Efflorescence using plumbers flux

[brandon7861]

A while back I built an active probe for my oscope and fc, but I had no electronic flux. I figured I could get away with using plumbers flux as long as i washed it off really well. I used acetone, water, acetone again with a fine paint brush trying to get under the parts the best I could. It looked good and worked good, for a while. it didn't work today, so I pulled it out of its copper pipe. This is what happens when the solvent traps a little flux under the part. The solvent migrates to the surface where it evaporates depositing the zinc chloride over everything.

Now I need to decide if these heat gun victims go back in, or new ones.

 

[nomadradio]

Zinc chloride is an electronic component's worst enemy. After a solvent wash, we're use to flooding each affected solder joint with liquid or paste rosin flux, melting the solder on each affected spot, and repeating the solvent wash. Might have to lather, rinse, repeat depending on how concentrated the contamination is.

If the leads are rotted off the component, you're too late. Try the same procedure for affected component leads. Hard to predict how many of them will still have enough metal on them to reuse.

73

 

[brandon7861]

I stayed up half the night removing all components and replacing them with new ones. Also found more errors in my schematic. I found out small signal amplifiers work best with lower bias voltages and each stage needed its own bias voltage to achieve the gain I wanted.

That David Jewsberry "poor mans 1GHz probe" schematic I used for the BF998, the bias is not ideal there either. His schematic shows 4.7k and 6.8k for a bias divider, but I found with 5v at the drain, the ones I bought from mouser had much nore gain and drew less current with only 1v of gate bias so I changed the 6.8k to 10k and the 4.7k to 2.2k. Had to bias the bc848 seperately too. It works great now.

When listening to a received tone on the radio (uniden washington 8719), probing the 10.24MHz crystal pins on the PLL (8 and 9) only cause a 20Hz change in received tone, so if thats all it pulls a crystal, this probe shouldn't load down the output of anything.