Homebrew PCBs

Hi Brian,

This is in regards to your post on No Clean Flux.

" Noclean is hard to clean and may be conductive. I had problems in rf circuits even with my mega8 running at 8MHz. Just today I reworked 21 QFN's with water soluble flux and just swishing them in a tub of 140 degree water and drying the board they worked at 1 GHz. There was no problem washing the flux from under the part, it seems."

I am an analog design engineer an have had a number of problems in the past with "conductive flux" in high impedance circuits. In fact I did a study on different fluxes back 2001 to better understand the issue. Bottom line is that all the flux residues were conductive with the No Clean flux being one of the worst. 

What a lot of people don't understand is that No Clean flux is a "heat activated" flux that under goes a chemical change to a non conductive state when heated above a certain critical temperature. Boards that were coming out of manufacturing reflow process were fine. The problem was that the rework folks were brushing it on when reworking boards and the soldering iron only heated the lead they were working on which left the rest of unactivated and hence conductive. They would use a brush with flux try to remove it but that would only spread it around the board. The only way to remove it was to do a total immersion flush.

I finally got the rework folks to change from No Clean to a standard flux, apply it only to the lead that was being soldered and to use enough flux remover to flush the flux residue off the board into a small tray. After that things went pretty well. Btw, the flux residue on the board is highly hygroscopic and very sensitive to humidity. During the tests that I ran the impedance between two traces on the board to decrease from 1 meg to around 80k by breathing on the board.  

Also the principal industry standard for soldering flux is:  "IPC/EIA J-STD-004 - Requirements for Soldering Fluxes"


Ron

[Non-text portions of this message have been removed]

rbert99@... wrote:
> Hi Brian,
> 
> This is in regards to your post on No Clean Flux.
> 
> " Noclean is hard to clean and may be conductive. I had problems in rf circuits even with my mega8 running at 8MHz. Just today I reworked 21 QFN's with water soluble flux and just swishing them in a tub of 140 degree water and drying the board they worked at 1 GHz. There was no problem washing the flux from under the part, it seems."
> 
> I am an analog design engineer an have had a number of problems in the past with "conductive flux" in high impedance circuits. In fact I did a study on different fluxes back 2001 to better understand the issue. Bottom line is that all the flux residues were conductive with the No Clean flux being one of the worst. 
> 
> What a lot of people don't understand is that No Clean flux is a "heat activated" flux that under goes a chemical change to a non conductive state when heated above a certain critical temperature. Boards that were coming out of manufacturing reflow process were fine. The problem was that the rework folks were brushing it on when reworking boards and the soldering iron only heated the lead they were working on which left the rest of unactivated and hence conductive. They would use a brush with flux try to remove it but that would only spread it around the board. The only way to remove it was to do a total immersion flush.
> 
> I finally got the rework folks to change from No Clean to a standard flux, apply it only to the lead that was being soldered and to use enough flux remover to flush the flux residue off the board into a small tray. After that things went pretty well. Btw, the flux residue on the board is highly hygroscopic and very sensitive to humidity. During the tests that I ran the impedance between two traces on the board to decrease from 1 meg to around 80k by breathing on the board.  
> 
> Also the principal industry standard for soldering flux is:  "IPC/EIA J-STD-004 - Requirements for Soldering Fluxes"

What's wrong with rosin flux and not cleaning it? I've been doing that
for 10 years. Old valve gear works for eons. My stuff is hand-soldered
leaded components.

If you need to clean rosin flux off, there's a really good aqeuous
cleaner i've used, and it makes all the solder and other metals look
extra shiny too.

I've found another factor in the choice, too...

It's harder to clean no-clean solder paste off stencils :-P

To add to this we just made 200 RF boards that required rework, for some reason the QFN's wernt reflowing right. The solder paste templets were made from the customer gerbers.

All the rework was all done with water soluble flux as I recommended. Using a water washing system and drying provided great results since noclean is hard to clean from under these parts.

We only use noclean on boards that are coated.

Brian

--- In Homebrew_PCBs@yahoogroups.com, <rbert99@...> wrote:

>
> Hi Brian,
> 
> This is in regards to your post on No Clean Flux.
> 
> " Noclean is hard to clean and may be conductive. I had problems in rf circuits even with my mega8 running at 8MHz. Just today I reworked 21 QFN's with water soluble flux and just swishing them in a tub of 140 degree water and drying the board they worked at 1 GHz. There was no problem washing the flux from under the part, it seems."
> 
> I am an analog design engineer an have had a number of problems in the past with "conductive flux" in high impedance circuits. In fact I did a study on different fluxes back 2001 to better understand the issue. Bottom line is that all the flux residues were conductive with the No Clean flux being one of the worst. 
> 
> What a lot of people don't understand is that No Clean flux is a "heat activated" flux that under goes a chemical change to a non conductive state when heated above a certain critical temperature. Boards that were coming out of manufacturing reflow process were fine. The problem was that the rework folks were brushing it on when reworking boards and the soldering iron only heated the lead they were working on which left the rest of unactivated and hence conductive. They would use a brush with flux try to remove it but that would only spread it around the board. The only way to remove it was to do a total immersion flush.
> 
> I finally got the rework folks to change from No Clean to a standard flux, apply it only to the lead that was being soldered and to use enough flux remover to flush the flux residue off the board into a small tray. After that things went pretty well. Btw, the flux residue on the board is highly hygroscopic and very sensitive to humidity. During the tests that I ran the impedance between two traces on the board to decrease from 1 meg to around 80k by breathing on the board.  
> 
> Also the principal industry standard for soldering flux is:  "IPC/EIA J-STD-004 - Requirements for Soldering Fluxes"
> 
> 
> Ron
> 
> [Non-text portions of this message have been removed]
>

Hi Russell,

Yes based on the "Flux study" we changed to rosin flux for hand soldering. Please note however that rosin flux is still hygosopic and absorbs moisture that makes it conductive and hence sensitive to humidity. The inter trace impedance is proportional to length of the adjacent traces that the flux makes contact to, the amount of flux residue and the humidity. It is usually in the range of 400k to 1meg, so low in medium impedence designs (< 100k) it is rarely a problem. 

Also, if you are not going to clean the flux off the board completely, you are better off not cleaning it at all. If the flux is applied only to the point where the soldering is done the amount of trace that it makes contact to is small so its impedance is realitively high. If you use a brush with flux cleaner it just spreads the flux over a larger area and, although it may look better, the flux contacts more trace area and the impedance will drop noticably. Also please note that use of solder mask on the board really improves the situation since the flux can't come in contact with the etch where there is solder mask.

BTW, if anyone wants to experiment with this, take a clean blank PCB (without solder mask) and meausre the impedance between two adjunct traces (usually >1 meg) Then apply flux to a small area between two adjunct traces and when dry measure the resistance between the traces. Slowly breathe on the board to increase the effective humidity and you will see a noticable decrease in resistance (usually <1meg) . Then take a brush soaked with flux remover and "clean the board" by brushing it around between the two traces. When dry redo the above measurements ( I have seen the impedance go down to 80k with heavy breathing). Finally use the brush with enough flux remover to flush the flux off the board and redo the measurements.


Thanks

Ron
 -------------------------------------------
>What's wrong with rosin flux and not cleaning it? I've been doing that
>for 10 years. Old valve gear works for eons. My stuff is hand-soldered
>leaded components.

>If you need to clean rosin flux off, there's a really good aqeuous
>cleaner i've used, and it makes all the solder and other metals look
>extra shiny too.


[Non-text portions of this message have been removed]