EEVblog Electronics Community Forum
General => General Technical Chat => Topic started by: TheEENerd on July 05, 2013, 12:35:03 pm
-
Thinking of starting up a PCB assembly factory here in Sydney, just wondering if anybody has any experience or thoughts about this idea, like total cost, annual return or any problem with it etc. It would be a single SMT line + wave solder at the begining. I'm a desperate EE.
-
SMT with no oven, wave with no through hole pick and place?
-
It seems like an odd choice of business; there's not much overlap between the skills required to be an EE and the skills needed to run a manufacturing facility. The fact that you're manufacturing electronic items doesn't really matter that much - the CEM business is all about managing suppliers, lead times, stock levels, pricing and scheduling. Very little of it has anything to do with engineering, it's all commercial and management.
Obviously you need operators with basic practical skills who can solder and build stuff by hand, and technicians who can drive the machinery and get it working again quickly when it breaks down, but in terms of techical skills that's about it. Very few CEMs that I know of employ design engineers - they like to be given detailed instructions which they follow literally.
-
I know some guys who have done it. Biggest thing is that you need to find customers first, none of this if you build it they will come bs. A lot of CMs start with guys in their basement hand soldering stuff until their workload gets so big (and they have accumulated enough cash) to invest in some automated equipment.
It can be done with lots of sweat, two toaster ovens and a solder stencil though.
-
Thinking of starting up a PCB assembly factory here in Sydney, just wondering if anybody has any experience or thoughts about this idea, like total cost, annual return or any problem with it etc. It would be a single SMT line + wave solder at the begining. I'm a desperate EE.
A friend and former co-worker did just this. But he had four partners - all wealthy. Trust me on this one thing (if you listen to nothing else). You will be bored but constantly fighting fires. Is that the life you really want?
-
If you where to start up, you would be looking into small company's that don't want to pay the EE's to assemble a few PCB's. Past that, your time (and you do deserve to be paid well for your time) would cost to much to make it worth using you instead a of a PCB fab house. :-\
-
Thanks so much for all the replies. Really appreciated. Just want to quit 9-5 and start sth of my own.
-
Quitting 9-5 may end up with you working 7-6 ! ;)
-
Thinking of starting up a PCB assembly factory here in Sydney, just wondering if anybody has any experience or thoughts about this idea, like total cost, annual return or any problem with it etc. It would be a single SMT line + wave solder at the begining. I'm a desperate EE.
Wait until September, and head to Melbourne for Electronex.
http://www.electronex.com.au/ (http://www.electronex.com.au/)
Any PCB firm "having a go" should be there - talk to them about the business.
However you would only start a new business if the demand is there. Otherwise you'll go nuts and probably broke finding work.
-
That's right, but if you are the boss, you manage your own time. ;D
-
Yes, absolutely true, get customer first. However I wonder how can any customer trust you if you don't have anything running already? That's like egg and chicken problem, isn't it?
[/quote]
Wait until September, and head to Melbourne for Electronex.
http://www.electronex.com.au/ (http://www.electronex.com.au/)
Any PCB firm "having a go" should be there - talk to them about the business.
However you would only start a new business if the demand is there. Otherwise you'll go nuts and probably broke finding work.
[/quote]
-
That's right, but if you are the boss, you manage your own time. ;D
Hmmm... That's true to some extent, but if you want to make good money you still need to put in the hours. A lot of people (not specifically you) who work 9-5 don't appreciate what goes into working for yourself. Starting a business "on the side" can be a good thing, but may not be practical for many situations.
-
Yes, absolutely true, get customer first. However I wonder how can any customer trust you if you don't have anything running already? That's like egg and chicken problem, isn't it?
It depends on whether or not you know anyone in the industry already who knows you, your work, and would be willing to trust that you'll do a good job on that basis alone.
Most of my customers have come via industry contacts - people I've worked with before who have heard that I'm now doing consultancy work. Occasionally I get an email saying "we need a piece of hardware designing, can you help us out?", or sometimes "I know someone who's looking for a hardware designer, could you give this guy a call", and it goes from there.
Finding new customers without that introduction being done on your behalf is much harder. I've done it, but it involves a lot of hard work and more than a little luck - you have to approach the right people at the right time or you're lucky to get an answer at all.
Re: working hours - sometimes you'll be quiet, and you can spend time on admin: web site, taxes, learning new skills, that sort of thing. Then several customers will come along at once, and none of them cares about the others - they all want your attention, right now, until their job is done.
That's why right now, despite the fact that it's a sunny Sunday morning - a rare thing indeed here in the UK - and I have a motorcycle sitting in the garage that desperately needs riding, I'm sitting in my lab about to start a full day's work testing power supplies :'(
-
Yes, absolutely true, get customer first. However I wonder how can any customer trust you if you don't have anything running already? That's like egg and chicken problem, isn't it?
It is but starting a business is all about solving chicken and egg problems. I would say that at first you would need to find a customer who is as desperate for a cm as you are for cm work. Work for the small customers that none of the big guys want to do business with. Execute on those builds in a very lean and cost effective way and just make sure that every new deal that you book is just slightly bigger than the last. All of a sudden you'll have a nice portfolio to catch the bigger fish. Very important to stay very lean, when your small, about the only way is to compete with price first.
Like everyone else says too, having connections is a great thing too... Most importantly, make sure your customers are happy, one sour customer and growth could be stunted really fast!
-
Yeah when working for yourself you do get to pick your own hours. Any 12 a day that you like. :-DD
-
We do our own PCB assemble in-house, and while I am very glad we have this capability in-house, I shudder at the thought of doing it for others on a contracted basis. My god... it would be close to my idea of hell.
-The machines are really expensive to buy and maintain. And there are always new machines coming out that are better in terms of running costs and cost-per-component. I imagine it would be a nightmare trying to keep on top of technology, and having a certain kind of machine would relegate a person to a certain kind of work. And if that work went away (either less popular or losing a big client), it could really railroad someone. I'm lucky in that we have infinitely more capacity than we need and our machines were a luxury, so I don't need them to hit any specific efficiency targets. But a contract manufacturer does
-Running a batch of boards is such a pain in the ass. I am able to forgive a lot of mistakes (or keep my tolerances really high), but in a contract setting, you can't. So I can use solder paste way past it's expiration, and I can hand fix boards that got screwed up, and I can twiddle with the placement locations to correct errors. In a production setting, all of that means $$ lost. Also, I *hate* loading and unloading feeders, especially mid-reel. It would be an absolute nightmare dealing with customers when you have to load/unload every single component for one job. I even go so far as to design parts into jobs based on ease of assembly. Like I might parallel two 10k resistors when I need a 5k, just because I always have 10k's on the machine. But for other customers, you'll always be loading/unloading and picking up dropped parts and dealing with hand placements and trying to find that $4 MCU that fell into the back of the machine and you only had the exact amount, etc.
I also find it pretty boring work. It's not mechanical enough to be fun (like working on a race engine can be), and I derive little sense of satisfaction from the work. The only pleasure I get from it is hearing the machine bleeping and blooping and working while I surf youporn EEVBlog. But it's rare to get more than 10-20 minutes of uninterrupted rest... because you're always either running out of reels or the machine needs adjusted or something else.
I'd say it is as related to electronics as selling car tires is related to the automotive industry. It's very tangential.
Did I mention the huge capital equipment cost to get started? :)
-
Yeah when working for yourself you do get to pick your own hours. Any 12 a day that you like. :-DD
On point.
-
The 12 hour days aren't so bad, at least as long as you're getting paid. It's the 0 hour days that really grind you down :-//
-
Again thanks very much everyone for all the inputs, they are really helpfull :)
-
Napkin calculations.
At least $250K for some decent gear, maybe some 2nd hand stuff.
$50K / year for enough space to run it.
You have no employees and run it yourself.
Let's say you could churn out a uCurrent level job of 100 panels 3 days out of every week, at say $5K a job, with some hand soldering thrown in.
That's $15K / week in jobs for 50 weeks of the year = $750K worth of jobs a year.
Ballpark it does seem possible to pay off your capital investment in the first year.
But really, the error bars are big here.
There are not that many small assemblers in Sydney, but you'd still have to fight for jobs, and maybe have some innovative tie-in with the hacker/maker/crowdsource crowd.
The key of course is having a source of customers up from to get your started.
It won't make you rich, and you'll work your butt off. But it's doable.
-
just wondering if anybody has any experience or thoughts about this idea, like total cost, annual return or any problem with it etc. It would be a single SMT line + wave solder at the begining. I'm a desperate EE.
Talk to the Australian company that sells 2nd hand SMD gear, the name escape me though, they exhibit at Electronex.
Ramzonics paid $500K+ for the line you see in the video I did a while back, minus feeders.
If you are serious about this, and know what you are getting into, 2nd hand gear might be a better bet (or more trouble?) and should the venture fail, the value of gear likely would hold up reasonably well compared to new gear?
-
maybe have some innovative tie-in with the hacker/maker/crowdsource crowd
Personally that's one potential source of custom from which I'd run screaming. The last thing you want are customers who are only going to want one or two boards at a time, and who will undoubtedly need an above average amount of hand-holding and technical support per order. Your overheads will skyrocket.
-
maybe have some innovative tie-in with the hacker/maker/crowdsource crowd
Personally that's one potential source of custom from which I'd run screaming. The last thing you want are customers who are only going to want one or two boards at a time, and who will undoubtedly need an above average amount of hand-holding and technical support per order. Your overheads will skyrocket.
I have dealt with a number of hobbyists in my assembly business and they don't seem to need very much hand-holding at all.
Maybe I've just had a good bunch, idk.
-
Ramzonics paid $500K+ for the line you see in the video I did a while back, minus feeders.
Is it really? I thought that line worth 1 mil, :phew: looks like that was a brand new line and quite advanced.
just wondering if anybody has any experience or thoughts about this idea, like total cost, annual return or any problem with it etc. It would be a single SMT line + wave solder at the begining. I'm a desperate EE.
Talk to the Australian company that sells 2nd hand SMD gear, the name escape me though, they exhibit at Electronex.
Ramzonics paid $500K+ for the line you see in the video I did a while back, minus feeders.
If you are serious about this, and know what you are getting into, 2nd hand gear might be a better bet (or more trouble?) and should the venture fail, the value of gear likely would hold up reasonably well compared to new gear?