Another LM4780 for Bi amping

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi Fellow Diyers,

Here is a Diyable single sided LM4780 PCB if you are interested.

Nothing exotic or Fancy about the schematic ,it is from manufacturers Datasheet.

I have built this Amp recently a month ago for my Dipole speakers driven by miniDSP.

Good things as in manufacturers datasheet
1.Approximately 50watt X 4 channels
2.Onboard AC rectifier (directly connect the 25-0-25 AC transformer)
3.No output clicks and pops while startup to blow out tweeters (mute function used) .
5.Tested and working since last 30 days.
6.No ground loop or Hum issues faced so far.
7.Single layer diyfriendly PCB
Bad things
1. Would need a massive Heatsink
2.Fixing the chip on Heatsink can be tricky which plays crucial part in performance and life of this chip.
3.Sometimes LM4780 might be unobtainable.
 

Attachments

  • Lm4780.png
    Lm4780.png
    191.5 KB · Views: 349
  • Lm4780 Schema.jpg
    Lm4780 Schema.jpg
    244.7 KB · Views: 360
I had to contain 4 amplifiers with 200Va trafo into 30cmx20cmX10cm cabinet.
So tradeoff was small heatsink.
Lm4780 is dual Lm3886 in one package hence double the heat,I ended up forced cooling with a Fan which is triggered by a thermistor above 80dgreeC.

LM4780 is far better than any 50 w Class D amp I have heard,Initially my plan was to use class D amps for the available space but Dipoles speakers were not forgiving with class D amps Then I went back to LM3886 and TDA7293.
I preferred LM3886 as whatever TDA7293 chip I got my hands on were all counterfeits (from UTstar and ebay) .

After fiddling with LM3886 I finally decided upon 2 in 1 LM4780 .It did not dissapoint except for super heat.
In order to further reduce the the thermal resistance I discarded the heatsink Insulator and bolted the Chip directly on the heatsink which makes heatsink active negative.whole PCB and heatsink has been isolated from cabinet using spacers.
Rgds.
 

Attachments

  • 60999132-9638-45ed-8a4d-14d6ad354613.jpg
    60999132-9638-45ed-8a4d-14d6ad354613.jpg
    123.8 KB · Views: 414
  • fba9a303-b85c-4fb2-b59b-7eae65d8c74b.jpg
    fba9a303-b85c-4fb2-b59b-7eae65d8c74b.jpg
    123.9 KB · Views: 297
Last edited:
Hi Tom. I have read your LM3886 site. Thanks for sharing those thoughts.

My interest is speculative at this point. I unearthed a pair of long forgotten 4780's from my parts box, and now am looking for mischief to create. It was the dipole remark earlier that caught my interest more so than bridging.

Bridging always seemed a bad idea, although I've never seen any words written on what seems to me is a guaranteed DC offset. Any -- seemingly unavoidable, actually-- slight mismatch between the two amps or just a variance in gain puts a DC component across the speaker.

Anyway... the two thoughts are related. I'm out of DSP channels and am in possession of some spare parts that are compatible with the excess volt amps of my newly larger toroid.
 
My interest is speculative at this point. I unearthed a pair of long forgotten 4780's from my parts box, and now am looking for mischief to create.

Fair enough. :)

Bridging always seemed a bad idea, although I've never seen any words written on what seems to me is a guaranteed DC offset. Any -- seemingly unavoidable, actually-- slight mismatch between the two amps or just a variance in gain puts a DC component across the speaker.

Not necessarily. You can put a DC servo on a bridged amp and drive the DC offset to (near) zero if you wish.

The gain mismatch can also be dealt with by using 0.1% tolerance resistors, which are now available for really not that much money. That doesn't eliminate the mismatch, obviously. It only makes it smaller.

If you're stuck at a lower supply voltage, bridging can be the ticket to getting the output power you want. As you've noted, it does come with a set of drawbacks, however.

Tom
 

PRR

Member
Joined 2003
Paid Member
...Bridging always seemed a bad idea, although I've never seen any words written on what seems to me is a guaranteed DC offset. Any -- seemingly unavoidable, actually-- slight mismatch between the two amps or just a variance in gain puts a DC component across the speaker.....

All OCL amps have a DC offset.

For audio amplification we normally roll-off subsonic gain to unity, so "gain resistor match" is completely off the table.

I won't touch the idea of "bridging bad", except to point out two extremes:

Load offset doubles. But power output has quadrupled, meaning you probably have hefty woofers which will ignore tiny or 2X tiny offset.

Load offset *cancels*. This can happen (with good tracking even) in a monolithic dual.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.