Yet Another LME49811 + STD03 Build

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The question is: shall I share the heatsink with power transistors or not?

I wouldn't. In my first setup, I put the LME on the same heat sink as the power transistors, thinking it would be convenient. However, I don't think those LME chips are intended for that. They don't have an obvious attachment hole, just those 'nibs' on the ends of the package. They do need to be isolated from the heat sink, so I cut some thermal washers to fit. I ended up cracking two packages before calling it quits and putting the LME on its own separate heat sink.

There are many heat sinks available for this purpose. Look for one that's intended for a TO-247 package and uses a clip for attachment. You'll have to add a spacer between the clip and the package to get a good fit.

~Tom
 
Moderator
Joined 2002
Paid Member
jackinnj did some tests on THD and sharing the heatsink with the outputs, and found it best to keep them separated, so the mechanical issue isn't the only one. I think they're on his site, and pertain to the 4702 though I suppose the issues would be similar.
 
Thanks to everybody. I seem to have found a nice heatsink. Well, it looks a bit heavy - but not smaller than required, and this is good.
But about D5 - I'm still unclear. The bias network can conduct only in one direction - from top to bottom. And the bias pot is shunted by D5 exactly in this direction. Why?
 
D5 across the POT

One reason could be:
The pot and the D5 have opposing thermal coefficients. Thus when the ambient temperature varies over a large margin say 18 deg C to 55deg C, the diode compensates for the change in resistance of the pot and thus helps in bias current stability over temperature variations.
 
Hello tomchr,

While I wait for my PCBs of the scalable IRS2092 amp, I wish to try your circuit in post #88 but with only one pair of STD03x and for +/-35V that will be allowed to go upto +/-42V. Do you have any special recommendations to do this? What should be the size of the heatsink for the LME49811 and for the STD03x?

Thanks in advance!
 
While I wait for my PCBs of the scalable IRS2092 amp, I wish to try your circuit in post #88 but with only one pair of STD03x and for +/-35V that will be allowed to go upto +/-42V. Do you have any special recommendations to do this? What should be the size of the heatsink for the LME49811 and for the STD03x?

What is your experience with STD03x or the ONSemi Thermaltrac Transistors used with LME49811? Is any one preferred?

For my experience with the STD03 and LME49811, I suggest that you read through this thread as it sums up my experience rather nicely.

I do recommend putting the LME49811 on a separate heat sink. I used a small one intended for use with two TO-220 packages. It's probably 5 K/W or so. You should be able to calculate the power dissipated in the LME49811 from the supply current listed in the data sheet and your supply voltage. Pick a heat sink that will ensure that the IC remains below the limits listed in the data sheet.

For the power dissipated in the output devices and the heat sink calculations, I suggest consulting a textbook, such as Sedra & Smith: "Microelectronic Circuits". I've linked to the 5th edition, but any of the editions (at least back to the 3rd) will work and will be less expensive than the 5th (and much, much less expensive than the most recent, 6th, edition).
With +/-35~42 V, you'll probably find that even with a large heat sink (say <0.7 K/W) you may need more than one pair of STD03s.

~Tom
 
The pot and the D5 have opposing thermal coefficients. Thus when the ambient temperature varies over a large margin say 18 deg C to 55deg C, the diode compensates for the change in resistance of the pot and thus helps in bias current stability over temperature variations.

No. The diode across the pot is to limit the bias current under conditions such as start-up and in the event the pot fails. The diode limits the maximum voltage across the bias trimpot. The current normally flows through only the trimpot. The diode hardly carries any current during normal operation.

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