9
\$\begingroup\$

I'm a student living in the UK.

I have this great idea for a product that fills a niche market and it interests me. It's a powerful on screen display module with datalogging capabilities, and it's open source hardware and software. See http://code.google.com/p/super-osd/.

I want to start selling these modules. I have PCB designs and I'm ready to order the components, but what route should I take for manufacturing the modules? I see three main possibilities:

  • Etch my own PCBs and solder components on - pretty much a no go as I'm dealing with double sided surface mount boards.
  • Order PCBs cheap from China and solder all 40 odd (0603, TQFP44 etc.) components on myself. (Probably the cheapest option but lots of manual labour.)
  • Get a quote for PCB+assembly.

Any ideas?

\$\endgroup\$
2
5
\$\begingroup\$

I would say there are three major factors that would need to be considered before making a decision.

  1. How many products are you planning on selling yearly? If it is a lot, say more than a hundred, you probably don't want to be doing much assembly. You need to save your time for future product development!
  2. How many surface mount components are on the board and can you reduce that number more. Yes you can solder surface mount chips in your house and there are tons of cool methods for simplifying that process but still, nothing beats the speed and precision of a machine. (If you want some tips on how to solder surface mount chips ask that in another question. I have found some very good techniques) Anyways, thru-hole components are obviously way easier to solder by hand and if you are going to be doing more than a few boards and actually selling them to customers they are a lot easier do not screw up.
  3. Do you enjoy circuit assembly. You need to consider what you like doing. If soldering boards all day gives you jollies and you have the time then maybe you should go for it. But if you are also interested in design and development of new stuff than you may not want to be tied down with your soldering iron all day.

Whatever you end up with, Good luck!

\$\endgroup\$
3
  • 1
    \$\begingroup\$ just one other note. Board house and assembly house prices drastically decrease with product volumes so it can be hard to afford them with low quantites. \$\endgroup\$ Sep 29 '10 at 19:44
  • \$\begingroup\$ 1) I am planning to sell more than 50 yearly but less than 200. 2) There are about 20x 0603 resistors, 10x 0603 caps, and various chips, regulators and a few electrolytics. 3) I don't really enjoy it too much. \$\endgroup\$
    – Thomas O
    Sep 29 '10 at 19:51
  • \$\begingroup\$ Sounds like you should outsource the assembly. Make a lot of phone calls and you can find assembly houses that work much better with small-quantity folks than the usual assembly house which is only interested in very high quantities. Look for the smaller shops. Also, ask the bigger shops if they know of other shops that work will smaller quantities. I have had good luck with that. \$\endgroup\$ Sep 30 '10 at 13:19
6
\$\begingroup\$

In terms of quality and reliability, you cannot beat machine placement on professionally manufactured PCBs.

You will most likely find that option #2 + #3 will work best - get the PCBs made offshore by a reputable low-cost shop that deals in prototype volumes (which may or may not be easy) and shop around for contract assembly places.

Have you considered sourcing the parts yourself (i.e. issuing a complete parts + boards kit to prospective builders) or are you looking for someone to supply parts as part of the service?

\$\endgroup\$
1
  • \$\begingroup\$ I have also considered making kits, and I will be selling PCBs with instructions and all components for less. \$\endgroup\$
    – Thomas O
    Sep 30 '10 at 12:19
4
\$\begingroup\$

I would start by ordering 10 .. 100 PCBs, soldering maybe 10 by hand, and start selling. Meanwhile, get a quote from an assembly house for maybe 100. From the rate at which you sell the systems (hand-soldered) you can decide on your next action. If the sales are disapointingly low you don't have lost too much money. If they are high you have soldered obnly a few yourself. Both ways you win (or don't loose too much).

\$\endgroup\$
1
\$\begingroup\$

Don't forget the marketing!

Apart from the hardware assembly and manufacturing side of your project, you may also wish to consider how to market your product earlier in the piece.

If it really turns out to be as good as you expect, having one of the review forums like hackaday or hacked gadets review the product will help you achieve critical mass a lot faster and move the product out the door.

As you are a student and investing in 50-100 built units is going to cost you a serious chunk of change. Once they are built and have them sitting on the shelf isn't going to get you a good return on investment. So you need to be just a bit clever about the way you intend to sell your product. Blogging the process, twittering, tumblr'ing will hopefully get you a bit of a following and make it just that much easier to sell some product.

Think about what support or other options you might also provide purchasers with. Building a logging device is probably one of those things which would probably benefit from regular firmware updates to improve or add features to the logging. (Bug fixes even!) Make it easier for purchasers to access this info and update their product.

\$\endgroup\$
3
  • \$\begingroup\$ In the words of Dave L. Jones, "don't even think about marketing." I am not going to be building 50-100 immediately, I will build them on a per-order basis only. My marketing comes from interested people on the 'net, such as this RCGroups thread: rcgroups.com/forums/showthread.php?t=1290963. \$\endgroup\$
    – Thomas O
    Sep 30 '10 at 12:20
  • \$\begingroup\$ Ahhh that explains why I am so wealthy and Dave L. Jones isn't (or may not be! heh). Because I do consider marketing. :-) hobby businesses can be very profitable if done properly. There is no harm at all in giving it a go, in the immortal words of Gordan Gekko.. Greed is Good. Please do keep us posted on your build. I am very interested in data loggers myself! \$\endgroup\$
    – Rob
    Sep 30 '10 at 13:02
  • \$\begingroup\$ If you're building on demand, that strongly suggests doing it yourself rather than outsourcing it, as there will be high fixed costs per order. \$\endgroup\$
    – pjc50
    Aug 28 '12 at 9:51
1
\$\begingroup\$

I would ask for a quote from a shop. Yes, it's obvious that it's going to be more expensive than doing it yourself, but at 100+ it may be worthwhile. How much is your own time worth to you? You already said you don't enjoy it too much.

If you're soldering manually it will probably show, and that's BAD! Your customers won't appreciate it and most likely you'll lose them after a first sale. Plus they're not going to recommend you to their 1000 Facebook contacts, and you'll be up to an impossible-to-win marketing battle. Non-professional looking products are only accepted if the price is Real Low. So low that it may not be interesting to you. (The Chinese have professional-looking products real cheap, so you're up against them as well.)

Just to say that it is probably a good investment it to spend some money on quality.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.