Weird assembly issue
Phil Pilgrim (PhiPi)
Posts: 23,514
Sorry, software guys! Not that kind of assembly. 
I was assembling a proto Propeller board that uses the QFP44 package. The board also has a video DAC terminating in a 3.5mm phone jack (for compactness, compared to an RCA jack). I fired up the board and loaded a video demo program to check the video output, and everything was fine, except the color was missing! I looked at the signal on a scope, and the color burst looked okay, as did the chroma in the horizontal lines. I even compared it to the output from a known good Prop board, and the two were identical. What the ... ?
While handling the board, I noticed that the color would sometimes return, but disappear when I set the board down. After awhile, I realized all I had to do was press on the Propeller chip to make the color return. So I resoldered and double-checked all the pins on the DAC side of the chip and tried again. 'Still no color, except when I pressed on the Prop. @#$%^!!!
To make a long story bearable, it was a bad crystal connection. Apparently one of the Prop pins going to the crystal had a bad connection to its pad. But there was enough capacitive coupling for the crystal to function anyway — just at a slightly off frequency. My TV monitor could forgive the slightly off horizontal frequency but not the skewed color burst frequency. Resoldering the crystal conneciton fixed the problem.
The lesson is to make sure to inspect each pin under a magnifier for a visible solder fillet. The pin in question was missing the fillet, a fact which I missed on the first — and subsequent — inspections.
-Phil
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Just a few PropSTICK Kit bare PCBs left!
I was assembling a proto Propeller board that uses the QFP44 package. The board also has a video DAC terminating in a 3.5mm phone jack (for compactness, compared to an RCA jack). I fired up the board and loaded a video demo program to check the video output, and everything was fine, except the color was missing! I looked at the signal on a scope, and the color burst looked okay, as did the chroma in the horizontal lines. I even compared it to the output from a known good Prop board, and the two were identical. What the ... ?
While handling the board, I noticed that the color would sometimes return, but disappear when I set the board down. After awhile, I realized all I had to do was press on the Propeller chip to make the color return. So I resoldered and double-checked all the pins on the DAC side of the chip and tried again. 'Still no color, except when I pressed on the Prop. @#$%^!!!
To make a long story bearable, it was a bad crystal connection. Apparently one of the Prop pins going to the crystal had a bad connection to its pad. But there was enough capacitive coupling for the crystal to function anyway — just at a slightly off frequency. My TV monitor could forgive the slightly off horizontal frequency but not the skewed color burst frequency. Resoldering the crystal conneciton fixed the problem.
The lesson is to make sure to inspect each pin under a magnifier for a visible solder fillet. The pin in question was missing the fillet, a fact which I missed on the first — and subsequent — inspections.
-Phil
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Just a few PropSTICK Kit bare PCBs left!
Comments
"What? the mighty Phil is stuck in pasm and asking for help?! no way"
Meanwhile, I learned something else about surface mount soldering using solder paste: Don't let a half-populated board sit overnight! I was populating a second copy of the above-mentioned board, when I decided there was one more thing I needed to test on the first one before wasting parts on the second one. My modus operandi is paste-and-place, paste-and-place, rather than doing all the pasting first. I had about eight components placed this way before setting it aside. This morning I finished populating the board in put it in the IR toaster oven to reflow. During the reflow stage, I heard some pops. 'Turns out, these were from the components I had pasted the night before. The had literally leapt off the board and lay on the bottom of the oven. My theory is that the flux dried around the solder blobs leaving liquid flux inside. When this came to a boil the blobs exploded, propelling the parts off the board.
-Phil
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Just a few PropSTICK Kit bare PCBs left!
Never had experience with paste, ovens and reflow. Tried soldering .50 pitch LQFP64 and was amazed how easy it is with a decent iron and a soldering wick. I didn't even use any magnification. The hardest part was lining it up
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My Prop Info&Apps: ·http://www.rayslogic.com/propeller/propeller.htm
Phil,
I have seen this before. I have had it happen when using cold solder paste too. When the solder paste condenses water it will do the same.
James L
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer
Lil Brother SMT Assembly Services