4 layer PCB manufacturing - machine(s) requirements for complete production line

[perieanuo]

Hello everyone,
I need advice from people who worked in pcb manufacture.
I will sign monday a new job contract with a company who wants to produce 4-layer pcb's (from prototyping to low volumes, small pcb size, below A4 they said, but i'm sure they will need bigger size like maybe towards ~ 600x600mm, for led panels for example, they will produce also this type of pcb).
So, let's build a "machine" list.
Seems funny to start like this, but viewing the fact i always was on the other side of the pcb production process (maintenance, then r&d --> i worked in a company making the schematic and layout design, then prototyping, validation etcatera, and finally small production like 2k-5k pcb's per month, but pcb production and assembly was not in our company, we were only the designers).
I know in theory what they need, but as i never visited a pcb production site, i don't qualify as expert and i can do stupid beginner's mistakes like forgetting the air compressor, gas evacuation or some intermediary ultrasonic cleaning machine. you got the idea.
So i ask humbly for advice in what machines i need to propose for that 4 layer prod site, from the 2-sided untreated pcb's upto final pcb sent to clients.
My first idea is a complete production line (at least the main body), but as i said, i'm starting.
I saw lots of industry so i'm not really starting my first contact with this environment, i will understand the details as you propose them.

So, anyone experienced can give me some guiding lines in what machines are necessary, but detailed? I want to reduce the things i forget to count to a reasonable percentage with your help  so the final amount won't be doubled by my mistakes. I saw experienced engineers also forgetting to buy something they couldn't avoid in the prod line, and that something was over 6000 euros without transportation at the time (one example, the cooler for a compressor in a slaughter house, half of the machines with small pneumatics were not functional whet the external temp was over 35 deg C and condensations hit some tiny electrovalves)

thanks in advance!!!

[tszaboo]

Just think of a number, double it, and then write a zero after it, and add like 12 people's salary to it. And then realize you cannot hire those 12 people.

[mikeselectricstuff]

This video will give you an idea of the scale of what's involved


Making 4L PCBs in-house seems a pretty crazy thing to want to do though when there are already so many companies offering this service.

[KaneTW]

Evaluate the requirements whether you really need to produce in-house.

Then contact a local vendor and work it out from there.

[Alex Eisenhut]

I want to add a video too, just set the time machine to 1969! My little Tektronix/1960s fetish.



You're not opening a bakery here. Just dealing with the chemicals, supply and disposal, will probably keep you busy for months.

[coppice]

Are you sure you will really be expected to manufacture the actual boards, and not just assemble them? Is this for some clandestine activity? The crazy cost may be worth it for some specialist military work, but you need to be making boards by the million to make it economical to do. Unless you can hire a very experienced team to operate the facility, you are also likely to have a long period of quality problems as they gain experience.

[perieanuo]

This video will give you an idea of the scale of what's involved


Making 4L PCBs in-house seems a pretty crazy thing to want to do though when there are already so many companies offering this service.

that one is already +- 'reverse engineered', thanks
no matter how many they are, there is always room for another, viewing un EU we need to repatriate this type of production
we already are very aware of the enterprise level, marketing involved, skilled technicians and so on
what i ask is feedback of machining manufacturers, personal experience and traps to avoid, watching videos is done of course, no ofense

[perieanuo]

Are you sure you will really be expected to manufacture the actual boards, and not just assemble them? Is this for some clandestine activity? The crazy cost may be worth it for some specialist military work, but you need to be making boards by the million to make it economical to do. Unless you can hire a very experienced team to operate the facility, you are also likely to have a long period of quality problems as they gain experience.

my boss tells me what i do, i do. i will point all the marketing and complexity of prod line, but he said already he want to produce in-home, i expect he has clients for such a big enterprise.
a similar company took several years to start that pcb production (they make for years 4-layer and collateral products, so yes they found the market for this and i agree 100% we need to rebuild in Eu this type of prod lines. we have already a strong r&d industrial park (maybe 10k employees in different companies from all over Eu countries, maybe they found the clients there->i don't wanna discuss this 'find clients' subject as at this level this negotiation is not always what people think , i'm the prod engineer, not the commercial one, so only tech problems in my sphere of interest)
for assembly, i'll do my own research, we will do this also, in fact we already do this in some level of expertise (i've done similar r&d and small prod in other company)

[YetAnotherTechie]

Are you expected to go from 0% inhouse to 100% in house production with no experienced personal?
I recommend you set up a working relationship with a local pcb house, that will build your pcbs. In that relationship you will visit and see how things are done. In parallel you can set up a line in-house where your local technicians can learn the machines and processes in prototype runs. When they are experienced enough you can start producing small lots, and ship those. I've seen small companies have their own in-house pcb production(assembly) with low volumes and a single machine of each type, but the knowledge was developed over time. The usual time-money-quality triangle applies here, and it's seems you're being set-up to make it happen with two of those missing. Who's taking the blame when it fails?
Re-reading your post it seems like you want to make the pcbs, including copper deposition and etching?? Have you done any feasability study for this? The enviromental and chemicals? How much support can you get from the machine suppliers? They should tell you everything you need for a turn key solution.
As far as equipment needed you can look up auction houses, there are sales of production lines regularly, you can contact them for your needs, or buy something and then when you go pick up your items, take a tour. (i've done that, coincidentaly it was in France)

[iesn]

Personally I know a company that start to make PCB in '90. Is growing, they made up to 8 layers PCB (if I remember correctly) for almost an entire country (Romania) and for some neighbor country (Hungary, etc.). With 2 or 3 years ago I speak with a good old friend (that is in electronic prototipe industry and work with this company from start to... end...) and I ask them about this company because at this time I need to prototype a board. Well... my friend it say that the company is... CLOSED ! I ask why ?! Well... todays days... all guys work with Chinese company, nobody can fight with the prices that have the Chines company. I know... the quality are that are... but seems that this is the... philosophy... Included I... I have 2 or 3 project and the PCB... well... come from China. I will see if I need to kill my best old friend or... not.

[asmi]

Ten 10x10 cm 4 layer PCBs now cost about 30$ delivered and the lead time is usually 7-10 days (that includes everything from placing the order until the boards are in your hands). Think about how are you going to match this price and lead time. The level of automation you require (and consequently the amount of capital investment) to achieve that is mind-boggling. People love talking about the importance of local production and all that jazz, but talk is cheap, and I don't see a lot of willingness to put their money where their mouth is. So the price is everything. Unless some sort of public body foots the bulk of the bill for initial expenditures, I don't see anyone willing to invest so much money to make it a reality.

[mon2]

Do not want to be a Debbie Downer but as per above posts, you really think you can challenge the investments made by companies such as JLCPCB ? Keep in mind that in China, their government will assist such giants to be larger and stronger. For us, at least in Canada - when we landed a deal with Logitech many moons ago - the government rushed in to audit us. They could not believe we did this so quickly and left after a few hours with a congrats comment. Screw them and the horse they rode in on. Now we supply the world (with zero funding and no support from the leaching Canadian government) and the top tier OEMs. For us, it was important to find the balance to support the clients. I am sure you can find your unique vertical in this but do not underestimate the capitol and knowledge required to build a PCB factory.

Be sure to view this video before jumping in too deep. In my opinion, you will be better off with brokering your services to those that may not understand the complexity of out-sourcing and/or engineering. A technical middle-man / middle-woman / middle-whatever.

Video is brought to you from the Burger King guy. Its good to be the king.

[jduncan]

Are you sure you will really be expected to manufacture the actual boards, and not just assemble them? Is this for some clandestine activity? The crazy cost may be worth it for some specialist military work, but you need to be making boards by the million to make it economical to do. Unless you can hire a very experienced team to operate the facility, you are also likely to have a long period of quality problems as they gain experience.

my boss tells me what i do, i do. i will point all the marketing and complexity of prod line, but he said already he want to produce in-home, i expect he has clients for such a big enterprise.
a similar company took several years to start that pcb production (they make for years 4-layer and collateral products, so yes they found the market for this and i agree 100% we need to rebuild in Eu this type of prod lines. we have already a strong r&d industrial park (maybe 10k employees in different companies from all over Eu countries, maybe they found the clients there->i don't wanna discuss this 'find clients' subject as at this level this negotiation is not always what people think , i'm the prod engineer, not the commercial one, so only tech problems in my sphere of interest)
for assembly, i'll do my own research, we will do this also, in fact we already do this in some level of expertise (i've done similar r&d and small prod in other company)

If they are actually serious about this, they need to fire you yesterday and put someone with actual manufacturing experience on the project. you are in so far over your head you can't even tell you're underwater.

[Alex Eisenhut]

Personally I know a company that start to make PCB in '90. Is growing, they made up to 8 layers PCB (if I remember correctly) for almost an entire country (Romania) and for some neighbor country (Hungary, etc.). With 2 or 3 years ago I speak with a good old friend (that is in electronic prototipe industry and work with this company from start to... end...) and I ask them about this company because at this time I need to prototype a board. Well... my friend it say that the company is... CLOSED ! I ask why ?! Well... todays days... all guys work with Chinese company, nobody can fight with the prices that have the Chines company. I know... the quality are that are... but seems that this is the... philosophy... Included I... I have 2 or 3 project and the PCB... well... come from China. I will see if I need to kill my best old friend or... not.

https://www.olimex.com/PCB/

[Styno]

I recently read a story about a US electronics design company that did a lot of high reliability designs. They decided to do their pcb mfg in-house. The result was a fully automated, fully tracked, high capacity and near zero-waste factory that had a payback time of only 2 (or 4, I forgot) years over their Chinese suppliers.

So, it can be done but you need volume and a high upfront investment that will pay back in reduced labor cost and quality.

[nctnico]

Personally I know a company that start to make PCB in '90. Is growing, they made up to 8 layers PCB (if I remember correctly) for almost an entire country (Romania) and for some neighbor country (Hungary, etc.). With 2 or 3 years ago I speak with a good old friend (that is in electronic prototipe industry and work with this company from start to... end...) and I ask them about this company because at this time I need to prototype a board. Well... my friend it say that the company is... CLOSED ! I ask why ?! Well... todays days... all guys work with Chinese company, nobody can fight with the prices that have the Chines company. I know... the quality are that are... but seems that this is the... philosophy... Included I... I have 2 or 3 project and the PCB... well... come from China. I will see if I need to kill my best old friend or... not.

https://www.olimex.com/PCB/

Dang... did they close down their PCB service? I'm more inclined to believe that Eurocircuits killed them by being better at the logistics side while keeping costs down. AFAIK Eurocircuits has all the engineering / preparation done in India or so but small batch production occurs in Europe.

[YetAnotherTechie]

https://www.olimex.com/PCB/

Dang... did they close down their PCB service? I'm more inclined to believe that Eurocircuits killed them by being better at the logistics side while keeping costs down. AFAIK Eurocircuits has all the engineering / preparation done in India or so but small batch production occurs in Europe.

They closed the PCB service some 10-15 years ago, said they had no leftover capacity, and it was all for their internal projects, the stuff they sell on the webshop.

[perieanuo]

Just think of a number, double it, and then write a zero after it, and add like 12 people's salary to it. And then realize you cannot hire those 12 people.

hi,
i worked 10 days for the company
finally after realising what means production line for 4-layer they said let's forget that for now and focus on assembly

so reset, restart the 'dossier' with stencil printer, pick-and-place and oven. envelope 20 keuros. after realising the guys wanted to do this prod line with ONE employee for 1000 boards/month, also 3 pcb's not yet designed+prototypes+validation for the next 3 months (including starting from scratch tht pnp line). i was assigned to be r&d designer, prototyping technician, production line operator, production supervisor and why not, janitor and who knows what else. luckily they had already like 5 accountants and 2 rh plus a little football team as commercial diggers.
i liked the idea to repatriate the assembly in europe, but not 1 production employee for 10 non-productive, the percentage seemed wrong

they also 'heard' the autotronik offer (they sent me 3 smt line offers, 69keeuros, 102keorus and 200keuros without vat and transportation, can't disclose the details), way off the records for that 20keuros envelope...
so move on baby back to consumerism mode, we like this in EU, bruxelles tells us consuming is goooood (i'm talking about me here) and business owners think the chinese produce those magnific pcb's with those crappy cheap machines 

[perieanuo]

thanks everyone for helping, maybe the admin can delete this topic (i saw better info in parralel ones)?
you know if i can do this myself by deleting first post?
regards, pierre

[perieanuo]

ha ha, update, today they called me for renegotiating the r&d job (i'll do only the series 0 prototype), they decided the pnp part could be done with something like qm61 or smt660/880 pnp machines. and some other guy will manage this assembly line, works for me.
so much for my first chance to do something solid 
at least they realised the product conception is not a 2 weeks job ( i mean a competitive product not some pcb making popcorn when mains fluctuation comes)
btw, what ETA do you have for successful proto starting from scratch (what the un-informed client wants) for a medium-complex pcb (like 120x80mm, including 3-4 input modular sensors incorporated on-board, microcontroler core, 2-3 dcdc's, some opto isolators)?
where i worked, boards like this took sometimes 2-3 months to be up-and-running 100%, firmware included of course, with all client hw/sw validations passed correctly and some bigger/more complex pcb's even more

[Styno]

I recently read a story about a US electronics design company that did a lot of high reliability designs. They decided to do their pcb mfg in-house. The result was a fully automated, fully tracked, high capacity and near zero-waste factory that had a payback time of only 2 (or 4, I forgot) years over their Chinese suppliers.

So, it can be done but you need volume and a high upfront investment that will pay back in reduced labor cost and quality.

actually it’s from this video with pcb expert Happy Holden:

[Rat_Patrol]

Just think of a number, double it, and then write a zero after it, and add like 12 people's salary to it. And then realize you cannot hire those 12 people.

hi,
i worked 10 days for the company
finally after realising what means production line for 4-layer they said let's forget that for now and focus on assembly

so reset, restart the 'dossier' with stencil printer, pick-and-place and oven. envelope 20 keuros. after realising the guys wanted to do this prod line with ONE employee for 1000 boards/month, also 3 pcb's not yet designed+prototypes+validation for the next 3 months (including starting from scratch tht pnp line). i was assigned to be r&d designer, prototyping technician, production line operator, production supervisor and why not, janitor and who knows what else. luckily they had already like 5 accountants and 2 rh plus a little football team as commercial diggers.
i liked the idea to repatriate the assembly in europe, but not 1 production employee for 10 non-productive, the percentage seemed wrong

they also 'heard' the autotronik offer (they sent me 3 smt line offers, 69keeuros, 102keorus and 200keuros without vat and transportation, can't disclose the details), way off the records for that 20keuros envelope...
so move on baby back to consumerism mode, we like this in EU, bruxelles tells us consuming is goooood (i'm talking about me here) and business owners think the chinese produce those magnific pcb's with those crappy cheap machines 

Wait, they are giving you 20k Euros to setup for assembly?

Yeah, unless you are going to be rocking Charmhigh 48VB machines, lol

[tszaboo]

Just think of a number, double it, and then write a zero after it, and add like 12 people's salary to it. And then realize you cannot hire those 12 people.

hi,
i worked 10 days for the company
finally after realising what means production line for 4-layer they said let's forget that for now and focus on assembly

so reset, restart the 'dossier' with stencil printer, pick-and-place and oven. envelope 20 keuros. after realising the guys wanted to do this prod line with ONE employee for 1000 boards/month, also 3 pcb's not yet designed+prototypes+validation for the next 3 months (including starting from scratch tht pnp line). i was assigned to be r&d designer, prototyping technician, production line operator, production supervisor and why not, janitor and who knows what else. luckily they had already like 5 accountants and 2 rh plus a little football team as commercial diggers.
i liked the idea to repatriate the assembly in europe, but not 1 production employee for 10 non-productive, the percentage seemed wrong

they also 'heard' the autotronik offer (they sent me 3 smt line offers, 69keeuros, 102keorus and 200keuros without vat and transportation, can't disclose the details), way off the records for that 20keuros envelope...
so move on baby back to consumerism mode, we like this in EU, bruxelles tells us consuming is goooood (i'm talking about me here) and business owners think the chinese produce those magnific pcb's with those crappy cheap machines 

Boards like that can be ordered for about 2 EUR per piece from Europe, probably even cheaper if you shop around a abit. So they planned to hire someone to make this production, probably even part time, the ROI of the line is negative. But I would argue, that even without it, you wouldn't be able to match those prices, so it would be a loss leader forever.
The fact that they want to make a production line for such a tiny production runs tells me that the MBAs in that company are completely clueless about the electronics market, and I highly recommend you not to take a full time position there. Charge them by the hour. Or have an alternative ready to go on a moments notice, when they have run the business into the ground. Or maybe they are the geniuses, who gets government funding to cover everything, and they can do whatever they want with the endless supply of money - no matter how stupid the company is.