recommed me a leveling agent for copper plating

[T3sl4co1l]

Negative, like I said, reversal. AFAIK just a polarity switch is fine. But maybe varying current density is something to try.

Already said frequency and duty.

Tim

[coppercone2]

Its probobly like welding where you set offset and balance. Both have to do with transfering something.

my wavetek has been consistently turning on after some resocketing and cleaning so I think it might be a go

I have been thinking about the magnitudes. Aggressive plating is a thing, I don't know if agressive cleaning is a thing.

Given the fact that electrocleaning solutions are totally different then electroplating solutions, none of the information I have really crosses over at all.

The high throw solution can get things into a hole. Having it go into the hole is very much wanted, for filling voids and such. You don't want those copper ions to come out. I am going to make a guess and say that 10% reduction in cleaning current vs plating current might help keep copper in holes... you want the bumps to come off but the holes to get filled. Maybe more cleaning can get vias to prevent getting clogged or misshapen, because you basically do not care about the copper on the top when you are doing that, infact I bet the pros use two solutions, ones to get the holes finished to 90% then go to a low acid bath to get the top of the via finished without effecting the hole too much.


I wonder what tricks they have for the very reliable PCBs

[coppercone2]

Lol that wavetek is harder to set then I thought for some reason. I set it only using the analog gauges on the amplifier, I got some pulse that spend 80% of time @ plating current and 20% of time at ~20% reverse current.


The first coin looked really shiney at first, this one looks matte.


I need to get a scope down there but anyway I wonder if those dreaded bumps will show up. If it stays flat, it will still be interesting.

I can't seem to get it into the mHz, so its at 1Hz for that generator

this shit is ridiculous

its most amusing to see it go down on a +- dial with the electrons doing tug of war

[coppercone2]

Ehh its looking like etched PCB.

I need to make sure that bath is still good to see if it plates nice on DC still with a fresh coin, because if this is it, the reverse current makes the finish crappy. Maybe it  could delay or prevent bumps, which would be useful for electroforming thick parts, but this would make crappy waveguide finish.... I can't polish the inside of a apature easily if it looks like this 

I will run a test on the other power supply later to see if its the solution or if reverse current makes chalky surface finish...

All these experiments will require a control to verify the solution for a short duration plate, because with just DC it took 3 hours to get it to this texture. Again, maybe it prevents bumps. Maybe the frequency is just way too high.


did you notice if surface finish was effected by the polarity reversal?

[coppercone2]

Perhaps I trusted the gauge dial too much for current and voltage. I set the duty cycle to 50% to see what is actually going on to give the dials time to respond. I don't want my scope near that bath lol

I am trying it at a much lower interval, that is -100mA 0.05 S and 250mA 0.95s for plating a nickel

I think last night I was pulsing it like 400mA -> -300mA at 20%. so now its 250mA -> -100mA @ 5%

[coppercone2]

lmfao I figured out the problem with my coin

I thought pi*D was area of circle, not circumference.

I guess I was flooring that plate job!

I kept thinking something aint seem right with 6 square inches LOL

I think it was going like warp 10


If the results I got is for 400% of recommended current, I say they were damn fine 

[IanB]

I thought pi*D was area of circle, not circumference.

pi * D has dimensions of length. Area has dimensions of length squared. Therefore it is impossible for any formula that only has a single length dimension in it to be an area formula. Dimensional analysis is your friend.

[coppercone2]

its a good proofing test I am glad it happened

[coppercone2]

do you think using a BeCu wire to suspend stuff is a bad idea for this bath if its on the object being plated?

I can make a neat BeCu spring that can clamp to the object and then go above water to connect to power source, its combined electrical contact and mechanical support.

I don't think any Be can leach into the bath unless the polarity is wrong?

[T3sl4co1l]

Cathode? It'll get plated like anything else. Titanium is the usual stuff, which may still get plated but the metal can simply be cracked off, it won't ever stick to it.

Tim

[coppercone2]

well I tried it and it looks like its just being plated I did not want Be to be dissolving in there but I think its sealed up by now


you just wind 4 turn spring on a shaft and you can clip it on the part

like 1mm is plenty unless you need a wide spring

the whole 'cake string' setup was pissing me off. they make god damn cake box windings around the part, or expect someone to buy a 10$ titanium clip. the hell with that, its not even good conductor. Its like a proper connector for a part.

you can probobly take a thicker one and make a C spring clamp to hold a part with minimum contact too

[coppercone2]

Is there a way to make a ultra deep throw copper solution?


This one gets my horn about 1/2 depth good, the last 1/4 is really thin compared to the first 3/4.


I think you can do a final plate by fixture a copper rod in the middle of the object, but is there maybe a different approach? Say adding even more acid to lower the conductivity. Or I guess you can mask it.

Because they optimized it for a PCB thickness via. I need a wave guide.


If I could get away from having to do masking and special electrode it would be better IMO


I think I will add a drop of tarn-x to the mixture


I bent a piece of refrigerator tubing in a shape so it can clip on the side of the tank (dewalt toolbox organizer) and go down the middle of the horn. Maybe that will put some meat on the aparture cuz it has chicken legs right now.



It looks like I need to build some kind of complicated dielectric apparatus with scaffolding and such.......

[T3sl4co1l]

Well, two things: electric field, and ionic concentration.  The leveler is supposed to adsorb to the metal surface, blocking current flow particularly on prominent surfaces, thus encouraging deep and even plating.  So it changes the electric field, microscopically: in the ionic boundary layer around the cathode itself.  For this to work, you also need enough ionic concentration, especially in crevices, holes, pits.  If you aren't circulating electrolyte, you're relying on ionic diffusion, extremely slow, and so the plating current must be extremely low to allow ions to diffuse about.  With agitation, stirring, or even spray, current density can be higher, but keep in mind, any pores where fluid isn't flowing, will have poor coverage.

Reversal helps by preferentially removing prominent features, or perhaps passivating the surface (say by depositing Cu2Cl2 or whatever?), and may be necessary for the deepest-throw processes.  I don't know what the practical limits are of purely chemical vs. electrically assisted plating are.

Tim

[coppercone2]

ok the thiourea "tarn-x" addition made the antenna into almost exactly carrot color. I think it was the wrong move. 

it looks like i carved it out of a large carrot

need to run test coins


there is copper under neath, but right now its optimally camouflaged for a fruit platter or chaquita banana hat, maybe useful for cuban embassy

home depot can also treat their roof this way, perfect.

[coppercone2]

well i finished and its a bit of a fail,

10.1 mm vs 4.8 mm for a WR42


and it should be

10.7 x 4.3

I guess I screwed it up somehow ... 


so

6% x 11% error for the launcher waveguide input interior dimensions.

My print must have been scaled and too much paint etc 

I think I scaled it with a guess to what it might look like after paint and plating.


Ferengi wave guide, the kind the grand magus used to contact deep space nine.


the paint is too much of a wild card, I need to get electroless copper down, then I can just deal with plating , meaning I can periodically measure it while its going. trying to predict what that paint will do is too hard


and forget about doing a "nice" job on the waveguide launcher bit without precise fixturing and possibly masking etc. its not too pretty in there The flange is great but the ID I guess needs a special short plating pin to go in there about 1/2 inch deep , plate that and then plate the interior with a longer bit and then plate the exterior


the sharp interior corners are also difficult to polish, I used silicone rubber abrasive points, but its wonky, and the silver plating (kool amp) did not do a 100% job on that bit either, I get a bit of dark spots that are really hard to clean up. You probobly need to scrub it using a tooth pick that has diamond abrasive on it to get those "nice". Because I had to polish it and add kool-amp silver coating, the kool amp had a few inner corner spots that are black, but under it it has copper metal if you scratch it. 

I put off buying alot of the polishing tools I wanted for a long time, those will be a expensive key..


But this waveguide is outsdie my VNA capabilities. The next smaller one requires a few liters of solution to properly plate that I can try to measure

[coppercone2]

I got a <10 um woven polypropylene filter sheet to try to make anode bags from.

non woven looked like it would soak up stuff, so I opted for woven. Non woven is like felt I think, its gonna be mad heavy/damp

wet felt is like god damn metal weight lol

[coppercone2]

of course the tin chloride from the other suppliers also makes milk.

I am going to try to dissolve in the minimum amount of HCl water, filter it, then freeze crystalize


all the tin chloride sellers are trash and they make this endevour nearly impossible


just to make sure I will make the degassed water with nitrogen in the special degassing glassware and add fresh HCl and then try to dissolve it again . This problem will not hide behind some nebulous claims of oxygenated water

I heard tall tales about nitrogen tires from the shitty car dealership too. i feel like when nitrogen starts getting referenced there is usually a bunch of hooey behind it. The nitrogen tire man sold a car that needed $5000 in repairs and was still a piece of shit, and his tires were leaky, fat load of good nitrogen would do for that problem 

[coppercone2]

the third tin chloride worked to dissolve clear, this one came not as a powder but lumps.

Turns out the old palladium chloride I had from a few years ago was bad too, it dissolved clear. When I added the one from a more reputable supplier the solution turned jet black

too many bad reagents were giving me severe agita

[coppercone2]

i did the steps in that PDF with a piece of alumina, it made black stuff on it that is conductive and well adhered to the ceramic but it does not look like nice copper

[coppercone2]

I electroless plated a nylon washer.

It mostly worked but there is a few black spots. Like ink black.

The resistance is medicore (<20 ohms)

Slow electroless copper bath with dark spots – A slow copper bath exhibiting large dark areas can be a symptom of a copper content that is too low or an insufficient bath temperature.

[coppercone2]

its possible to solder to the electroless plating on ceramic. It beads, but it does adhere fairly strong, and if you rip it off it takes off the underlying copper, whatever did not melt into the solder.

[coppercone2]

I wonder if I wore out the tin chloride solution or if the copper plate is more sensitive then I thought.


There is a problem with mixing it, if you do not dissolve the lye first, it will heat up and burn the EDTA, so efforts to shake and bake the stuff are kind of fruitless, it turns into black carbon in the solution.

I feel that the glass-silver substrate bond is stronger then the electroless copper soldered bond, but neither is terribly impressive, well neither is standard PCB material, if you yank hard on a solder pad it will come off. All three have a satisfying "snap" for pull tests. I don't think that electroplating the copper on top of the electroless copper will result in stronger boards.

While silver on ceramic is interesting for ceramic PCB, so is chemical copper... this adds another dynamic to tests, to compare silver-glass alumina substrate with electroplated and etched substrate for making boards.


If you don't have thick film resistors on the PCB, is there any benefit towards using silver glass instead of just electroless copper plating, if thickened by electroplating? I wonder what will happen when I solder down parts, despite there being substantial strenght in the electroless on ceramic copper bond, it might not be enough when parts are actually soldered to the PCB, with thin traces, etc.


It does add a option though, for hybrid chemical and sintered boards, where the ground plane can be made with chemical plating and the signal layer can be made with silver ceramic. The via might have more standard RF properties if chemical plated, and the added complexity of masking and stuff makes it a bit of a moot point, even for a medium board silver is cheap enough that the time saved is worth more then a bit of silver


For a connector or turret terminals, the silver glass would make for impressive strength, but for say... 0.3mm vias stitching, I am not sure its the best choice, I suspect the "paste in hole" approach might result in a substantial amount of defects if you really stitch something   


however if the silver glass is drillable, making full vias and drilling them out might be very perfect

[coppercone2]

the blackness of the solution is not a indicator that it has enough palladium. if you do everything right for electroless copper and it looks right but its not working, add more palladium