SMALLSMT Vision Placer Pick & Place machines

[Smallsmt]

Yes the software is common to all machines we make. You can use a copy to do offline preparations but need to use a second dongle if machine is operating parallel.

[IconicPCB]

Thank You Michael.

[kgbxyz]

Thanks to all the participants for the comments.

By reading the thread, it is now clear to me what is the situation with Chinese SMT machines (Neoden, Charmhigh, Xjyingxing, etc. and unfortunately Smallsmt/Yuchengtech).

Taking into account the fact that we can not build a castle from the crap, I have a question in my mind. Namely, why can not work Open-source hardware approach in practice? We have created a lot of things in the West over the last centuries. Unlike the Chinese, what is that we do not know? (not video editing, photoshopping, etc.)
OK. it is the world's second largest economy by nominal GDP, and labor is inexpensive. Apart from this, what makes it possible for them to produce so cheaply, but we can not? I apologize for the off-topic question.

I just write the above, because this topic has started more than 2 years ago. This time for a professional would have been abundant enough to build a really good machine, write the software and test it. Not to mention if that this is done in teamwork.
Apart from some semi-finished work, there is no opensource solution currently available on this topic. In other words, can a good machine be opensource or it is conceivable only as a commercially available product?

[IconicPCB]

It was worlds largest economy just prior to industrial revolution.
Industrial revolution simply produced harder steel, bigger bombs, more potent drugs in support of bigger and stronger empires.
Ask any Chinese what they think of the the destruction of their nation brought about by Opium wars and gunboat diplomacy.

[Kjelt]

By reading the thread, it is now clear to me what is the situation with Chinese SMT machines.
I have a question in my mind.
Namely, why can not work Open-source hardware approach in practice? We have created a lot of things in the West over the last centuries. Unlike the Chinese, what is that we do not know?

two questions two personal answers:

-Open source be it hw or sw is a filantropic thought that works in societies where the salaries are good/great, a worker can have free spare time to work on it and have hobbies and the subject serves no or small economic interest for companies.
Example after WW2 my country was only in a built up and restore the country mode with 6 workdays almost no holidays and low wages. Today we work 5 days or less in parttime have enough salary to cope with that and have 35-40 holidays a year. China is just leaving the built up state IMO where almost anybody still works long days and need to make enough money to live.
In our society OS does work, I am building an OS HW CNC machine, still costs $9000+ so it never said that OS is cheap. Also look at OpenPNP for OS SW, if more people would contribute it would even be further ahead. But OS is luxury something the chinese are not yet ready for, hell they even scratch the surface of the ics to prevent their neighboor from copying the design.

-What we here in the West do not do is create thousands of small businesses and government support them by providing cheap rent housing in industryparks. Yes we have one person businesses which are often only consultants or it jobs related, only brainpower not creating hands. We have almost no single or two person businesses making small machines and then hundred a year. Perhaps because we do not have hundreds of small businesses that need them. We have a few large companies that will place parts on pcbs , they serve big companies and charge abundant so they can pay their parttime holiday enjoying employees and their expensive machinepark and high rent.

[Smallsmt]

Namely, why can not work Open-source hardware approach in practice?

Because it's a small market, complicated machines and software, high development costs so why should a company do and open HW / SW design???

Why do people think they get everything for free?
Who pay for our work?

The people who buy these machines never spend money for a professional machine.
This is not our market.

How many developers working on OpenPNP software and how many people using it?
But most of the users do not share the work on the open source projects.


Regards
Michael

[mikeselectricstuff]

Namely, why can not work Open-source hardware approach in practice?

Because it's a small market, complicated machines and software, high development costs so why should a company do and open HW / SW design???

Why do people think they get everything for free?
Who pay for our work?

The people who buy these machines never spend money for a professional machine.
This is not our market.

How many developers working on OpenPNP software and how many people using it?
But most of the users do not share the work on the open source projects.


Regards
Michael

Exactly. There simply aren't enough people interested. OpenPnP is essentially a one-man project. And the software is the easy part. 
A few people are looking at feeders with 3d printed parts but not sure if anyone has made serious progress.

[IconicPCB]

And the prices need to be what they are simply because the cost of a liter of petrol is what it is.. and more importantly a loaf of bread is what it is.
And so in order to be able to afford bread and petrol the developers have to be paid a decent wage...just like you and all the people around you.
Otherwise the sacher torte will no longer be available to the Viennese and the Danube will no longer be blue.

[rx8pilot]

Anyone that struggles to understand the costs involved in producing a low-volume P&P and developing an effective assembly line has a lot to learn before they should expect success in this area.

[mrpackethead]

Anyone that struggles to understand the costs involved in producing a low-volume P&P and developing an effective assembly line has a lot to learn before they should expect success in this area.

**[irony - sarcasm]***
But the dream is still alive and strong.   Its just a few stepper motors and and ardunio surely?
**[/irony - sarcam]***

[Simon]

Anyone that struggles to understand the costs involved in producing a low-volume P&P and developing an effective assembly line has a lot to learn before they should expect success in this area.

But the dream is still alive and strong.   Its just a few stepper motors and and ardunio surely?

Well good for you for volunteering, when will you be presenting your first open source PnP machine to the world?

[Dubbie]

Simon, I think you missed a tongue planted firmly in a cheek there...

[Simon]

yea, I did wonder about that after posting but decided I didn't mind being put right as you have.

[aandrew]

4/ Recognise that the software is possibly the most important part. I suggest you engage with openpnp and make your entry level machine fully compatible with that, or find a developer who can contribute your input to that project to ensure it works with what you offer. If you can do this you might just find the maker community want to embrace your machine.

This is the single biggest thing smallsmt can possibly do to win over a large customer base.

Make your machine 100% compatible with openpnp. Give them the protocol specs, or better yet, hire an OSS dev to add (and maintain!) support for your product lines in openpnp.

This will be a big cost savings to you, and really stand out in this area.

I too am a potential customer. I'm tired of trying to work with assembly houses for smallish prototype runs. It'd be so much easier to roll my own, so to speak. I'd love to see blog entries about setting up a machine (what it needs, how to tune, common issues and how to resolve them, etc.). I'd like to see comparisons between your low, mid and high end models (is it simply number of heads and speed, or do the  high end models also have better accuracy/repeatibility?)

Lots of things for me to learn, but making your system "just work" with openpnp would be a huge, huge bonus for someone like me.

[mrpackethead]

4/ Recognise that the software is possibly the most important part. I suggest you engage with openpnp and make your entry level machine fully compatible with that, or find a developer who can contribute your input to that project to ensure it works with what you offer. If you can do this you might just find the maker community want to embrace your machine.

This is the single biggest thing smallsmt can possibly do to win over a large customer base.

Make your machine 100% compatible with openpnp. Give them the protocol specs, or better yet, hire an OSS dev to add (and maintain!) support for your product lines in openpnp.

This will be a big cost savings to you, and really stand out in this area.

I too am a potential customer. I'm tired of trying to work with assembly houses for smallish prototype runs. It'd be so much easier to roll my own, so to speak. I'd love to see blog entries about setting up a machine (what it needs, how to tune, common issues and how to resolve them, etc.). I'd like to see comparisons between your low, mid and high end models (is it simply number of heads and speed, or do the  high end models also have better accuracy/repeatibility?)

Lots of things for me to learn, but making your system "just work" with openpnp would be a huge, huge bonus for someone like me.

SmallSMT has a large team of developers and engineering staff ( Micheal described his team to me, when i was contemplating buying a machine from him ) so, it shoud'tn be too hard.

[Gary.M]

I think the Chinese use their software as a fence around their hardware design. Scared that if the two are not tied together someone will just copy their hardware and takeaway their place in the market. This is a barrier to progress, particularly at the affordable entry level. I don't believe they think in a way that would let them do anything else.

[Simon]

Fact is if someone wants to copy it they will. So either they have really hot software that everyone wants and in some way cannot be copied (impossible?) or they do as suggested and make good hardware that everyone know's how to drive making them the defacto machine. I have contemplated what it would be like having my own machine and for the moment will leave it to assembly houses.

[Gary.M]

Fact is if someone wants to copy it they will. So either they have really hot software that everyone wants and in some way cannot be copied (impossible?) or they do as suggested and make good hardware that everyone know's how to drive making them the defacto machine. I have contemplated what it would be like having my own machine and for the moment will leave it to assembly houses.

Following the work being done to try to reverse engineer the software for the Charmhigh Chmt36v and then to make it work with Openpnp would suggest otherwise.

[Smallsmt]

This is the single biggest thing smallsmt can possibly do to win over a large customer base.

Make your machine 100% compatible with openpnp. Give them the protocol specs, or better yet, hire an OSS dev to add (and maintain!) support for your product lines in openpnp.

This will be a big cost savings to you, and really stand out in this area.

You assume that OpenPNP is a usable software solution.

[Simon]

isn't it?

[Kjelt]

I think the Chinese use their software as a fence around their hardware design. Scared that if the two are not tied together someone will just copy their hardware and takeaway their place in the market.

You can abstract and create the glue layer between OpenPNP and the machine in such a way that it does not run on other machines, eg you give the customer the choice to use the companies software or the OpenPNP software. But as SmallSMT already hints at, OpenPNP is far from a mature P&P software. Look at the technical debt on their website and you realize quickly this is far from reliable, perhaps usable for hobbyists but not for a company that depends on a timeboxed finished product IMO.

[Kjelt]

isn't it?

Not for companies. Look at their technical debt and the timeperiod they tackle and solve it, if your bug is being fixed within a year you are lucky.

[Simon]

Yep, that is why i bought a commercial E-CAD solution, in the hope that I would not be at the mercy of hobbyists.

[mikeselectricstuff]

isn't it?

Not for companies. Look at their technical debt and the timeperiod they tackle and solve it, if your bug is being fixed within a year you are lucky.

The whole point is that with OS, you can solve problems yourself, and adapt it to your needs if you want.

[Koen]

"You can solve problems yourself" now and move on. No need to wait for the manufacturer's new revision. Horses for courses.