So why do I want to remove no-clean solder flux from a PCB? Well, I have a non-audio PCB that will have up to 900V across an SMT-style capacitor, and I'm just not comfortable with the idea of having that slimy stuff in that area. The other side of the capacitor feeds a current-to-voltage converter with a gain of 1e6 so it won't take much to screw things up!
The reason I'm asking in this forum is that the subject may be (is?) relevant to those among us who are vacuum tube aficionados; and, in fact, my application DOES use a vacuum tube (but it has nothing to do with audio -- it's a photomultiplier tube).
The best approach most-likely is to just use a water-soluble flux but then there's the fact that I assembled my board with the no-clean stuff before I thought of this potential problem 🙄. Water sure doesn't work...
The reason I'm asking in this forum is that the subject may be (is?) relevant to those among us who are vacuum tube aficionados; and, in fact, my application DOES use a vacuum tube (but it has nothing to do with audio -- it's a photomultiplier tube).
The best approach most-likely is to just use a water-soluble flux but then there's the fact that I assembled my board with the no-clean stuff before I thought of this potential problem 🙄. Water sure doesn't work...
Hi I use isopropyl alcohol 99% (not the usual 70% as that gives stains and corrosion) for most older fluxes. I have found some recent fluxes need water and a cloth/cotton swab, these are quite hard to remove. In persistent cases I even have used a tiny bit of thinner, a toothbrush and/or cotton cloth and afterwards I redo with isopropyl alcohol 99%. I like clean boards and the how is less interesting than the result to me. Most methods require to dry the PCB with cotton cloth to avoid residue, fog like "ghosting" and/or the white corrosion of solder/PCB pads.
My approach is trial and error as the type of flux I generally meet is unknown (nor am I an expert or chemist).
https://www.chemtronics.com/flux-of...u6B65ufHC1Pbspk1wAuBfC7A-hW94QKhLXluhTyaXzd7h
My approach is trial and error as the type of flux I generally meet is unknown (nor am I an expert or chemist).
https://www.chemtronics.com/flux-of...u6B65ufHC1Pbspk1wAuBfC7A-hW94QKhLXluhTyaXzd7h
Last edited:
A first rinse with deionized water with a drop of surfactant (a generic, soft détergent) then a rinse and scrub with a mix of water +alcohol/ ether, then a rinse with isopropanol followed by a compressed air blow to remove everything.
Process can vary according to the circumstances , but it is the basic procedure: it is safe and effective
Process can vary according to the circumstances , but it is the basic procedure: it is safe and effective
Bob Pease advised the dishwasher sometimes. But what does the maker of the solder recommends? Doesn't hurt to ask them.
I don't know what it is in my life that removing flux has always been such a pain.
In the end, I discovered high percentage alcohol and a toothbrush with repeated scrubbing gets fantastic results. Repeat over and over until the flux is gone.
OR
Acetone and toothbrush and its gone in seconds.
In the end, I discovered high percentage alcohol and a toothbrush with repeated scrubbing gets fantastic results. Repeat over and over until the flux is gone.
OR
Acetone and toothbrush and its gone in seconds.
I would have thought that as the flux from SMT solder pastes is only on the pads, it stays there or becomes a thin splattering of tiny drops around each pad, i.e. not a continuous film that might allow conduction. Attempting to clean such a surface runs the risk of smearing the flux drops into a thin film whereas originally it was isolated spots?
the conductive no-clean flux sits under the part. you have to remove the part and resolder a new clean one
Why would there be solder paste under the part? Does the flux wick under parts sometimes? I guess if that happens its not going to be easy to spot.
I use a 2-stage cleaning process:
1. Isopropanol to clean most of the flux.
https://en.uraltone.com/ipa-plus-1l-bottle-very-pure-isopropanol.html
2. Ultrasonic cleaner with warm water and cleaning fluid.
https://www.amazon.de/-/en/Oregon-Ultrasonic-Cleaning-Concentrate-Bottle/dp/B07X5NK3Z6
After ultrasonic bath 2 hours baking at about 100°C.
1. Isopropanol to clean most of the flux.
https://en.uraltone.com/ipa-plus-1l-bottle-very-pure-isopropanol.html
2. Ultrasonic cleaner with warm water and cleaning fluid.
https://www.amazon.de/-/en/Oregon-Ultrasonic-Cleaning-Concentrate-Bottle/dp/B07X5NK3Z6
After ultrasonic bath 2 hours baking at about 100°C.
I once had a circuit with a capacitive voltage divider on a PIC MCU input. the divider was precharged by setting the pic pin temporarily to output. Then we did few A/D conversions of that node , and saw that the cap was drooping, caused by the "no-clean flux" . after removing the cap, cleaning the pads and surfaces, and placing a clean cap back we had no leakage anymore. This was also an effective leakage test. The circuit was used as an impedance measurement circuit for implanted electrodes, hence the dc free method.
Last edited:
Not with a package this size (U2):
The "big" SMT devices are 0805's, except for PD1 which is a photodiode. U2 is a TI JFE150. Pins adjacent to the gate are biased +/-10V so there WILL be leakage to the gate if I don't remove the flux -- the JFET is used as a TIA and has a 10M feedback resistor. I would have designed-in a guard ring but one of the pins is too close. The gate is the upper-right pin.
Hand soldering will spread flux all over the place. NOT using flux is an almost-sure way to end up with solder-shorted pins. I've been there too many times, even if doing the operation under a stereomicroscope.
Back to the no-clean flux issue, I have found that denatured alcohol does a nice job of removing the flux. I have a small ultrasonic cleaner so I put the board in a small mason jar, added the alcohol, then screwed-down the lid to keep the alcohol from evaporating. Water in the ultrasonic bowl transferred the ultrasonic energy into the jar (it's necessary to have water in the ultrasonic anyway, to protect the transducer). It worked a treat, so I'm moving on to actually assembling the board.
R5 is the 10M resistor. It's an axial because I'm going to wrap some fine wire around the body and solder the wire to ground. it's a way to short out the parasitic capacitance across the resistor, which greatly reduces the bandwidth of the TIA.
Anyway, this post was about the flux not my pie-in-the-sky circuit.
Thanks to everyone for their input!