LME49830 Project/Boards...

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
So ever since OPC introduced me to the LME amplifier chips I have wanted to make an amp based on them. Unfortunately I was unable to buy the boards he offered and so have designed my own. They work and feature a better heatsink for the LME.

There's an option to run the LME and output stage on separate rails.

Here's some pictures and specs...

LME49830.png


LME49830Performance.png


Blue line is 1W,
Green line is 50W

Both were run into 4R, 80KHz bandwidth, 35V rails (LME on the same rail) and at 300mA bias current. This is just with initial component values, I expect to get better performance once I have fiddled and run the LME on higher regulated rails.

LME49830Layout.png


image-1.jpg


You don't have to use an SMA connector for the input there are just holes there.

If anyone is interested in boards I will get some made. I have had a quick look at prices and for an order of 50 boards I will have to sell them at US$15+postage.

Last time I posted there were a few people interested so to get an idea of a number post your name and the names previous in your post with the number you want.

Thanks
Boscoe
 
I realize that it is a OPC clone, but what if by accident, you get a short on the load side? What protection measures are you incorporating or is this different discussion all together?
Blow a fuse all the way back to the mains fuse?
10uF/50V/1206 COG, bias cap, I bet that ain't cheap. Why 50V? I wonder if you use a X7R, would it make a difference?
I know Bob C. suggested in his book a optional R-C snubber, across the MOSFET Gate Drain?
Gate Zeners? Optional TPC?
Diodes across Drain/Sources?
Just some suggestions.
Rick
 
I realize that it is a OPC clone, but what if by accident, you get a short on the load side? What protection measures are you incorporating or is this different discussion all together?
Blow a fuse all the way back to the mains fuse?
10uF/50V/1206 COG, bias cap, I bet that ain't cheap. Why 50V? I wonder if you use a X7R, would it make a difference?
I know Bob C. suggested in his book a optional R-C snubber, across the MOSFET Gate Drain?
Gate Zeners? Optional TPC?
Diodes across Drain/Sources?
Just some suggestions.
Rick

It's not an OPC clone otherwise the board would be a lot like The Wire. The schematic is kept simple which makes it similar to OPC's design.

Use fuses on the outputs of you PSU or output protection off the board.

The bias cap is 100nF 50v NP0/C0G.

I don't think gate zeners are necessary. I haven't really got space for diodes across the outputs!
 
The National Semi prototype boards (for the 2SK1530/2SJ201) are showing up on EBay -- I've found that once an item starts trafficking they come out of the woodwork:

An externally hosted image should be here but it was not working when we last tested it.


Should be easy enough to use with lateral mosfets by rearranging the pinout, or the readily available IRFP or Fairchild devices.
 

AKN

Member
Joined 2005
Paid Member
Oh, I thought the layout was finalized - fixed. Now we have a new PCB layout.

My immediate response is like this:
Ground plane partitioning between high - low currents. High decoupling current directed through cap pins. And how about implementing gate zener protection?
 

Attachments

  • boscoe.png
    boscoe.png
    51.4 KB · Views: 465
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.