How good/bad investment in PCB-A service is?

[nardev]

Ok, so i live in a pretty undeveloped country but it's only 5-8 hours drive from Vienna or Munich Germany. Express shipping available, everything pretty well connected. (Bosnia and Herzegovina) And i'm thinking about to establish PCB-A service.

Average payment officially around 460 EUR, i think that skilled technician that would help, can cost in total with taxes, around 800.
Renting or even purchasing manufacturing place is insignificant expense. Don't plan to employ more than one technician, there is available and experienced person for communication with clients other than me.

Also, there is almost no need for PCB-A service in this country, and there is only two companies that i know, that do PCB-A, both for foreign clients and they have pretty long wait list for clients.

The question is, does it make sense, to invest up to 20-25k$ for just and only PCB-A service, to purchase small and simple pick and place machine, like Neoden 4 and Neoden IN6 and just offer custom, low batch, PCB-A and testing service with the limitation and capabilities of the equipment that i mentioned. And perhaps some manual work about the products.

The service would require a client to ship all parts and perhaps even PCBs (although i can get those made in my country pretty affordable but bit slower) and my service would be only as i said, assemble, test, package, custom package even, possible even ship to desired addresses etc...

Does this business make any sense?

Can this be profitable and under what conditions?

How hard is to find clients for such a service?

What are disadvantages and advantages for this business if living in undeveloped country?

What else would you point out?

[nctnico]

I'd strongly suggest to buy components and boards yourself. Then you know they where stored & handled right and / or whether you need to bake them or not. You'd be amazed how some people don't care about handling components the right way. A couple of years ago one of the assemblers I use received a bag with fine pitch SMT connectors thrown in. There wasn't one connector without bent pins. The same for your boards; if your clients tries to save money by using HASL instead of ENIG you end up with boards which are hard to P&P and solder. Likely you'll need to tune things like the paste mask too depending on your paste screening machine.

Maybe you can start smaller using a semi-manual pick & place and see how that goes. If the jobs get bigger then it is time to get a P&P machine. By then you'll also have a better idea on what P&P to get (maybe not the cheapest but the most versatile for low volume jobs). A big cost of a P&P machine is not the machine itself but setting it up for each job.

[nardev]

I'd strongly suggest to buy components and boards yourself. Then you know they where stored & handled right and / or whether you need to bake them or not. You'd be amazed how some people don't care about handling components the right way. A couple of years ago one of the assemblers I use received a bag with fine pitch SMT connectors thrown in. There wasn't one connector without bent pins. The same for your boards; if your clients tries to save money by using HASL instead of ENIG you end up with boards which are hard to P&P and solder. Likely you'll need to tune things like the paste mask too depending on your paste screening machine.

Maybe you can start smaller using a semi-manual pick & place and see how that goes. If the jobs get bigger then it is time to get a P&P machine. By then you'll also have a better idea on what P&P to get (maybe not the cheapest but the most versatile for low volume jobs). A big cost of a P&P machine is not the machine itself but setting it up for each job.

Does it make sense to:

1. have a stock of standard, most common components? Is that very expensive to have "frozen money" in that?

2. yea, right, obviously it's smarter to have one supplier for PCBs, where i'm sure that i'm gonna get good boards, the potential complication here is if i get it from abroad. Than it's hard to avoid import tax and other fees. In case if i have a contract with a client, he can send me any material without any tax, and i'm only doing service locally and i return it back easily.

3. I have been thinking about "semi-manual pick & place" but good machine costs huge money. It's 60-70% of some simple pick and place with 30-40 feeders.

What about current situation in the market, are people in need for this service?

Can you also tell me bit more about this "A big cost of a P&P machine is not the machine itself but setting it up for each job."? Any example why?

[nctnico]

Your biggest competitors are quick prototyping services from China. The advantage you have is that you are local and easier to communicate with.

Regarding the P&P machine: an employee may cost you around 50 euro a day. If that person spends 1 day to re-program your P&P machine and needs to do that 100 times per year for low volume products then the costs are 5000 euro for each year excluding cost of down time for the P&P machine. If you buy a P&P machine which only takes half a day to setup then you save 2500 euro per year and adds 400 hours of extra runtime to your factory. Slightly bigger assemblers often have 2 P&P machines: one for small jobs and one for larger jobs.

It won't hurt to have a standard stock of components. Passives are dirt cheap. Just buy the ones from a reputable brand (instead of cheap crap) and all your customers will be happy with the components you use.

[nardev]

Your biggest competitors are quick prototyping services from China. The advantage you have is that you are local and easier to communicate with.

Regarding the P&P machine: an employee may cost you around 50 euro a day. If that person spends 1 day to re-program your P&P machine and needs to do that 100 times per year for low volume products then the costs are 5000 euro for each year excluding cost of down time for the P&P machine. If you buy a P&P machine which only takes half a day to setup then you save 2500 euro per year and adds 400 hours of extra runtime to your factory. Slightly bigger assemblers often have 2 P&P machines: one for small jobs and one for larger jobs.

It won't hurt to have a standard stock of components. Passives are dirt cheap. Just buy the ones from a reputable brand (instead of cheap crap) and all your customers will be happy with the components you use.

Thank you so much for that explanation.

[nuclearcat]

Your biggest competitors are quick prototyping services from China. The advantage you have is that you are local and easier to communicate with.

And if China doesn't get you, Chinese capitalists will go to your country, invest in a 100x larger factory hiring Middle East and African refugees at a scale and get you, if your business plan ever makes money in any slightest way.

Probably they wont do this, if expected amount of clients wont match this 100x investment.

[olkipukki]

And i'm thinking about to establish PCB-A service.

Out of interest, where is your inspiration came from? Did you ever worked in PCB-A or related industry directly?

[nardev]

And i'm thinking about to establish PCB-A service.

Out of interest, where is your inspiration came from? Did you ever worked in PCB-A or related industry directly?

Well, i have been managing some small batches to be PCBA by some Chinese manufacturers and i even did manual PCBA for at least 1000 boards in past several months. I have some idea about how that goes, not completely out of nowhere.

[nardev]

Your biggest competitors are quick prototyping services from China. The advantage you have is that you are local and easier to communicate with.

And if China doesn't get you, Chinese capitalists will go to your country, invest in a 100x larger factory hiring Middle East and African refugees at a scale and get you, if your business plan ever makes money in any slightest way.

Probably they wont do this, if expected amount of clients wont match this 100x investment.

You are both "right" in a sense but what if, the world turns and "western developed" countries start imposing more and more bans on China or simply start preferring local or nearby companies? If the "openness" and liberalism as we know it simply turns to be something different?

In my country, we already have foreign ambassadors warning politicians not to turn to Chinese technologies and investments for example...

I don't have any personal preferences, i'm simply just trying to find something interesting to invest time and money into. That's why i'm asking with a hope to get proper reply by someone who have experience with this in the western countries.

[nardev]

Well, i have been managing some small batches to be PCBA by some Chinese manufacturers and i even did manual PCBA for at least 1000 boards in past several months. I have some idea about how that goes, not completely out of nowhere.

The only way I think you might stand a chance is:

1. You make export controlled products, which you may or may not qualify.
2. You make industrial/medical products, and you need millions upon millions to certify your factory.
3. You offer extremely fast turnaround time, which you need to be constantly on swift alert and be ready to dive into the shop 24*7.

Pick your poison. Go big and expensive, or go swift and diligent.

Aaaa.. no. I'm primarily asking about just PCB-A for other companies and only devices that are not completely at the edge for manufacturing with the machines that i mentioned.

Don't have knowledge, money and time for more than that. Just wondering about the market and if that can be profitable with such limitations.

[nardev]

Aaaa.. no. I'm primarily asking about just PCB-A for other companies and only devices that are not completely at the edge for manufacturing with the machines that i mentioned.

I understand, but why would a company go to your business?

If they can get cheaper options from China, they will. If not, there will be a huge demand in the West and Germany and Switzerland will swiftly populate the need with full automation.

OSHPark pioneered fully automated PCB pooling and panelization, JLCPCB pioneered the same on assembly. What makes you think the Germans and Swiss will not take their industrial advantage and offer similar services in Europe?

Manufacturing is a game that only the biggest fish survives. If they can get it done at zero labor cost, 1 day turnaround and fully automatically inspected, you are toast. If China is a complete dead end, there will be more than enough money poured into this market, and it will happen.

Regardless how streamlined you think your business is, you are simply no match to algorithm and automation.

Well, ok, an opinion accepted.

I worked with three different companies from central europe, they pretty much think the same as you described but there is one place which is 1000-3000 batch which costs quite some money as they are charged pretty high money per board. Also, post production custom work, i don't know how flexible are those companies that you have in mind. They have pretty much everything very well optimized but also quite constrained.

I thought of that type of scenario.

One example, a fellow of mine had an issue with PCB-A services in china. They have asked for more than 9EUR per board, which is like 22 different parts and maybe 30 parts in total.

The jlcpcb couldn't do that, he also have some parts that are not easy to provide and ship there. He has the pre-programmed controllers that he is not willing to ship to China etc.

for a batch of 3000-5000 boards, i would be pretty profitable if i could do it in a week, and i would not even mention that my price would be only a fraction as my cost i very very low (even when i add unexpected expenses and other guarantees). Plus shipping would be pretty quick and cheap, without any taxes.



 

[olkipukki]

What are disadvantages and advantages for this business if living in undeveloped country?

Is B&H become part of EU? I doubt it...

In this case, what about VAT, import/export duties and taxes, and just simple shipping in/out country?

As example,
https://www.eevblog.com/forum/chat/strange-claim-and-reply-by-digikey-about-min-120-eur-shipping-rate/msg2939204/#msg2939204

[nardev]

What are disadvantages and advantages for this business if living in undeveloped country?

Is B&H become part of EU? I doubt it...

In this case, what about VAT, import/export duties and taxes, and just simple shipping in/out country?

As example,
https://www.eevblog.com/forum/chat/strange-claim-and-reply-by-digikey-about-min-120-eur-shipping-rate/msg2939204/#msg2939204

You are right, it's not in EU but "there is plenty of ways" in my situation to "make as if it is". And besides all, as it's a candidate, we have some tax free relations with EU countries.

[tom66]

From what I understand there are two big costs to pick and place and PCBA.

For a new design, a significant amount of time is taken up by the operator configuring the machine.  This is one area JLCPCB have innovated in, they have an automated service and a database of known components and many machines configured to pick components according to customer configured program.  This means that they can keep many of the ~1,500 parts in their catalogue on machines and just shoot random customer boards through these machines. For other components they have machines with interchangeable reels that allow them to access a larger catalogue but they charge more for these parts.

If you can produce many hundreds of boards for a given order the operator set up time is amortised, and maybe less essential to the overall cost. So it depends entirely on whether you expect to get mostly prototype orders of a few boards at a time or do mass manufacturing.

The secondary cost is maintenance on the machines, loading new reels as reels are exhausted, inventory, and purchasing.  At the company I work at we do most assembly ourselves, and we have estimated that *all in* the cost per component placed on the board is around 4 pence.  Now that actually is quite a lot of money when you consider the average part cost might be a tenth of that (e.g. 0402 resistor) but you have to remember a pick and place machine is not a one-time buy, it is a precision piece of instrumentation that requires annual servicing (a day's work for the engineer so I imagine £500 minimum), plus it needs to be cleaned and parts may break down from time to time.  There is also the energy consumption of the reflow oven.  The oven takes 40kW three phase power, and takes about 10 minutes to warm up and uses about 70% of its rated power once at temperature.  If you are only doing a few boards/hour at an industrial rate of £0.18 per kWh then the oven is costing you £5/hour in energy costs alone - which could be half of what you are paying your employee!   Our boss has taken to suggesting that in summer months we bunch up jobs into the early lunchtime/afternoon period and run them through while sun shines on our 50kW rooftop solar array. It reduces costs a fair bit.  The PnP machine we have also requires dry compressed air, so we have two air dryers and pumps on site, which cost another £20,000 to install and have a bi-annual service schedule.

Purchasing at our company is a full time job for one person, plus we have goods in and a production manager, so that is three full time salaries there, probably £80,000 minimum in salaries, plus taxes and HR costs.  This is for a company with business ca £1m per year. 

Not to put you off, it's great to see people in this field. Just think about the costs and margins.

[nardev]

From what I understand there are two big costs to pick and place and PCBA.

For a new design, a significant amount of time is taken up by the operator configuring the machine.  This is one area JLCPCB have innovated in, they have an automated service and a database of known components and many machines configured to pick components according to customer configured program.  This means that they can keep many of the ~1,500 parts in their catalogue on machines and just shoot random customer boards through these machines. For other components they have machines with interchangeable reels that allow them to access a larger catalogue but they charge more for these parts.

If you can produce many hundreds of boards for a given order the operator set up time is amortised, and maybe less essential to the overall cost. So it depends entirely on whether you expect to get mostly prototype orders of a few boards at a time or do mass manufacturing.

The secondary cost is maintenance on the machines, loading new reels as reels are exhausted, inventory, and purchasing.  At the company I work at we do most assembly ourselves, and we have estimated that *all in* the cost per component placed on the board is around 4 pence.  Now that actually is quite a lot of money when you consider the average part cost might be a tenth of that (e.g. 0402 resistor) but you have to remember a pick and place machine is not a one-time buy, it is a precision piece of instrumentation that requires annual servicing (a day's work for the engineer so I imagine £500 minimum), plus it needs to be cleaned and parts may break down from time to time.  There is also the energy consumption of the reflow oven.  The oven takes 40kW three phase power, and takes about 10 minutes to warm up and uses about 70% of its rated power once at temperature.  If you are only doing a few boards/hour at an industrial rate of £0.18 per kWh then the oven is costing you £5/hour in energy costs alone - which could be half of what you are paying your employee!   Our boss has taken to suggesting that in summer months we bunch up jobs into the early lunchtime/afternoon period and run them through while sun shines on our 50kW rooftop solar array. It reduces costs a fair bit.  The PnP machine we have also requires dry compressed air, so we have two air dryers and pumps on site, which cost another £20,000 to install and have a bi-annual service schedule.

Purchasing at our company is a full time job for one person, plus we have goods in and a production manager, so that is three full time salaries there, probably £80,000 minimum in salaries, plus taxes and HR costs.  This is for a company with business ca £1m per year. 

Not to put you off, it's great to see people in this field. Just think about the costs and margins.

Thank you for some thoughts.

Just to point out. The 1kWh in my country, costs 0.04 to 0.08 EUR depending on day period. During the day it's cheaper from 1300-1600 and also night from 2200-0700. Much cheaper different than in most places.

[tom66]

Is that kWh for domestic or industrial users?

Industrial users with 3ph pay quite a bit more than domestic customers over here, it's about 30-40% higher plus you are billed if your power factor is poor and in some cases if your phases are badly balanced.

[nardev]

In my case i can use Domestic one for "small manufacturing". Because they don't consider my as a industrial but like a "craft".

[tom66]

Just be aware that most equipment used in factories requires 3 phase power. Reflow ovens almost certainly. Pick and place generally low power single phase but the air compressors that feed the PnP need 3ph.

[nardev]

Just be aware that most equipment used in factories requires 3 phase power. Reflow ovens almost certainly. Pick and place generally low power single phase but the air compressors that feed the PnP need 3ph.

Good point, though it's pretty common here because from previous times, a lot of people used electricity for heating, so, it's common installation. But should be confirmed...I agree.

[nctnico]

OSHPark pioneered fully automated PCB pooling and panelization, JLCPCB pioneered the same on assembly. What makes you think the Germans and Swiss will not take their industrial advantage and offer similar services in Europe?

Manufacturing is a game that only the biggest fish survives. If they can get it done at zero labor cost, 1 day turnaround and fully automatically inspected, you are toast. If China is a complete dead end, there will be more than enough money poured into this market, and it will happen.

Regardless how streamlined you think your business is, you are simply no match to algorithm and automation.

That isn't entirely true. There are a huge amount of PCBA companies and they are all making money. Being local and understanding their customers in their own language helps a lot.

[nardev]

Where did you get the idea about that?

Integration of Bayer and Monsanto?

[tom66]

Not until someone in Germany decides to f* the entire European PCBA industry by integrating Digikey/Mouser's infinite stock selection and fully automated assembly with no MOQ, maybe with more tricks under the sleeves like HDI, embedded components and printed resistors.

Wurth is dangerously close to that.

Unaware of this.  What are Wurth doing?