JLCPCB lead time

[derGoldstein]

PCB manufacture should almost be a light's out process from the front-end through to packing - with a small number of skilled technicians keeping the machinery & processes in trim.  I suspect that the first fully automated plant will actually be in China though.

Judging by the number of times you have to move the panels from one machine to another, I don't think it'll happen any time soon.
The individual steps may become automated -- for example the screen printing processes where at the moment there's (sometimes) a person "wiping on" the substance -- but there are too many times where you have to unclamp the boards from one machine and then clamp them in/on another. Sometimes you only have to flip the panel within the same machine, but in order to do that in an automated fashion you'd need to double the size of that machine.

In mass production you'll see this done with kuka or fanuc robot arms, but you have to restructure your facility for that, and the problem here is that the dipping-plating processes have to be done in separate rooms from other machines because of the corroding fumes.

Hypothetically this is all doable, but in practice you'd need to redesign a facility from scratch. Right now it ain't broke, so they don't fix it.

[mrpackethead]

having been to some PCB plants in China, some of  the 'automation' is provided by low cost humans.    One of the factorys i visited had 1800 staff..  Its hard to fathom the scale of the operation, and how big it is untill you see it.     When you see several hundred pallets of FR4 base material, and are told this is what will get used today!    Acids arriving it double truck/trailer tankers.     They had an old site school for the children of the workers,  400 kids there!     

[fcb]

Whilst land, labour and the 'environment' is cheap, the incentive to automate isn't really there. Yet.

Foxconn (Apple) are already using robots to replace many workers, and that's in China.  Valve built a heavily automated factory in USA to build their Steam controller. My view is that China will lead the world in automated factories and they will upskill their workers to become engineers that build and run these factories. I've been going round UK PCB plants since the early 90's, and the pace of change is glacial - they are so ripe for an Elon Musk level re-think.

VALVE:
CIG:

But back to JLCPCB, be careful if you mix a batch of prototypes and production boards together, obvious I know, but the production boards do take much longer to build.

[coppice]

I've been going round UK PCB plants since the early 90's, and the pace of change is glacial - they are so ripe for an Elon Musk level re-think.

They are responding appropriately to a dying UK electronics industry. Being a vibrant go ahead company won't get them anywhere without a vibrant go ahead market for their products. The potential for UK PCB makers to develop a strong export market is rather limited.

[mikeselectricstuff]

I've been going round UK PCB plants since the early 90's, and the pace of change is glacial - they are so ripe for an Elon Musk level re-think.

They are responding appropriately to a dying UK electronics industry. Being a vibrant go ahead company won't get them anywhere without a vibrant go ahead market for their products. The potential for UK PCB makers to develop a strong export market is rather limited.

They won't be exporting much, but it might get them ahead of other local competition, and swing a few decisions away from Far-East suppliers on more time-sensitive jobs.
Of course one advantage of "old fashioned" businesses where you deal with a person and not a fully automated system is flexibility.

[coppice]

I've been going round UK PCB plants since the early 90's, and the pace of change is glacial - they are so ripe for an Elon Musk level re-think.

They are responding appropriately to a dying UK electronics industry. Being a vibrant go ahead company won't get them anywhere without a vibrant go ahead market for their products. The potential for UK PCB makers to develop a strong export market is rather limited.

They won't be exporting much, but it might get them ahead of other local competition, and swing a few decisions away from Far-East suppliers on more time-sensitive jobs.
Of course one advantage of "old fashioned" businesses where you deal with a person and not a fully automated system is flexibility.

You have a history of dealing with UK PCB makers. A good sales person is everything in getting good service. As the good sales people get poached from company to company a lot of the business moves with them. Keeping the best sales people is the company's focus. Investment in greater efficiency less so.

UK PCB makers have never really been focussed on advanced PCBs. In the 70s and 80s we had to get all our advanced (i.e. fine traces, lots of layers, buried vias, etc.) PCBs made abroad. It was usually North America in those days, so the choice of import source has changed.

[mikeselectricstuff]

You have a history of dealing with UK PCB makers. A good sales person is everything in getting good service. As the good sales people get poached from company to company a lot of the business moves with them. Keeping the best sales people is the company's focus. Investment in greater efficiency less so.

I have always had the best service from PCB places that are too small to have sales people, or at least where the people you routinely deal with are not sales people.
Good places get repeat business and referrals.
I HATE dealing with sales people, as they are often clueless about technical aspects and will tell you whatever you want to hear.

[mrpackethead]

You have a history of dealing with UK PCB makers. A good sales person is everything in getting good service. As the good sales people get poached from company to company a lot of the business moves with them. Keeping the best sales people is the company's focus. Investment in greater efficiency less so.

I have always had the best service from PCB places that are too small to have sales people, or at least where the people you routinely deal with are not sales people.
Good places get repeat business and referrals.
I HATE dealing with sales people, as they are often clueless about technical aspects and will tell you whatever you want to hear.

I agree,  my rule is that if i cnat' talk to the boss of the company then they are probably too big for us to deal with  ( unless its a 'trivial supply like cellphone  provider etc )..    This rule breaks with the parts supply chain..  But it works pretty well mostly.   'Boss' just needs to be someone who can fix problems when they happen.

[fcb]

A link to this just arrived in my inbox:

http://www.smtnet.com/library/index.cfm?fuseaction=view_article&article_id=2782

http://www.smtnet.com/library/files/upload/pcb-fab-factory-design.pdf

Quite interesting, especially the "zero liquid discharge" maths.

[derGoldstein]

A link to this just arrived in my inbox:
http://www.smtnet.com/library/index.cfm?fuseaction=view_article&article_id=2782
http://www.smtnet.com/library/files/upload/pcb-fab-factory-design.pdf
Quite interesting, especially the "zero liquid discharge" maths.

That's a fascinating document, but I thought that copper recovery was something that was done in every PCB fabrication facility... Isn't that the primary expense in terms of raw material use? Did they just used to dump the copper compounds without recovering the copper?

[mikeselectricstuff]

A link to this just arrived in my inbox:

http://www.smtnet.com/library/index.cfm?fuseaction=view_article&article_id=2782

http://www.smtnet.com/library/files/upload/pcb-fab-factory-design.pdf

Quite interesting, especially the "zero liquid discharge" maths.

Interesting, but you have to be making a hell of a lot of PCBs to spend $12M on the factory....

[coppice]

A link to this just arrived in my inbox:
http://www.smtnet.com/library/index.cfm?fuseaction=view_article&article_id=2782
http://www.smtnet.com/library/files/upload/pcb-fab-factory-design.pdf
Quite interesting, especially the "zero liquid discharge" maths.

That's a fascinating document, but I thought that copper recovery was something that was done in every PCB fabrication facility... Isn't that the primary expense in terms of raw material use? Did they just used to dump the copper compounds without recovering the copper?

I think what they are trying to say is other plants have a net balance of $6100 between recovering the copper, and selling the recovered copper, so there is a net cost. For their new system the difference is $-6758, so there is a net gain. I don't know how realistic those figures are, but I feel like the document is not comparing their solutions with the current best in class.

[coppice]

A link to this just arrived in my inbox:

http://www.smtnet.com/library/index.cfm?fuseaction=view_article&article_id=2782

http://www.smtnet.com/library/files/upload/pcb-fab-factory-design.pdf

Quite interesting, especially the "zero liquid discharge" maths.

Interesting, but you have to be making a hell of a lot of PCBs to spend $12M on the factory....

What would be the point of starting a PCB making business, if it isn't going to make a hell of a lot of PCBs? It was only an industry that favoured the small business when PCB manufacture was quite crude. $12M is hardly a major investment in most industries.

[mikeselectricstuff]

A link to this just arrived in my inbox:

http://www.smtnet.com/library/index.cfm?fuseaction=view_article&article_id=2782

http://www.smtnet.com/library/files/upload/pcb-fab-factory-design.pdf

Quite interesting, especially the "zero liquid discharge" maths.

Interesting, but you have to be making a hell of a lot of PCBs to spend $12M on the factory....

What would be the point of starting a PCB making business, if it isn't going to make a hell of a lot of PCBs? It was only an industry that favoured the small business when PCB manufacture was quite crude. $12M is hardly a major investment in most industries.

I wonder how many Chinese places are investing that much in new equipment. The document is suggesting significant savings over outsourced Far-East PCBs, but you have to be doing big volumes to amortise that $12M investment

[coppice]

Interesting, but you have to be making a hell of a lot of PCBs to spend $12M on the factory....

What would be the point of starting a PCB making business, if it isn't going to make a hell of a lot of PCBs? It was only an industry that favoured the small business when PCB manufacture was quite crude. $12M is hardly a major investment in most industries.

I wonder how many Chinese places are investing that much in new equipment. The document is suggesting significant savings over outsourced Far-East PCBs, but you have to be doing big volumes to amortise that $12M investment

The presentation tries to paint a picture of crude Far-East PCB makers versus a highly automated plant in the west, and this is misleading. The presentation talks about the 2015 HKPCA show (this is the Hong Kong Printed Circuit Association, but nobody makes PCBs in HK any more. Its all about production in China, even if a lot of it is owned by HK interests), and says that automated solutions for every step in the PCB making process were on display. They were not displaying in Hong Kong to achieve sales in the west. With the great majority of PCB manufacture now occurring in China, the market for all this automated equipment is mostly factories in China. If the better makers in China were not heavily automating, the market for all that equipment would be extremely small. The market actually seems to be big enough to encourage a number of suppliers to bring a number of advanced manufacturing systems to market.

With salaries rising in China, and automation reducing the manpower in PCB plants, the costs of east vs west are certainly evening up to some extent. However, nothing is described in the presentation which would actually make western production cheaper than China's. If the costs are comparable, most PCBs will be made close to the board assembly companies - its a basic JIT principle. Those are mostly in Asia, so Chinese production of PCBs still seems to have an advantage over western production. With more electronics assembly moving out of China to places like Vietnam, the Chinese PCB makers seem more threatened by PCB manufacture moving to other Asian countries than it moving back to the west.

[fcb]

With 100% automation of PCB processes will likely come an increase in quality.

Already I get equivalent or higher quality boards from China than the UK and on shorter lead-times at a fraction of the cost. For years I've supported UK PCB manufacturers, but now I feel they are 'doing it to themselves' and this under-investment is really showing.

Still feel guilty not sourcing PCB's from the UK, but then most of the bits I stick to PCB weren't made in the UK (or even the EU).

[derGoldstein]

With 100% automation of PCB processes will likely come an increase in quality.

Already I get equivalent or higher quality boards from China than the UK and on shorter lead-times at a fraction of the cost. For years I've supported UK PCB manufacturers, but now I feel they are 'doing it to themselves' and this under-investment is really showing.

Still feel guilty not sourcing PCB's from the UK, but then most of the bits I stick to PCB weren't made in the UK (or even the EU).

Interestingly I've found that a very convenient parametric website will overtake other drawbacks. Where I am (Israel), there are about 15 places that I know of that will laser-cut sheets of metal/plastic/wood, etc., at competitive prices. The bureaucracy is terrible, however. If you finally get your simple-and-to-the-point DXF file "understood" and approved (and we're talking about a dozen emails back and forth to get to that point), they take their sweet time making the parts, and then if you're lucky, they'll deign to inform you that the parts are done (otherwise you have to nag them, often on the phone).

So what companies I've been working with have been doing lately is getting prototypes from Sculpteo, and then get higher volumes from china. There's no lack of available local services, there's just a lack of convenient ones.

[Falcon69]

I placed an order with them. The price really goes up if you change anything, i.e. soldermask color or copper thickness.  I needed 2oz. copper, and the price went up from $2 to $20 for 10 boards. It's funny how 10 boards cost the same as 5 boards. 

I just hope that a 2.5mm trace width with 2oz. copper and then solder applied the length of the trace on BOTH sides of the PCB along that trace can handle 30A (will probably never see 30A, but the relay I have is capable of that for switching).

My design was 100mm x 100mm and 3 different designs (8 total boards on the 100x100 panel).  There was no option to select different designs, so I hope they don't come back and say they need more money from me. If that is the case, I may cancel order and order from elsewhere, as the cost would then be the same as another manufacturer, and I could get a different color solder mask (I really hate green, but for prototype is okay).

The board I designed was for my Vehicle Alarm. It is used to cut of the fuel with a manual switch, but the remote start would still work (remote start on remote will override the manual switch).  I tried to buy this circuit from the Alarm Installers here in my city, and it seems that nobody makes this.  Basically, if you have remote start, you cannot have any manual switch (hidden) to cut-off the fuel pump, as the vehicle won't start unless the switch is manually turned on.  My circuit solves that, I hope.

I'll let you know how the order goes.  I almost didn't order from JLCPCB as there was a post some time back of that manufacturer spamming this forum.

Total Price $49.70 with Stencil and 10 panelized boards shipped to USA with DHL.

Order placed at 5:30am PST on 11/6/2017.

[Falcon69]

Got the boards and stencil today. The stencil was bent from the package being dropped by DHL. But the actual area where solder paste will go is okay, but I'll have to take out the tin snips and cut off the bent area so it lays flat.

The boards look okay. The screen printing is total crap though compared to other companies I've ordered from.

I am upset with them though. I specifically asked them not to print any in-house identifying numbers on it, and they did anyway. One thing I cannot stand are people/businesses who cannot follow instructions. For that reason, I will not order from JLCPCB/EasyEDa again.

They did NOT charge me extra for having 3 designs either.

I'll try and post up some pics tomorrow after I have time to assemble one.

[Monkeh]

They need to have some way to identify your PCBs. Either pay for an entire panel, or accept they're going to mark your PCBs.

[FrankBuss]

They need to have some way to identify your PCBs. Either pay for an entire panel, or accept they're going to mark your PCBs.

Some companies don't need it, for example the PCBs from https://oshpark.com look exactly as you ordered them, with no id number. And they have a preview when you upload your design, very nice, but more expensive than other companies for bigger boards. For smaller boards it is cheaper, because they provide free shipping.

Might be interesting to have a list of companies who don't print an id on it.

[Monkeh]

They need to have some way to identify your PCBs. Either pay for an entire panel, or accept they're going to mark your PCBs.

Some companies don't need it, for example the PCBs from https://oshpark.com look exactly as you ordered them, with no id number.

OSHPark is an unusual, and very much more expensive when comparing apples to apples (at least on size - ENIG is costly..), service.

Putting an identifying number on the PCB to maximize yield is standard practice for low cost fabs. Provide somewhere for it to go, or if the stellar cosmetics of your PCB are that critical: PAY TO ENSURE IT!

[derGoldstein]

OSHPark is an unusual, and very much more expensive when comparing apples to apples (at least on size - ENIG is costly..), service.

OSHPark's price is optimized for many small boards. If you only need 3 of each, you can send 20 different designs and you'll only pay for area used. You're right they're not the place to get larger PCBs, but for a large number of test PCBs they're great.

One thing I've used them for is when testing the efficiency of DC-DC converters. I'll design maybe 15 different layouts (using the same components), and order 3 of each. I'm able to explore a much wider range of layouts and test which one gives me the optimal efficiency for its size.

Another thing I use them for is when testing previously-untried footprints. Companies are coming out with some *very* weird footprints for power ICs -- multiple under-chip pads of irregular shapes. You have to design the footprint yourself, and sometimes the tolerances aren't clear. Getting several boards with different footprint versions lets me do multiple tests for much cheaper than if I had to use one of the chinese manufacturers.

Horses for courses.

[Falcon69]

Well, I finally had a chance to solder up those boards and take some pics.

Overall they soldered just fine and the stencil lined up perfectly.  The hole alignment in the vias and middle of pads (where wires go) seems to be centered, not perfectly, but pretty good.  I've had other board manufacturers that milled the holes in the vias where the hole is almost touching the edge of the via.

I have not had time to hook it up and see if this thing works, but that is not on the manufacturer of the boards, that's on me. All the traces looked good. No bodges or anything.  They actually sent 11 boards (22 circuits total) instead of the 10 ordered.  I question whether it is 2oz copper, as it looks like it's the same thickness as my previous boards (from another supplier) that was only 1oz. copper.

[pilleya]

Just received a second order from JLCPCB, I ordered on 3/12 and received them today 15/12. Pretty happy with the leadtime especially for the price.