How can I make YOUR pcb? (not for profit)

[Merlysys]

Making a few PCBs prior to larger production runs seems to be a problem for many hobbyists or even companies. It often takes several weeks when we prefer it be ready in days and priced closer to what we pay to order from China, and also be of high quality.

I want to produce such PCBs, and I will explain why you will benefit. I am an Electronics Engineering Technologist so I have knowledge and experience in this area, though I lack knowledge in some areas. I want to make the PCBs for hobbyists and companies in my local area only, otherwise the shipping costs will make it expensive and shipping times will mean the product won't be available for over a week. So once I get going I want to help others get into this and serve their local area. If several of us get this type of production going then there should be no need to send the work to China, or at least some work can be done in NA/Europe etc. as we will be able to get the PCBs in a few days at reasonable cost.

I work at my own small design company, only today I decided to this, then went a few doors to another little design company and told them, they instantly wanted me to make their pre-production boards as at present it takes weeks to get them done at a reasonable price.

So can you tell me from the beginning what equipment to get to do this? I am thinking of spending a few thousand dollars rather than a hundred thousand, which means some people's needs won't be met as they need perfection. I likely won't be able to do high frequency RF or boards having more than a few layers, along with other limitations. If there is a big demand then investing more will be worthwhile.

So what physical equipment shall I look into? I want to look into the used market. I have plenty of free time, so the equipment not need be very efficient, as long as it gives very high quality product I am ok.

[wraper]

I am thinking of spending a few thousand dollars rather than a hundred thousand

Do you understand that you won't be able to do any serious board production for this money. Yet even do plated visas (with any serious quality). IMO all you would be able to do are max two layer beards with non plated vias.

[free_electron]

a plating installation alone costs several 10 thousand dollars. add an imager, a dry film press, a lamination oven , a drill stage and before you have made a single board you are over 100.000$. and that is if you buy used stuff.
then there is the data preprocesing software and expertise you need there.
and the room to put all that equipment


think i'm exaggerating ?


Drill :
http://www.ebay.com/itm/Excellon-Uniline-2000-Machine-w-UCS-Controller-/251164690238?hash=item3a7a95073e

and that's a simple one...

the rest of the equipment you can't even find on ebay ...

to make a plated double sided board you need a minimum of

- a cnc drill
- a dry film applying machine
- a photoplotter  and exposure unit   or LDI machine
- a film developing machine
- a plating setup ( cleaner bath , actuator bath , carbon ink bath , plating bath and plating power supply all interspersed with various drying and washing stages. )
- an electroless tin plating bath
- a film stripper
- an etching system
- a soldermask ink application machine
- a drying oven
- a silkscreen ink application machine using actual silkscreen stencils or printer
- a drying oven

if you want to do 4 layer or more you will also need
a lamination oven

then there is all the chemicals
all the know how to set up the baths and control them
the preprocessing software to prepare the data all these machines need (Genesis Frontline or Ucamco )

if you want to deliver electrically tested boards you will need a flying probe machine and or a bed of nails tester.
you may also want to invest in an AOI machine

and to deliver the boards you will need a final board milling / depanelisation machine and possible a v-scoring machine

add in a vacuum pack setup and a source of various laminates and a wastewater treatment plant and you are looking at a million dollars.
without building and or permits for all your chemicals and waste handling

in short : can't be done. Not at your set budget , not without the appropriate know how.

I am thinking of spending a few thousand dollars

Clearly you got no clue ...

I likely won't be able to do high frequency RF or boards having more than a few layers

RF boards are no different process than others. As 'technologist' you should know this. Only the substrate material changes.
As for the layers ... for a few thousand dollars you'll be able to make singlesided boards or non plated double sided without masks. Anything beyond that is out  of reach.

I have plenty of free time, so the equipment not need be very efficient, as long as it gives very high quality product I am ok.

High quality ? Today 'chinese' boards are even run on LDI machines and those start at 200K US $ ...

Electronics Engineering Technologist

-cough- riiiight ...

[marshallh]

Just send it to china with rush processing, then put the puts on yourself with a cheapie oven and stencil. I can get a pcb shipped to my door in 4 days with steel stencil, and still pay less than 1/10 the cost of having it produced domestically.

I am an Electronics Engineering Technologist so I have knowledge and experience in this area, though I lack knowledge in some areas. I want to make the PCBs for hobbyists and companies in my local area only, otherwise the shipping costs will make it expensive and shipping times will mean the product won't be available for over a week

Typical tooling could be $100-1500. Overnight shipping costs under $100. It is really not a big deal.
Your time is better served on other aspects of the process, doing things the chinese can't, like QA, burn-in, shipping and fulfillment, things that take knowledge of the product.

[ez24]

I am thinking of spending a few thousand dollars rather than a hundred thousand

Since you do not have your country flag set nor said what country you are in, then if you live in Fairy Land, I would say no problems.

[RJFreeman]

Making a few PCBs prior to larger production runs seems to be a problem for many hobbyists or even companies. It often takes several weeks when we prefer it be ready in days and priced closer to what we pay to order from China, and also be of high quality.

Last batch of prototype boards I had made up, 10 each of three different boards, one 62 x 65mm, then 109 x 23mm and 84 x 107mm cost me $150 AUD including shipping, all double sided, plated through holes, masked, tinned, stencilled overlay etc.
Sent them the Gerbers on Sunday, received the boards on Friday.

ok, $150 might seem a lot for a hobbyist, but in the scheme of things isn't really (you can easily spend that on a case for a project alone) and you are going to need to spend big $ for a set-up that can do the standard of board that I received.

[CatalinaWOW]

If you are willing to invest the time and energy you can keep costs down below those stated (maybe even within your budget).  But you will be a slave to your hobby, and will not match the quality of existing sources.  An area where you will be spending a lot of time that you might not have thought about is complying with local environmental regulations.  There are mechanisms for hobby people to deal with this, but even at the small scale you propose you will have to go full bore commercial compliance.  That reason alone would put me off of this idea.

[KL27x]

a plating installation alone costs several 10 thousand dollars. add an imager, a dry film press, a lamination oven , a drill stage and before you have made a single board you are over 100.000$. and that is if you buy used stuff.
then there is the data preprocesing software and expertise you need there.
and the room to put all that equipment

I never did the math on it, but I figured as much. When I received quotes from a couple of "fast turn-around" pcb "manufacturers" in California, I compared board cost, assembly cost, and turn-around time to overseas manufacturers. I came to the conclusion they were ordering boards from china and populating them in their mothers' basement. These are the outfits with nice websites that score high on Google. And the owner answers the phone, personally. And the address might even be residential.

One of them was indignant when I inquired about assembly, only, supplying the boards myself (at approx half the cost), saying he only works on boards manufactured in-house.... oddly, prior to that, he only used the term "assembled in the US." Suddenly he's a full fledged board manufacturer. He claimed his cost of raw materials was more than the finished FPC's delivered to my door from China. Since his turnaround times were no better, if that were true, he's in the wrong business.

I agree, if you can even just turnkey order and assemble protos/small production runs on boards that are made elsewhere, and not jerk your customers around with BS, I think you can provide a good service. Just to avoid component shipping/customs back and forth, internationally. You can provide a good service out of your mother's basement.

Even if you had all the right equipment and space for FREE, you would still suck at making boards if you didn't do it, constantly. It's a process. The only way to do it, efficiently, is with a lot of people manning the equipment day in and day out. If you tried to buy all the equipment just to provide protos to a few local businesses, you would spend more time tweaking and redoing things than making boards. That's too much to take on, alone. And once you have employees, your costs are going to skyrocket and your time is going to disappear.

[Siwastaja]

I like the concept of DIY PCB's - even "high-tech" ones, meaning through hole plating and solder mask - but then the idea is that you gradually learn the process, build your own machinery (or get lucky and find something for almost free), play around a lot - and then it's your lab. You will have inferior quality to cheap Chinese manufacturers (but not by much, necessarily!), lower yield, and you'll never want to do PCBs for anyone else; it will be labor-intensive and you have risk of your process being out-of-order.

We run an advanced PCB process at a club. Every once in a while, we have a working through hole plating process, other times, it's "under maintenance". But we are still working on it to make it more robust. Some time ago, our CNC drill broke down (it was free; it was old; the over-compilated control side self-destructed), and we are rebuilding it. So now we are limited to double-sided, non-plated-through, with manual drill.

We use dry film photoresists, which is very robust and easy to use. 8/8 mils no problems at all; can go lower in a pinch. We also have soldermask, which looks quite nice; not as nice as professional work, but definitely usable.

I really like the idea of having your own small PCB fab. Go fancy with through-hole plating and soldermasks, but do it gradually, and only do it for yourself (or close enough friends who are interested in the process itself, not only in the result). Done right, you can save a little bit of money, but not much; mainly, you are getting your prototypes the same day. With our process, I can do:

- Single-sided SMD board in 30 minutes
- Double-sided, non-TH-plated board in an hour.
- Double-sided, through-hole plated, soldermasked fancy board in about 4 hours, a lot of it is waiting.

This is, unless parts of the process is "out of order", for some reason. We also had varying level of yield problems in through hole plating, still to be fixed. It's easy to make a demo board with 50 perfectly plated holes, but when you need to have hundreds of reliable holes any day, it's different.

I have never, ever done PCBs for others with this process. But I help people help themselves all the time (for free). This way, there is no responsibility for delays, problems etc. And then they are able to help others further.

I really like having my prototype ready without waiting for days or weeks. It speeds up development considerably. Also, when I have "the flow", I need to work before it dissipates.

If you are still interested in this, try to integrate to a local hacklab or start looking at founding one. That way, you can build the PCB process where the "customers" are, and they will be their own workforce, and everyone can contribute in maintaining the process. That is, if they like it. We have about 10-20 people doing PCBs with the process and a few contributing more.