LM323K installation with heatsink

8 posts / 0 new
Last post
Online
Last seen: 1 hour 29 min ago
Joined: Sep 20 2019 - 10:37
Posts: 11
LM323K installation with heatsink

I have been installing parts on my NTI Newton board and I just put on the LM323K.  I haven't powered it up yet, because I am unsure how to know if the thermal grease or the heatsink will cause a short.  Should I put some type of insulation around the pins on the regular? Is this just one of those things where it works or you fry it?  The holes in the lm323k are smaller than the heatsink, so trusting those to take up the slack and center it seems risky too.  Maybe this isnt a big deal, but I have poured a considerable amount of time and money into this board and it seems like I should ask if anyone else has had a problem at this point.

Offline
Last seen: 1 day 5 hours ago
Joined: Oct 9 2011 - 12:54
Posts: 1243
WrongDateCode wrote:

WrongDateCode wrote:

I have been installing parts on my NTI Newton board and I just put on the LM323K.  I haven't powered it up yet, because I am unsure how to know if the thermal grease or the heatsink will cause a short.  Should I put some type of insulation around the pins on the regular? Is this just one of those things where it works or you fry it?  The holes in the lm323k are smaller than the heatsink, so trusting those to take up the slack and center it seems risky too.  Maybe this isnt a big deal, but I have poured a considerable amount of time and money into this board and it seems like I should ask if anyone else has had a problem at this point.

 

I'd be more worried about your heat sinks on the smaller regulators touching.  

 

If you use the correct RAM and have a small fan blow on the whole area, you don't need the heat sinks on the smaller regulators. Some RAM chips like the Intel ones can use extra current and cause the smaller regulators to really heat up.   

 

 As for the large LM323k, you should be fine, in fact you put a lot of thermal paste on it.  you may want to clean some off using a qtip and ISP. It won't harm anything with the extra compound, it's just cosmetic.

 

Cheers,

Corey

Offline
Last seen: 1 day 9 hours ago
Joined: Jun 5 2008 - 07:26
Posts: 422
You can use an ohm-meter to

You can use an ohm-meter to check for shorts between the input (pin closest to top of board), output (the other pin) and ground (case of the LM323k) or any other signal that you are concerned about.

 

regards,

Mike Willegal

 

 

Online
Last seen: 1 hour 29 min ago
Joined: Sep 20 2019 - 10:37
Posts: 11
Thanks for the suggestions.

Thanks for the suggestions.  I think I am ok as far as the short goes, did use the ohmmeter to test.  The lm323k seems to only be putting out 4.8v.  I tried both of the ones I have (same part from same place) and they are the same.  When I light up the board with the video section populated I get a flashing between D and $ and @ and blank.  See images.  I had lots of fits and starts because I am not sure what parts are good and not.  However, while playing with a logic probe I noticed that when I touch the 2504 chips at D5 on pins 5 or 6 it seems to start smoothing the problem out.  Is this a voltage drop problem?  Is the 4.8v too low?

Online
Last seen: 1 hour 29 min ago
Joined: Sep 20 2019 - 10:37
Posts: 11
E1624EFD-3017-41DD-9342
Online
Last seen: 1 hour 29 min ago
Joined: Sep 20 2019 - 10:37
Posts: 11
Did a bunch more

Did a bunch more troubleshooting and also found a neat chart from another post on walking through the 25xx chips to validate they work.

  • all 25xx removed flashes boxes, pass
  • add 2513, solid @ grid, pass
  • add 2519, blinking @ grid, pass
  • add 2504 to d4a and d4b, blinking @ in every other column, alternating "L" and "," in other colum, pass
  • add 2504 to d5a and d5b, this should give blinking @ in one colum and some other alternating characters in the other colum, but instead the characters are all garbled.  fail
  • removed 2504 from d4 and populated d5 again, blinking @ column and alternating "C" and "#" on other column. pass
  • populate d4b and d5a, so one chip per socket, blinking @ in one column and blinking "I" and ")" in the other

for what ever reason I cant have more than 2x 2504 chips in d4 and d5 or it garbles everything.  Looking at the scematic I replaced all chips that could be in the line, c4 74157, c3 2519, and even d2 2513.  No change. Am I missing something obvious?  Should I start removing and replacing sockets? I did retouch all solder joints. My 7.5k resistors all test ok, but seem to vary between 7.8k and 8.1k. The board is an NTI Newton AC1 rev B1.

Offline
Last seen: 1 day 5 hours ago
Joined: Oct 9 2011 - 12:54
Posts: 1243
WrongDateCode wrote:

WrongDateCode wrote:

Did a bunch more troubleshooting and also found a neat chart from another post on walking through the 25xx chips to validate they work.

  • all 25xx removed flashes boxes, pass
  • add 2513, solid @ grid, pass
  • add 2519, blinking @ grid, pass
  • add 2504 to d4a and d4b, blinking @ in every other column, alternating "L" and "," in other colum, pass
  • add 2504 to d5a and d5b, this should give blinking @ in one colum and some other alternating characters in the other colum, but instead the characters are all garbled.  fail
  • removed 2504 from d4 and populated d5 again, blinking @ column and alternating "C" and "#" on other column. pass
  • populate d4b and d5a, so one chip per socket, blinking @ in one column and blinking "I" and ")" in the other

for what ever reason I cant have more than 2x 2504 chips in d4 and d5 or it garbles everything.  Looking at the scematic I replaced all chips that could be in the line, c4 74157, c3 2519, and even d2 2513.  No change. Am I missing something obvious?  Should I start removing and replacing sockets? I did retouch all solder joints. My 7.5k resistors all test ok, but seem to vary between 7.8k and 8.1k. The board is an NTI Newton AC1 rev B1.

 

OK, so you did the first obvious thing and looked for cold solder joints by retouching up everything, how about the opposite.  Did you check if you had solder bridges?    before I put my brain in troubleshooting mode to help, visually inspect the board under magnification. 

 

 

Online
Last seen: 1 hour 29 min ago
Joined: Sep 20 2019 - 10:37
Posts: 11
Ok, I combed over the board

Ok, I combed over the board with a magnifying glass again and it looks good.

 

New thought and question, should the @ blinking always have a blank space next to it?  Mine doesnt.  I put the 2513 on a logic analizer and I was a little shocked by how noisy the address pins are.  This seems like it has to be part of my problem. It feels like the lows are not sticking, but again I have nothing to compare with.  can someone post a good 2513 capture screenshot?

 

Log in or register to post comments