Skip to main content
Topic: Really universal soldering controller (Read 818300 times) previous topic - next topic

Re: Really universal soldering controller

Reply #4126
Did you try the last V1.33? My intention was to be used inside station case , as it is in mine case.
Sorry... i did not try it. It's too big for my holder... and i did not want to add it inside the main case, since i wanted to maintain a kind of modular construction to be able to switch to a totally different holder with maybe an optical sensor... if a new tool needs one (i have at least 3 fully equipped and programmed optical sensor pcbs still laying around here somewhere).

Re: Really universal soldering controller

Reply #4127
I don't have components connected to GND at input except for ESD protection.

Yes you have. First you have the same diode in the chip as I do outside, and second, that 200k resistor at your input must be rated at 2KV, just as on mine, by definition. The fact that in real life plain 0805 or 1206 resistor works does not change this in any way.

Re: Really universal soldering controller

Reply #4128
I still look forward to sparkybg's MCU solution

It won't be any time soon, although the code should be pretty much the same as on my optical sensor board. The only difference will be that I will use internal comparators in the MCU instead of sending a signal to the IR transmitter and checking if it is received by the receiver, as now I do.

But given the fact that both mine an minkok's pure analog versions work, I don't see much reason of doing something else soon.

Re: Really universal soldering controller

Reply #4129
It won't be any time soon, although the code should be pretty much the same as on my optical sensor board. The only difference will be that I will use internal comparators in the MCU instead of sending a signal to the IR transmitter and checking if it is received by the receiver, as now I do.

But given the fact that both mine an minkok's pure analog versions work, I don't see much reason of doing something else soon.

Understandably, at the moment, they work steadily, and I just like to try different things :)

Re: Really universal soldering controller

Reply #4130
"Quote from: baihui on Today at 12:54:29 pm
I want to say that I have a pile of 4001 thrown into the garbage can, because I have a lot of this project, and I also have many friends who use 4001. It has problems among most people, but only a few people can work
You can just "add a C at input".  Who's care! ... What I mean is, Do not do that !!! There is latest revision."
And what I am continuing to ask is why???
That wasn't pointed to you, but anyway look at its sch and don't ask anymore.

Re: Really universal soldering controller

Reply #4131
Yes you have. First you have the same diode in the chip as I do outside, and second, that 200k resistor at your input must be rated at 2KV, just as on mine, by definition. The fact that in real life plain 0805 or 1206 resistor works does not change this in any way.
Is it really 4148 the same as in to gates and it is used for ESD protection?

Re: Really universal soldering controller

Reply #4132
4148 is larger then modern device internal diodes. I don't know about 4001 - it may be smaller or larger. But it is similar enough. It handles larger current for sure, bit this is more due to it's package than anything else.

As I said, it is there on my schematics just because LM339 does not have internal one to ground. It does not affect the circuit performance at all, or at least to a degree that makes it's influence irrelevant. 4148 has around 10nA leakage at room temperature and around 4pF capacitance.

BAV199 can be used with even better leakage if this is a concern, but it is not - LM339 has around 30nA bias+leakage current already, and 10 megaohms resistor gives 330nA current - the circuit leakage is a decade less. And all this leakage can be easily compensated changing the reference voltage if needed.

Re: Really universal soldering controller

Reply #4133
I did a non scientific heat generation test of the new 4 layer back board equipped with STD40P8F6AG p-channel fets compared to the 2 layer board equipped with SQD50P08-25L. I can feel a slight difference after one minute at full power, with the STD40P8F6AG being a little bit cooler, but both fets do not need additional cooling. I will put a 40x40mm heatsink behind the FERD20H100 rectifiers though, those get too hot to touch, so it's nice that they are now surface mount devices so they can be cooled easily

Re: Really universal soldering controller

Reply #4134
You may put a heatsink on top of them too. In fact I am pretty curious which solution would work better - in one case you have a PCB between their back and heatsink, in the other you have a body plastic/ceramic between the die and the heatsink.

P.S.: I found a handbook with these experiments. It appears that top-side cooling is a pretty good solution, sometimes even better than bottom side PCB cooling. You may try this if you want. I will try it for sure when i get an assembled back board.

Although, the TVS can be a problem making this, if it is higher than DPAK.

P.P.S: Thecked this out - SMC package of the TVS has almost the same height as DPAK, so it may , or may not be a problem.

Re: Really universal soldering controller

Reply #4135
You may put a heatsink on top of them too. In fact I am pretty curious which solution would work better - in one case you have a PCB between their back and heatsink, in the other you have a body plastic/ceramic between the die and the heatsink.

P.S.: I found a handbook with these experiments. It appears that top-side cooling is a pretty good solution, sometimes even better than bottom side PCB cooling. You may try this if you want. I will try it for sure when i get an assembled back board.

Although, the TVS can be a problem making this, if it is higher than DPAK.

P.P.S: Thecked this out - SMC package of the TVS has almost the same height as DPAK, so it may , or may not be a problem.

Height of both packages looks the same by eye, there's not much in it for sure:

Back board

I believe it will work with self adhesive heat sink tape, I will try it out when the heatsinks arrive

Re: Really universal soldering controller

Reply #4136
Self adhesive will work on the back for sure. I am not sure about the front. You may have to reheat the DPAKs and press them to the PCB. A small difference in solder/solder paste quantity may give a small difference in height of different parts. I always press them ehile the solder is still liquid when assembling by hand, but I don't know what can happen when assembled in oven with solder paste.

Re: Really universal soldering controller

Reply #4137
Self adhesive will work on the back for sure. I am not sure about the front. You may have to reheat the DPAKs and press them to the PCB. A small difference in solder/solder paste quantity may give a small difference in height of different parts. I always press them ehile the solder is still liquid when assembling by hand, but I don't know what can happen when assembled in oven with solder paste.

I put a dab of flux on the solder pad before placing the component. I then heat he dpaks from the rear tab and feed solder to the edge of the solder pad and the tab while pressing down on the component with tweezers. This way, solder will wick into the joint and the components are completely flat and even on the board surface. I hope the adhesive will work. What other method would there be to secure the heatsink?

Re: Really universal soldering controller

Reply #4138
Pretty the same of what I do. I sometimes put a spot of solder paste before doing it. But anyway, i press down the DPAK waiting for solder it to solidify, although I never used heatsink on the top so far

There are more thick materials, but they are usually not adhesive, but needs to be pressed by a spring or something. There are some thermo glues with silicone-like consistency after solidifying. I bought some from TME some time ago.  Thay also can be unglued if needed, just like silicone.

This is pretty much everything I've used so far, excluding thermopastes.

If you pressed the package until the solder becomes solid, this is enough for same parts form the same manufacturer to be at same level, so I think the self adhesive tape should work well enough.

...and after all, the top-side heatsink may not be needed at all. Even the back heatsink may not be needed at all. 80 degrees celcius is safe enough temperature, and I stop being able to touch them already at 60 degrees, so they may not be that hot after all, although I definitely don't like anything so hot that I cannot hold my finger on it for extended periods.

And another solution is a small heatsink on each component, but we loose heatsink surface this way. This is pretty common for small FPGAs when they get hot. For example 10x10mm heatsink is not that rare thing to see. Although I never thought about their performance nor analysed it.

 

Re: Really universal soldering controller

Reply #4139
What other method would there be to secure the heatsink?
I used alumina/boron-nitride/epoxy-based 2k thermal glue in the past. With great success. I even dare to say: with too much success. Because i had a heatsink glued to an ADC ic in the past and the ADC failed and had to be replaced. Just out of couriosity i wanted to remove the heatsink. I tried brute force with a craft knife and even cranked my hot air soldering station up to 450°C in my attempt to remove it from the ic with lots of heat. All without any success :o .

So that kind of 2k thermal glue seems like a pretty reliable solution for mounting a heatsink... except when you are planing to remove the heatsink in the future ;) ...