Shop OBEX P1 Docs P2 Docs Learn Events
Double sided but not... [Now a review of my new PCB Fab in a Box kit! pg 2] — Parallax Forums

Double sided but not... [Now a review of my new PCB Fab in a Box kit! pg 2]

photomankcphotomankc Posts: 943
edited 2010-01-11 03:05 in General Discussion
I'm working on a board for my thermostat project that will actually be a stack-on module to add to the Propeller Platform. When I investigated the cost to have it run I decided that I would probably prefer to get the PCB fab in a box kit and run my own. If I ask for the kit for Christmas and buy the laminator myself I'm looking at the same cost really.

Anyway one problem came to mind last night as I was looking over the board. I was routing in ways that were perfectly great on plated-through holes but that were going to be a problem without that. There are two difficult items. One is the DIP sockets. I have connections top and bottom so that rules out the cheapo CrapShack style. I was thinking that machine pin types might let me get the top side soldered as well but I'm not sure. I could get around that by soldering the chips directly to the board but the SSRs were expensive little guys and I'll never get them back off if they are soldered on both sides. Next is the pin headers that will carry my buttons, LEDs, and LCD to the top surface of the enclosure. I also very much need to be able to solder both sides of those or I'm looking at totally ripping up and starting over. I have added a few SMT parts that I plan to hand solder looking to help reduce the number of holes to drill so I really want to avoid adding vias everywhere to solder wire between layers.

Is there a way most people handle getting parts like that soldered on both sides? I was thinking I could use some wire under the parts to hold them up just high enough to get my small tip under them and reach the pads and hopefully not melt everything in the process. Solder paste comes to mind but having never worked with it it would mean some learning and at least a toaster oven and timer as I understand it. I would think that would still need something to hold the through-hole parts up so they don't drop right down into the solder.

Post Edited (photomankc) : 12/31/2009 11:01:20 PM GMT
«1

Comments

  • ElectricAyeElectricAye Posts: 4,561
    edited 2009-11-10 14:59
    It's not clear to me what you're trying to do, but if all you need to make are a few PCBs, then take a look at expressPCB or something similar. Their software for making schematics and PCBs is all free (but then you're stuck using their service). If you're not planning on going into full scale production anytime soon, I think this would be a reasonable route for prototyping and testing a few boards. They make double-sided boards and the holes will be drilled and plated-through for you. For extra money, they also will apply solder mask, which should make it a lot easier to solder SMTs, etc. I think they also do multi-layer boards. I'm a newbie at electronics and, for me, this sort of service was revolutionary. There are other similar services available, so other forum members can tell you about those.

    www.expresspcb.com

    If you decide to make your own PCBs at home, think about what you are going to do with all the nasty chemical waste you must dispose of.


    hope that helps,
    Mark
  • photomankcphotomankc Posts: 943
    edited 2009-11-10 15:58
    I'm aware of ExpressPCB but this is a one-off design as will be most of my stuff. I don't need three boards of each design and I really don't want to spend upwards of $70 to $90 per project with shipping. For what it will cost me to run 2 designs I can make them at home knowing there are some limitations. What I ran into was the fact that I wasn't thinking of the limitations of a double-sided board without plating as I was laying out the board. Where I have IC sockets and pin headers I have connecting traces on top and bottom layers attached to legs of the ICs and the headers. So I either have to start over or come up with a way to get at least the IC sockets soldered to the top and bottom pads of the board.

    I keep looking at the hobby PCB houses but in almost every case they double the cost of the whole project if not triple it. I'd rather point to point wire it in that case and design a prototype board for the Prop Platform that suites my needs better than the present version on Gadget Gangster. At least there the other 2 boards would be useful where for this they would simply be waste.
  • ElectricAyeElectricAye Posts: 4,561
    edited 2009-11-10 16:08
    photomankc said...
    ...Where I have IC sockets and pin headers I have connecting traces on top and bottom layers attached to legs of the ICs and the headers. So I either have to start over or come up with a way to get at least the IC sockets soldered to the top and bottom pads of the board....

    I'm still confused. Are you saying that the legs of your IC sockets should have gone through the board so they would be solderable to traces placed on both the top and bottom of your board?
  • kwinnkwinn Posts: 8,697
    edited 2009-11-10 16:20
    The simplest way I have found to do this is to mount the through hole components on one side (top) and the SMT's on the other (bottom). With careful layout and the use of the top side mainly as a ground plane only a few traces need to be run on the top.
  • ElectricAyeElectricAye Posts: 4,561
    edited 2009-11-10 16:28
    photomankc said...
    ...For what it will cost me to run 2 designs I can make them at home knowing there are some limitations. ...

    Keep in mind that, depending on the size of your projects, you might be able to design 2 or 3 separate projects onto a single PCB and only use the traces you need. So even though you'd have 3 boards of identical design, they could be made quite versatile if you plan ahead.
  • photomankcphotomankc Posts: 943
    edited 2009-11-10 16:48
    I could see that on some stuff that I mght do later in robotics where i will have repeated designs for sensors and hints like that.
  • skylightskylight Posts: 1,915
    edited 2009-11-10 19:21
    put something under to stand off the holders so that you can solder top and bottom so that the legs make the thru connection
  • photomankcphotomankc Posts: 943
    edited 2009-11-10 20:06
    That's what I was thinking but I was wondering if maybe some solder-paste would make it easier. Someday I'll probably be doing the toaster oven method anyway.
  • WBA ConsultingWBA Consulting Posts: 2,935
    edited 2009-11-10 21:32
    This thread is basically the same topic:
    http://forums.parallax.com/showthread.php?p=827563

    You can create "vias" by using a piece of wire and solder on top and bottom as needed. Since it is just wire, you have access to solder both sides

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Andrew Williams
    WBA Consulting
    WBA-TH1M Sensirion SHT11 Module
    Special Olympics Polar Bear Plunge, Mar 20, 2010
    314 x 196 - 7K
  • photomankcphotomankc Posts: 943
    edited 2009-11-10 22:10
    No I get that. It's getting that kind of via on the legs of the IC socket and dual row headers that I'm going for. I don't want to add extra vias for them to join them up top and bottom with wire. It's that much more drilling and there's already a lot. I'll just have to try it out and see how feasible it is for me to get to them with my iron. Otherwise I'm just going to have to start over.
  • BradCBradC Posts: 2,601
    edited 2009-11-11 01:27
    I have a fab in a box here. I've just been waiting for the little Proxxon TBM220 drill press to arrive (it arrived yesterday), so this weekend I plan on making my first through hole board to try precisely this.

    Many years ago when we used to make our own boards, we used to use machined pin sockets and solder them both sizes. I suspect it will work, but it's been a lot of years, so I need to try it and find out.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    If you always do what you always did, you always get what you always got.
  • photomankcphotomankc Posts: 943
    edited 2009-11-11 03:01
    BradC said...
    I have a fab in a box here. I've just been waiting for the little Proxxon TBM220 drill press to arrive (it arrived yesterday), so this weekend I plan on making my first through hole board to try precisely this.

    Many years ago when we used to make our own boards, we used to use machined pin sockets and solder them both sizes. I suspect it will work, but it's been a lot of years, so I need to try it and find out.

    Oh yes, I would be most interested to hear how Pulsar's kit·works out!·· I'm planning to just use my little dremel press for the near future but if it works out well I will invest in a good one.· I figured the machine pins should work for the socket.· I know I have soldered those on both sides of protoboards.· I'm going to give it a try with some regular pin headers tonight and see if I can do it.


    ETA:·Well, it's doable.· Not something I would want to do very often but it's workable in a pinch.· I would probably need to do the headers first since the angle needed to get under them is so low.· Any parts on the edges will interfere.

    Post Edited (photomankc) : 11/11/2009 3:34:54 AM GMT
    640 x 427 - 51K
  • Erik FriesenErik Friesen Posts: 1,071
    edited 2009-11-11 13:56
    Solder paste won't work with through hole stuff, at least the way I tried it. The problem is that the heat does not flow from top to bottom the way it will with liquid solder.

    As far a pulsar's stuff, you will need a laminator to get the stuff to stick to the board if I don't miss my guess.

    It would be easier to design the boards combined, and send them off. If you can wait, A place like Ourpcb will get a few boards to your door for around $100 in less than 2 weeks.
  • dMajodMajo Posts: 855
    edited 2009-11-11 14:16
    This kind of sockets will do the job: you can solder them on both sides.



    BTW I use only this kind: they are much better, secure contacts

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    · Propeller Object Exchange (last Publications / Updates);·· Vaati's custom search
    640 x 640 - 94K
  • BradCBradC Posts: 2,601
    edited 2009-11-11 14:34
    Erik Friesen said...
    Solder paste won't work with through hole stuff, at least the way I tried it. The problem is that the heat does not flow from top to bottom the way it will with liquid solder.

    As far a pulsar's stuff, you will need a laminator to get the stuff to stick to the board if I don't miss my guess.

    It would be easier to design the boards combined, and send them off. If you can wait, A place like Ourpcb will get a few boards to your door for around $100 in less than 2 weeks.

    Well, you *can* do it with an iron and bit of doweling, but I do have the recommended laminator. Will see how it comes out.

    @dMajo, those sockets are always the ones to use. When I was an apprentice I spent about 50% of my time removing the cheap fold-leaf sockets and replacing them with spun machined pin.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    If you always do what you always did, you always get what you always got.
  • dMajodMajo Posts: 855
    edited 2009-11-11 17:48
    eod_punk said...
    I've been using the 3M sockets from digikey and have been happy with them, I'd be willing to try those out next time I order some do you have a part number, manufature number or any other info to help find them in the sea of sockets?

    @BradC I also got the laminator from Pulsar, I'm sure its not going to be the prettiest the first few times, but I do have high hopes for it. Pulsar also has a howto on using the Permatex for thru hole plating.

    Per visualizzare i prezzi in tempo reale, fare clic sul link del prezzo unitario o del numero di articolo Digi-Key.



    TABLE WAS SUPPOSED TO GO HERE



    11:40:02 11/11/2009 - Help con questo schermo

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    · Propeller Object Exchange (last Publications / Updates);·· Vaati's custom search
  • BradCBradC Posts: 2,601
    edited 2009-11-18 10:27
    Just a quick Fab in a box update. I did not get to it on the weekend, so I had a bit of a crack at it today.

    I obviously have some problems with my process. I got the first side of the board to transfer and etch pretty cleanly. I etched using a small piece of sponge soaked in hot FeCl and the etch was quick, easy and clean. The bottom side I'm having some serious issues with. I get incredibly clean toner transfers, but I've been having problems applying the TRF and having it smudge the tracks.

    The board I'm working on is about 4" x 3", and so far I've used 3 Letter sized sheets of transfer paper, so I've got some process refinement to do yet.

    I've got some ideas, but I ran out of time today. I've taken a boatload of photos, but they are pretty much all of failed bits of process. I'll keep at it though.

    I must admit, the results are getting better, but I'm obviously still not doing something quite right.

    One thing of note. I bought a second hand Laserjet 2200D to do this with as my colour LJ2550 does not do suitable blacks. The LJ2200D had a re-filled toner cartridge in it, and my last couple of attempts were after putting a new genuine HP cartridge in it. That improved the toner transfer results considerably.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    If you always do what you always did, you always get what you always got.
  • dMajodMajo Posts: 855
    edited 2009-11-18 14:11
    @Brad: My experience cames from high school where the first years we was etching pcb in school laboratory (the last years we was used to send the pcb design through modem to fabs and have it done in 10 days professionally: yes, one pcb per student per month and yes all paid by school) and then evolved during the years at home.

    At the beginning (from school) I was used to put the pcb in the liquid so you have to go in with hands to take it out and see it it's ok. Now I use a teflon box fixed on a heated vibrator plate (t=50/55°C, liquid level 3 to 4 cm). I am etching the pcb side per side. You put the pcb at 30/45° touching the liquid on one edge and slowly layng down so that there is no air between the pcb and the liquid (the pcb is floating over the liquid) and then I turn on the vibrator. For single sided boards you see from the outside when the process is compleded because the pcb change the color where the copper on the other side is removed. For double sided boards just have a look and when is done repeat the process for the other side. Doing this way I never had problems also on the thinner traces.

    Usually I use precoated boards, for smaller simple projects/tests I use positiv20 photoresist spray (I have a lot of 50x50cm and 100x25cm double layer pannels) and UV lamp. As etching moisture I usually use ferric chloride because of easier handling (I am able to buy also already liquid) and greater efficiency, when I have much parts and time is an issue I use hydrogenperoxide+hydrochloric (muriatic) acid [noparse][[/noparse]parts:2/5/5h2o] (much quicker etching but saturates much before)

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    · Propeller Object Exchange (last Publications / Updates);·· Vaati's custom search

    Post Edited (dMajo) : 11/18/2009 2:27:48 PM GMT
  • LeonLeon Posts: 7,620
    edited 2009-11-18 15:16
    I do my etching in the kitchen sink with the etchant (feCl3) and board in a plastic food container, in an old washing-up bowl with about 1" of boiling water in it. Etching takes about 5 minutes with manual agitation. I wear rubber gloves, of course.

    I use pre-coated boards, and a home-made UV exposure unit. Artwork is printed on JetStar Premium film with a cheap HP DeskJet printer.

    Leon

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Amateur radio callsign: G1HSM

    Post Edited (Leon) : 11/18/2009 3:21:09 PM GMT
  • BradCBradC Posts: 2,601
    edited 2009-11-19 01:09
    In my case the etching worked perfectly. Less than 120 seconds and about 35ML of FeCl. Further investigation last night shows my issue is laminating the TRF onto the toner image. I'm stretching the film slightly as I apply it and as it springs back it smears the image. I'll have another couple of goes at it in the next couple of days. I have seen people do 6/6 mil reliably with this product so I know it will work, it's just a bit fussy and I'm a bit green [noparse]:)[/noparse]

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    If you always do what you always did, you always get what you always got.
  • LeonLeon Posts: 7,620
    edited 2009-11-19 09:49
    Heres's a little PCB for the Propeller that I made last night:

    www.leonheller.com/Propeller/XProp/XProp_photo.gif

    Tracks are 10 mil, I can go down to 8 mil if necessary. I have actually done 5 mil as an experiment on a test board.

    Leon

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Amateur radio callsign: G1HSM

    Post Edited (Leon) : 1/3/2010 10:30:59 PM GMT
  • BradCBradC Posts: 2,601
    edited 2009-11-19 09:52
    Ok, I had another crack at it this afternoon. The crux of the problem is my laminator is not getting hot enough. I measure about 130 Degrees C, which is about the softening point of toner, so I had to run the board through the laminator about 10 times to really get it hot enough. I've found the temperature regulator on the laminator, so over the weekend I'll see if I can wind it up to about 170 Degrees C. That is the temperature the guy at Ultrakeet.com uses on his boards and he regularly gets 6/6 accuracy with this stuff.

    It also turns out I *was* overstretching the TRF, and I still did a tiny bit right at the very end of the board. Next time I think I know how to solve that one. I ended up breaking one track when I got the knife out to scrape a little bit of the toner away as the smudged TRF had created a short.

    Both sides of this board were etched with the same 35mL of Ferric Chloride, heated to about 40 Degrees C and applied with 1/8th of a cut up dish sponge. I held the board in one hand and just wiped it slowly with the sponge in the other. Obviously I was not concentrating around the edges as I've left copper on all 4 corners, but those would get trimmed off if I was going to actually use this board.

    Unfortunately my registration was off far enough that the board is unusable as I can't reliably solder the top layer where I need to. I'm gonna have to work on that.

    The top looks dirty because I used an 8/20 mil crosshatch grid and the TRF was badly smudged. For the bottom I tried solid copper pours, and with plenty of heat in the laminator and real HP toner it was a *lot* better.

    The Proxxon drill and carbide bits were worth every cent. It makes it incredibly easy and accurate to drill the board.

    Minimum track size here is 10mil, most are between 12 -> 20 and clearances were set to 12.5mil. Even though the board is pretty much toast, I'm not incredibly disappointed for a first ever board.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    If you always do what you always did, you always get what you always got.
    2460 x 1832 - 1M
    2448 x 1813 - 944K
  • BradCBradC Posts: 2,601
    edited 2009-11-23 09:41
    Ok, a few tweaks to the process and we have what I consider a reasonably good result.

    I made some tweaks to the layout to increase the clearances, and re-footprinted all the components to give them bigger pads (just to make drilling easier for me). Now I've got the process sorted I can probably live with the old layout, but this one will make the DIP's easier to solder on both sides.

    The biggest tweak was to the thermostat in the laminator. I was getting about 130 Degrees C and having trouble getting the toner to stick properly on the first pass. Now it runs at 180 Degrees C and only *needs* one pass.

    I solved the double-sided alignment problems by drilling out the alignment holes to 1mm and sticking PCB pins in them. I then just laid the top toner paper on and taped the edges, flipped it over and laid the bottom toner paper on and taped its edges. Pulled the pins out and whacked it through the laminator. I drilled the alignment holes in the board and both toner transfers with my little drill press, so the holes were very clean and pretty close. Check out the through-board photo for a look at the alignment.

    I've still got some learning to do, as I did have a jaggy edge on one track. I'm quite confident of 10/10 mil spacings now, and I suspect if I'm not asking the printer to lay down huge slabs of toner it will cope much better with the finer tracks. It was having trouble keeping the density up with the earth planes. Now I've got the procedure right I'll try a hatch pour next time to reduce the amount of toner laid down.

    This one took me a little over an hour from sitting down at the desk to scrubbing the board clean and disposing of the etching gear. It will probably take me quite a while to drill though [noparse];)[/noparse]

    Overall I'd give the Fab-in-a-Box kit the thumbs up. I'm quite happy with it and now I've got the process working I can start having a bit more fun prototyping.

    <edit> Drill took 15 minutes including 3 bit changes.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    If you always do what you always did, you always get what you always got.

    Post Edited (BradC) : 11/23/2009 12:24:16 PM GMT
    1024 x 743 - 146K
    1024 x 694 - 166K
    2464 x 1480 - 345K
    2300 x 1756 - 640K
  • LeonLeon Posts: 7,620
    edited 2009-11-23 14:11
    People using laminators on the HomebrewPCB group generally seem to use multiple passes. Some of them even get results approaching those I get with my photo-etch process. smile.gif

    Leon

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Amateur radio callsign: G1HSM
  • BradCBradC Posts: 2,601
    edited 2009-11-23 14:26
    Leon said...
    People using laminators on the HomebrewPCB group generally seem to use multiple passes. Some of them even get results approaching those I get with my photo-etch process. smile.gif

    I'm sure they do. I had to use multiple passes too before I modified my laminator. It would appear the more passes you make, the more you smear the edges of the toner. Cheap laminators seem to have difficulty when you ask them to exceed about 160 degrees C.

    Photo etch is my next step, when I've got the toner transfer method perfected and I need to go smaller than about 8/8 mil. Remember, this is the first board I've ever done at home, so I've a bit to go yet. I'm sure when I have as many years experience as you do under my belt, I'll have ascended to the next level [noparse];)[/noparse]

    You also have to note, I live in the most remote capital city on the planet. It's not incredibly easy to get access to affordable pre-sensitized board. Plenty of good quality stuff you need to pawn a kidney to buy, and some very affordable stuff of incredibly questionable quality. The rest costs a fortune to get shipped in. If I can make the toner transfer method work for me I'll save a fortune using cheap board from china where I don't have to worry about the photoresist quality. The paper is cheap enough and available locally.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    If you always do what you always did, you always get what you always got.
  • BradCBradC Posts: 2,601
    edited 2009-11-24 05:36
    And to answer the original question, yes it's very, very easy to solder components on both sides where required. Radial caps can be a bit untidy though.

    Next time I'll clean the board first and maybe add a little bit of extra flux. The tip on my iron is not particularly small so there has been the odd slip to the ground plane. Nothing that shorted though, it just looks untidy.

    No need to space the dip sockets of the board, just pop the iron fairly firmly between the track and the pin and solder.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    If you always do what you always did, you always get what you always got.
    852 x 344 - 91K
  • dMajodMajo Posts: 855
    edited 2009-11-25 09:08
    BradC: a trick! If you have a suitable pcb holder/arm, so that you can have both hands free and both sides available, for sockets, you solder first the bottom side than you keep the soldering iron on to bottom side and solder wire on top side (no need to hassle with soldering iron in the small space on top)

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    · Propeller Object Exchange (last Publications / Updates);·· Vaati's custom search
  • photomankcphotomankc Posts: 943
    edited 2009-11-25 20:47
    Nice tip..... that would definitely help on the headers.
  • BradCBradC Posts: 2,601
    edited 2009-11-26 00:24
    dMajo said...
    BradC: a trick! If you have a suitable pcb holder/arm, so that you can have both hands free and both sides available, for sockets, you solder first the bottom side than you keep the soldering iron on to bottom side and solder wire on top side (no need to hassle with soldering iron in the small space on top)

    I'll give that one a go, but I have a concern as to how the pad on the top of the board gets heated without direct iron contact. I'd be worried about overheating the pin and damaging the plastic on the socket that retains the pin.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    If you always do what you always did, you always get what you always got.
  • BradCBradC Posts: 2,601
    edited 2009-11-26 00:25
    photomankc said...
    Nice tip..... that would definitely help on the headers.

    The headers were easy. Solder the bottom, slide the plastic carrier up the pins, solder the top and slide it back down again.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    If you always do what you always did, you always get what you always got.
Sign In or Register to comment.