Show us your SMD stuff.

[wavley]

^

I am surprised conductive epoxy would provide a good enough contact with low enough resistance to be usable for an RF amplifier.  It certainly does overcome the problems of excessive soldering heat.

It's actually quite great stuff =84]http://www.epotek.com/site/component/products/productdetail.html?cid[0]=84

It's really the only reliable way to assemble cryogenic chip and wire or MMIC components, a lot of components we use are only about 10 mils (the thousandth on an inch mil) square.  Because of it's relatively low frequency (we consider anything with a coax input low frequency, above that is waveguide) I could get away with a PCB design, but usually we glue components straight down to a gold plated chassis or CuFlon substrate with epoxy and wire bond the components.  I've built amps up to 120 GHz, above that the SIS Mixer guys come in.  Recently I've mostly been prototyping either low frequency (1-2 GHz) Phased Array Feed stuff, new ways of phase stable local oscillator distribution, some photonics stuff, and some 60-90 GHz amps.  When I first started 8 years ago I was just in low noise amp production for the VLA in New Mexico, but now I'm really enjoying R&D.  You'd be surprised how much my audio experience comes into play.

A lot of information about my lab is here: https://science.nrao.edu/facilities/cdl

[Cozybuilder]

Just wondering if this conductive epoxy will work as well with copper PCBs? It might open up a whole new way of building for those of us still in through-hole land.

[wavley]

Just wondering if this conductive epoxy will work as well with copper PCBs? It might open up a whole new way of building for those of us still in through-hole land.

The stuff is great and easy to work with once you get used to dabbing it on with a pin vice (I have needles bent, flattened, and sharpened all sorts of ways to get down into RF channels, but for regular PCB work a sewing needle or syringe would be just fine, you can even stencil it on for big operations)  I find it's a lot easier to epoxy some things because you can put the put the epoxy on the pad and drop the components in with tweezers and make sure it's settled in.  For me it's better than tweezers in one hand, soldering iron in the other.

Keep in mind, there is no cleaning action of the flux so whatever you're working on needs to be very clean, the Enepig board in the picture doesn't oxidize and takes solder, epoxy, gold wire, and aluminum wire equally as well which makes it great to work on.  When I use it on a copper board I make sure it's nice and shiny, then I keep a ceramic IC next to my microscope and I scrub all the tinned components on it before I place them, gold plated stuff is fine the way it is.

The other part to remember is that it shears pretty easy, if you have a nice stable substrate it will cool down to 15 Kelvin hundreds of times, no problem, but if you are going to flex or bend it at all it will break which is why my power connector is soldered and then glued with non-conductive epoxy.

And lastly, you need an oven or hot plate to cure it properly, otherwise you will have to wait a week for it to cure at room temperature, generally I do 20 minutes at 120C.   

[Cozybuilder]

Thank you,very informative 

I think this is something worth pursuing.

[alexanderbrevig]

Not a guitar pedal, but it's definitely SMT:


Several 0402 parts, I'll think twice before using that for hand assembly again.

[pickdropper]

Very nicely soldered.

I'm surprised you found the 0402 resistors more difficult than the higher pin count chips.

0402s are fairly easy to get used to.  0201s on the other hand are really really small.  I haven't tried 01005s yet.

[Ice-9]

Not a guitar pedal, but it's definitely SMT:


Several 0402 parts, I'll think twice before using that for hand assembly again.

Great job on the hand soldering, I agree that I find it easier to solder the IC's over the ting resistors. Tack one corner then the opposite and the rest of the pins are easy.

Next time I get a PCB made with that much SMD though I think it worth while getting a solder paste stencil included. 

Nice job.

[pickdropper]

Next time I get a PCB made with that much SMD though I think it worth while getting a solder paste stencil included. 

I think I'm weird.  I actually find it more difficult to apply solder paste evenly (with or without a stencil) than to solder by hand.  Reflowing does save a good chunk of time, however.

Obviously, if there are BGAs or QFNs or components with thermal pads, then hand soldering isn't an option.

[duck_arse]

ONLY for rough-hacking, slash-and-burn clots like me, or if you get depressed by winter:

shown is not really an smd build, but for build helping. it's my new heat-soaker, boasting a MASSIVE 1.2 WATTS OF PURE ... erm .... warmth. it consists of a 7805 reg and 47R load resistor (shoulda used that 10R) mounted under an intel-style pentium heatsink, all clamped up in a lump of wood.




turn it over and you see how it [is intended to] works. double sided tape yr board to the black, tastefully arrange yr parts about the place, place another heatsink on top and do something else for 20 minutes. when you come back, the parts and board have some preheat/soak to them, so you spend less iron-time soldering them.



it even has some bishops graphics taping-tape around the edges to stop runaways. it does work/warm, but is yet to be used in anger. it could have temp-sense/comparator/mosfet controller added (greaser_au - different, but same principle), but switching it off if it gets too hot will work as well.

[edit :] oops.

[alanp]

Mutable Instruments Elements. Another great Olivier Gilette design!

The biggest issue I had was getting the firmware to upload, until I realised that the TX and RX wires were crossed.

[Les Turnbull]

This conductive epoxy is pricey , any preference between the silver or carbon epoxy . http://uk.rs-online.com/web/c/adhesives-sealants-tapes/adhesives-glues/conductive-adhesives//?redirect-relevancy-data=636F3D3126696E3D4931384E4C446573635461786F6E6F6D794272616E645365617263685465726D325F74656D70266C753D656E266D6D3D6D617463687061727469616C6D617826706D3D5E5B5C707B4C7D5C707B4E647D3F5C707B5A737D2D2C2F255C2E5D2B2426706F3D31313326736E3D592673723D52656469726563742673743D4B4559574F52445F4D554C54495F414C5048415F414E445F4D554C54495F414C5048415F4E554D455249432673633D592677633D4E4F4E45267573743D636F6E647563746976652065706F7879267374613D636F6E647563746976652065706F787926&r=f#applied-dimensions=4294429411&sra=oss

[vigilante397]

I've decided I refuse to let this thread die

This is just a prototype (please excuse the mess, it's been hacked up and re-soldered a bit), but it's my first stompbox-related design using SMD. It's a programmable (4/8/16 ohm) dummy load with a speaker simulator on the end because I want to crank my amp in the middle of the night without waking up the kids The dummy load is rated for 64W, but gets a little warm Still a prototype, so I'm working on kicking it to 100W, adding a balanced headphone out, and adding a speaker passthrough so you can use a cab and the emulator at the same time if you want. It's going to be neat

[alanp]

Vigilante, how does that PCB sink 64W of heat? Or is there an offboard thing I'm missing?

[vigilante397]

Vigilante, how does that PCB sink 64W of heat?

The twelve larger package (2512) resistors around the switch are power resistors. They're rated for 16W each and I have four in parallel for each load. In open air these do fine without a heatsink, but I'm not about to throw it in a box like this I still have some bugs to work out on this prototype, but the final version will definitely have a heatsink.

[mth5044]

THOSE are rated for 16W? Now that's impressive.

[vigilante397]

THOSE are rated for 16W? Now that's impressive.

I know right?!?! I had used some for a project at work that required them to operate up to 320 degrees Fahrenheit and was impressed by how well they performed so I decided to try them out as a dummy load and they've been doing great. Most of my testing has been with an 18W amp, but I've tried them up to 45W. They get very warm, but no failures so far.

[thermionix]

THOSE are rated for 16W? Now that's impressive.

+1!  What are they made of?

[EBK]

At full power, wouldn't they be a bit like a hot soldering iron tip? 

[vigilante397]

At full power, wouldn't they be a bit like a hot soldering iron tip? 

They don't quite get that hot I was doing some testing with them at work today and I had a bunch in parallel inside a box taking 14W each, and they got to 170 celsius (338 in freedom units), not quite enough to melt solder, but definitely enough to take off your fingerprints!

What are they made of?

I think they're metal film. They're from Susumu, the CPA series: http://www.susumu.co.jp/common/pdf/CPA.pdf?v=20170706 They're available at Mouser and Digikey, but at about $3 per resistor they're the most expensive thing on that board. Part of why I want to go to 100W wire-wound resistors is so just my dummy load alone doesn't cost me $36

[thermionix]

some testing with them at work today

Oh I see, alien technology recovered from a crash site.

enough to take off your fingerprints

DB Cooper style.    (<--actually DB Cooper)

I think they're metal film

Pretty crazy they seem no bigger than 1/4W metal films.