
please don't feel that way about our discussion. It's not a question of who's got the uper hand at anything. We are here to exchange opinions and knowledges. Nobody is a all-knowing-deity, everyone has something to learn at some point. Your design proves that you are a contributing member of community, there is no reason to feel upset.
I'm sorry if I made you feel that way.

Best regards !

Cheers!
Click to expand...

Hi,

Just my quick POV having built BPA.

Depending on your power supply and load, you are either current limited, voltage limited or distortion limited w.r.t the output power deliverable to the load.

And in the case of speakers they usually are not 8 ohms over their entire range, can dip lower and usually do, so the comparison to a resistive load is a little problematic.

In a parallel connection each amp sees twice the load impedance, and in bridge each amp sees half the load impedance, so in a BPA each amp half sees exactly the same as the actual load impedance.

The logic of paralleling a bridged amp is to ensure the amp is not current limited by the output stage, and the logic of bridging a parallelled amp is to double the output swing into the load. So both views are correct....

But yes, in parallel mode the main benefit into say an 8 ohm load is the capability of the output stage to deal with transients and due to paralleled output pairs, reduction of beta droop (distortion reduction) rather than increased RMS power. AFAIK in theory a parelleled output stage will not increase output power into a purely resistive load...

I can't see any of the schema's tiltedhalo posted earlier, but he has a point. Say you have an 8 ohm load and want to run the chip off very high rails. In this example 40 volt rails was mentioned. with one chip you'd be running a risk of frying the chip by running close to the outer edge of the SOA. If you parallel another chip/amp, the total Pd comes down for the chips, and as long as the transistors can handle the voltage, the current dissipation will come down. As a result, higher power outputs become feasible into high (8 ohm) impedance load.
 
Roushon

Roushon

Member
Joined 2004
gmphadte said:
The pcb...
The pairs of capacitors are not needed as this is not dual mono design. U can put one double value cap instead. However two caps have smaller ESR.

U have to make one point as a main star at one cap lead, for all the supply (+/-/gnd) connections. Rest all will start from here, even the other parallel caps.

Tracks that run below the ICs (on component side) should preferably be gnds.

Gajanan Phadte
Click to expand...

Thanks. I will modify the circuits and the PCB accordingly and soon put them here.

Sorry, only today I checked my email and read your email. Not able to reply as there is a problem with the email server.

Regards
Roushon.
 
P

ppcblaster

Member
Joined 2006
tiltedhalo schmetic

Would someone please email me tiltedhalo's schmetic?

it did not display in my browser. I would also like to ask the

Moderators to contact tiltedhalo and ask him to please reconsider

his decision to drop out of the forum due to an unfortunate

insulting remark. Non of us are responsible for what is posted

here. I think a moderator may have stopped that insulting post

before it was posted, and tiltedhalo would still be here, with

what may have been one of the best designs so far.

Thank You
 
S

SpittinLLama

Member
Joined 2005
Both TiltedHalo and Juma are correct but only when things are clearly explained, which Juma did a good job of clearing up. As stated, you will either be limited by voltage swing, output current OR thermal limits. Having a single LM3876 or LM3886 or 1 ch. of the LM4780 into an 8 ohm load will give you X power. Now if you want more power into that 8 ohm load you have to know what is the limiting factor. For 8 ohm at any supply voltage for these chips it is NOT current. It is either voltage or thermals. So a bridge will solve the voltage issue but make worse the thermal issue. Parallel will do nothing to solve the voltage issue but will solve the thermal issue AND give more potential current drive if the load dips too much in impedance. You will never get 100W from a single chip or channel that National makes (class AB) into an 8 ohm load with less than 1% THD. Sure it can be done at 10% THD. Add a second chip and run at 42V and you will probably get there but only barely. Change to 4 ohms and thermal issues will limit you the most so a second parallel chip really helps plus you can then push a lot of current. It is just simple electronics but you have to know what is limiting your output power. I am pretty sure it is explained pretty well in the AN-1192 note but maybe not. If this is still confusing then I can explain more and how each amplifier sees a load impedance that is different than what is actually there in a bridge or parallel configuration. It is really just basic electronics.

-SL
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom