|
|
Post by leonski on Jan 7, 2022 2:37:46 GMT -5
I have a real good idea for going forward.
First? MOST parts are available for repairs for a long time period. Transistors? Capacitors? Resistors? Even surface mount stuff will be available for years to
come. Even the large PowerSupply caps can be had.
Nuts / Bolts are of course nearly forever. Case parts are rarely needed.
But ONE thing not stocked and NOT available are CIRCUIT BOARDS. One good zap might render such a board UnFixable. Carbon arc traces conduct
and are difficult ot 'fix'. I've bridged burned parts using wire Jumpers. Not pretty but it works and the machine I fixed was NEEDED.....
So? HERE IS MY IDEA.
Emotiva can maintain a FILE of board designs. They should be doing this, anyway. My solution is NOT easy or cheap, but getting easier and cheaper ALL the time,
Get a copy of hte board made with a 3D printer. It should even be possible to make a multi-level board with all the correct pads and traces.
Such a part will NOT be cheap. But it WILL be an exact duplicate of the original part....and made to the latest / best revision which EMO has on file.
Just a thought going forward. We might not be quite THERE yet, but certainly within 5 years this will be both affordable AND doable......
|
|
|
|
Post by vcautokid on Jan 7, 2022 5:40:53 GMT -5
It is a great idea but the reality of this is pretty low. It would be cool if they would do this though. Especially for legacy products where it is almost impossible to get the those boards, but I think you probably wouldn't get much on this idea.
|
|
|
|
Post by 405x5 on Jan 7, 2022 8:39:23 GMT -5
|
|
|
|
Post by marcl on Jan 7, 2022 8:59:43 GMT -5
I worked for Merck for a couple years before retiring, and one of our monthly team meetings was held at the facility in NJ where they have a 3D printing lab. It was the most fascinating, nearly sci-fi thing I've ever seen! Okay maybe I COULD have known what could be done with 3D printing, but I didn't. I remembered the earliest days in the mid 80's (yeah, EIGHT-IES!) when I could get prototype plastic parts made with a photolithography process that was very expensive, and limited to very few materials ... not to mention VERY slow computers. But this was just amazing.
You might wonder what a pharma company needs with 3D printing. But for example: model of a nasal passage with human cells printed on the inside to evaluate propagation of nasal sprays; an emergency part replacement to get a production line up in two days, when the manufacturer said it would take two weeks; pills with imbedded printed circuits that transmit when dissolved in the body to confirm a patient took the medication; etc.
And then the combinations of complex tools, multiple material and color models of parts and molecular structures. They guy said a transplantable human liver may be only 10 years away.
|
|
KeithL
Administrator 
Posts: 9,941
|
Post by KeithL on Jan 7, 2022 11:04:55 GMT -5
This is a nice idea - but isn't really practical with the current technology. This is in fact true for most things. For example, even though 3D printing of plastic parts is easy, and is commonly used to manufacture prototypes, it is rarely done for replacement parts. Odds are that, if you purchase a replacement knob for your radio, or a replacement fender for your recent model car, it has probably NOT been 3D printed.
(And, while 3D printing is occasionally used for actual production, this is still the exception... because it is usually far more expensive than other readily available options.)
With current technology, when making a multi-layer circuit board, you cannot simply "load in a pattern and print it". There is significant machine set up and configuration for each individual board. There are actually machines that can "custom wire" relatively simple boards whose layout isn't especially critical...
But the processes they use don't work very well for boards that handle high frequencies where component and foil layout is critical... (Which includes most of the boards in most modern digital products.)
There are also in fact machines that can do "one off production of high-precision multi-layer circuit boards". However, the cost of the machinery involved, and of maintaining it, is very high... (There are companies who do this for a living, as a service, so it can be done.)
Considering current trends what you suggest may in fact be practical in the next five to ten years... But for now the economics just aren't there...
With the complex boards used on most modern gear, the production of circuit boards is a complex process, and is largely automated. The board itself consists of many layers of metal and circuit card material, each printed and aligned with remarkable precision, and bonded together. Connections between the layers must then also be done with really impressive precision.
Most of the tiny surface mount parts, as well as the majority of other parts, are then placed and soldered by machine... And this is done in batches, by very complex and expensive machines, which must be configured for each specific type of board.
Specific parts that are difficult to install or solder automatically are then added or adjusted by hand... There is a huge difference between placing or replacing a few parts by hand and loading and soldering hundreds or thousands of parts by hand.
If you were to look at some of the more complex PCBs used in some of our processors... And assuming that we could have boards "one-off printed", or simply maintained a stock of spare blank boards, and had the parts already on hand... The cost to populate a single board with all of those parts, solder them, and test the finished board, all by hand, would be hundreds or thousands of dollars in labor alone. With batch production you might be talking about a multi-million dollar machine, that costs thousands of dollars to set up for a particular board, but can then produce as many boards as you want for $50 each.
With an ongoing product, which is still in production, you can simply "order some replacement boards along with the next batch". But, when production ends, or in a limited production run, it is far more cost effective to simply purchase extras, based on the number you expect to need.
I have a real good idea for going forward. First? MOST parts are available for repairs for a long time period. Transistors? Capacitors? Resistors? Even surface mount stuff will be available for years to come. Even the large PowerSupply caps can be had. Nuts / Bolts are of course nearly forever. Case parts are rarely needed. But ONE thing not stocked and NOT available are CIRCUIT BOARDS. One good zap might render such a board UnFixable. Carbon arc traces conduct and are difficult ot 'fix'. I've bridged burned parts using wire Jumpers. Not pretty but it works and the machine I fixed was NEEDED..... So? HERE IS MY IDEA. Emotiva can maintain a FILE of board designs. They should be doing this, anyway. My solution is NOT easy or cheap, but getting easier and cheaper ALL the time, Get a copy of hte board made with a 3D printer. It should even be possible to make a multi-level board with all the correct pads and traces. Such a part will NOT be cheap. But it WILL be an exact duplicate of the original part....and made to the latest / best revision which EMO has on file. Just a thought going forward. We might not be quite THERE yet, but certainly within 5 years this will be both affordable AND doable...... |
|
|
|
Post by LuisV on Jan 7, 2022 14:21:48 GMT -5
|
|
|
|
Post by leonski on Jan 7, 2022 15:38:38 GMT -5
I have just LOOKED at 3D printing for an idea I have......which requires parts which I can NOT easily fabricate from wood or even metal....IF I HAD the proper tooling, even....
The 'trick' with my suggestion is to stay JUST AHEAD of what is easily possible... For example? A simple Copper on substrate SHOULD be easy to do.....THan finish
in conventional etchant. Only a few steps. Size board? Print with photoresist media? Etch? Clean? Drill? With the right DATA file, you should be able to make
these by the bunch. Print 'em 2x or 4x (depends on size) at a time than Mass Etch......If you need several prototype SIMPLE boards? Can-Do.
As for more complex boards? I saw a list of printable materials. ALL SORTS of stuff, including 316 Stainless.....which is $$$ and finished by a sinter process.
I don't know IF any printer will do 2 or 3 material layers IN REGISTRATION. That's the big issue. In photography? I can make multiple exposures of a scene. Even come back
the NEXT DAY for more images.... THAN manipulate the files. BUT? they all must 'LINE UP' PERFECTLY or it'll have color fringine or blur....I've done Panos...consisting of 3 ROWS of 4 IMAGES EACH.....
And it looks like ONE really neat exposure. And is a really HUGE and detailed file. My prof at school did on with what must have been 80 or 90 images stitched to a single image.
And Keith is right. Expensive TODAY? Less so tomorrow. Just like some other things you can name. Cost of Semiconductors is a fraction of years past.
My point in dragging EMO into my thoughts? LACK OF SUPPORT for older gear which may be servicable but for lack of a part. While most 'parts' are available, the REAL clincher is that
circuit boards may be stocked for a couple years than NOT replensished due to cost of fabrication than Storage. If you could send someone a FILE? It should be possible (not CHEAP)
to fabricate a board......Just noodeling thru the posibles.....
|
|
|
|
Post by leonski on Jan 7, 2022 16:00:44 GMT -5
Interesting process but NOT rocket science. I worked with Pick and Place machines back in the late 70s.....which were ALL in-house built and proprietary. Our company employed a German Master Machinist to fabricate while as a peasant, I helped keep 'em all working. If you really want to see some neat stuff? Check out THIS artifact video from the 80s. SKIP AHEAD to about 4:45 to see a bit of the interior of an implanter. That wheel? OVER 5' in diameter and spins VERY fast.....in an Excellent vacuum. Loading and unloading is highly automated and has few problems....ever......after proper adjustment. Applied Materials issues an alignment kit used by the techs.. the wheel has a capacity of 25 slices which is one full cassette of wafers. Balance is CRITICAL. www.youtube.com/watch?v=qCYhHq9mlEk |
|
|
|
Post by leonski on Jan 8, 2022 2:36:58 GMT -5
I just looked up circuit board prototyping.....
CURRENT standard involves something called a 'Gerber File' which is basically the 'map' of the board. It is written in ASCII code, apparently and is THE recognized standard.
From the look of this, a 2 layer board can be made.
Now? If some bright type could TRANSLATE such a file into whatever a 3D printer would recognize? Such printing could be done more easily....
I'm going to look into this, out of curiosity.
Yes, stuffing a SMD board is a real expense. Not trivial. And ordering extras during hte produciton run OF any given product? YOu have to store, sort and I think even have
TAX issues with inventory. So If you think you'll need 50 spares in the next 5 years? You GOT 'em and that expense.
this kind of non-fix ENCOURAGES obsolescence. Since even IF your board is good, repairing SMD is a PIA, many people would simply junk it out and start fresh.....Too bad.
|
|
|
|
Post by 405x5 on Jan 8, 2022 13:47:03 GMT -5
|
|
|
|
Post by leonski on Jan 8, 2022 17:26:07 GMT -5
MANY years ago I taught a semiconductor class. The layers in a planar device and the order in which operations were done.
One of my compatriots started using PLAY DOH. Comes in a bunch of colors and with a few simple tools......plastic knife and roller, to start,
you can create visually stunning and easily understood devices. Neat idea using toys as a teaching aid of serious stuff...
|
|
|
|
Post by MusicHead on Jan 10, 2022 20:20:26 GMT -5
😁😁😁   |
|
|
|
Post by leonski on Jan 10, 2022 22:43:13 GMT -5
cuople 'a things?
We made models of semiconductor DEVICES....not circuits. Our goal was to illustrate to new and NON-Technical
operators how what they did on hte production floor related to making the final device.
We would make a model of hte layes in, for example, an 'N' Type MOSFET transistor....
The Squishy 'kit' uses parts to make circuits which are ACTIVE.
|
|
KeithL
Administrator
Posts: 9,941
|
Post by KeithL on Jan 11, 2022 11:55:28 GMT -5
That is a really cool teaching tool. 😁😁😁 |
|
|
|
Post by leonski on Jan 11, 2022 17:19:57 GMT -5
This is CIRCUITS, and NOT Devices........
This would have done me NO Good whatsoever. I needed the students to SEE how, just for EXMAPLE...the layers in a GATE of a transistor work...
You have gate oxide, poly silicon and how it 'switches' on or off.....as well as where the 'source' connects and how the drain metalizatin effects certain measurables.
If I sat down prior to the class? I could even simulate IMPLANTS which are vital to the operation of semiconductors......
Making visuals at such large scale really helps.
Not at ALL what 'squishy' does.
|
|
|
|
Post by leonski on Jan 14, 2022 3:21:32 GMT -5
I worked for Merck for a couple years before retiring, and one of our monthly team meetings was held at the facility in NJ where they have a 3D printing lab. It was the most fascinating, nearly sci-fi thing I've ever seen! Okay maybe I COULD have known what could be done with 3D printing, but I didn't. I remembered the earliest days in the mid 80's (yeah, EIGHT-IES!) when I could get prototype plastic parts made with a photolithography process that was very expensive, and limited to very few materials ... not to mention VERY slow computers. But this was just amazing. You might wonder what a pharma company needs with 3D printing. But for example: model of a nasal passage with human cells printed on the inside to evaluate propagation of nasal sprays; an emergency part replacement to get a production line up in two days, when the manufacturer said it would take two weeks; pills with imbedded printed circuits that transmit when dissolved in the body to confirm a patient took the medication; etc. And then the combinations of complex tools, multiple material and color models of parts and molecular structures. They guy said a transplantable human liver may be only 10 years away. Some of what you suggest MAY now be easily done with computer modeling. That really speeds up the process from 'idea' to 'result' or even 'product'. Right now? A San Diego startup is heavily involved. Company? APTERA. they are making a very revolutionary ALL electric car which they claim can be had with a 1000, mile range. And weigh <2000lb empty. Some very unique features, but I'll let your curiosity lead you to their site. ALPHA cars, 3 of 'em are running around the county AND Jay Leno has driven one......He has a couple videos, as well..... |
|
|
|
Post by ashleyrplayer on Jan 20, 2022 11:57:44 GMT -5
It was my first time utilizing a multilayer PCB, and I'm glad I did because it has a lot of advantages, such as increased assembly density. Due to their inherent electrical properties, multilayer boards provide high capacity and speed in a smaller footprint. Connectors for numerous distinct PCBs are reduced or deleted, simplifying design and lowering weight even further. Multilayer PCBs are available in rigid and flexible configurations. However, keep in mind that the more layers on a flex PCB, the less flexible it gets. |
|
|
|
Post by leonski on Jan 20, 2022 23:15:50 GMT -5
I don't know what the limit is, but I fixed a computer for a friend that had a multi-layer board.
The power jack had 'wiggled' enough to no longer be a relieable connection.....than it broke outright....
Friend brought over the whole thing AND the new part. Desoldering was easy using solder wick and the 'fix' involved more heat than
I was comfortable using...but the solder DID a complete flow.
Fixed and still working a year and change after I did the job......
the worse fix? HEADPHONE CABLE which was 'pulled' out of the can. That wire is miserable to solder. TINY and with a nylon reinforcing thread.....
CAMERAS frequently use flexible interconnects. Assembly is very tight and complex....
|
|
