Tips for Replicating Microchips

[abritinthebay]

Hey all,

Anyone have any suggestions as to how I would go about making a replica of a microchip? Obviously just a visual replica. I can't find any chips that are the same shape or size, which isn't surprising - as these are 36 years old now!

I'm thinking some kind of acrylic plastic cut to shape, but the texture of that is often too smooth; and that still leaves me out in the cold with the pins.

In my case the chips look a lot like 36 legged spiders - with the pins on the side like legs, rather than underneath the chip - which is why I need the pins to show. (they are also arranged down each side so 9 are longer than the other 9 in an alternating pattern, but that's less of a problem)

Anyone have any thoughts? Would be helping me a lot.

 

[exoray]

Well since I suspect you are trying to replicate the vintage Sanyo chips in the GB calculator circuit...

I would consider some SQUARE brass wire like this

http://www.brandywinejewelrysupply.com/cybermerchant/scripts/prodList.asp?idCategory=88

I don't have my calculator handy but 21 gauge is probably pretty close to the right size, I will verify that later for you, 18 might be closer...

For the chips I would probably just go with a white sheet of styrene, some dry or wet sanding with 200-400 grit sand paper will get you close to the ceramic finish of the real chips...

Short of that castings made with white resin and ceramic dust will probably give you the closest representation of the finish...

 

[ScooterinAB]

Tee he. I'm also trying to figure out a good way to build a GB belt circuit.

I was originally thinking IC sockets with the tops puddied in with either bondo of Green Stuff (expoy). There were a few things that stumped me though.

If we could get accuate measurements of the legit chips, that might halp with finding subsitutes.

Short of that castings made with white resin and ceramic dust will probably give you the closest representation of the finish...

That creates interesting ideas in my head. I've been wondering if a belt circuit would be even mildly profitable. Adding in resin cast chips could be interesting.

 

[exoray]

If we could get accuate measurements of the legit chips, that might halp with finding subsitutes.

I have provided them on another forum...

The white chips are 23mm tall by 16.5mm wide (excluding legs) Metal window cover ones are about 1m thick while the non windowed ones are 2mm They are elevated off the board give or take 3mm

http://www.gbfans.com/community/viewtopic.php?f=4&t=4156

The chips were a unique design, even for their time... I doubt you will find many substitutes that are even close...

That creates interesting ideas in my head. I've been wondering if a belt circuit would be even mildly profitable. Adding in resin cast chips could be interesting.

With diligence the real calculators can be found, and even found cheaply... Trying to build a good replica will probably push yourself past what the market value will allow... Now you could probably get some decent builds done within a decent price but as always compromises will need to be made to keep the price low... But, since the stunt ones were pretty ugly you should be able to do better then that for a decent price if you tried...

 

[Rik1138]

I found this site showing how someone did a complete scratch build:

http://www.tk2441.co.uk/gb-belt.htm

 

[abritinthebay]

Yup, the Sanyo.

I've worked out I can produce a copy of the PCB for $10-25 (depending on if I can get more people interested in one). So the PCB would actually be simple.

The other components are basic resistors, capacitors, and a few transistors. Nothing expensive. They are all made today too, and cheaply.

That brass wire sounds perfect, though attaching the wire to the plastic so it looks right might be a real pain in the ass. It's not like superglue would look right!

I'm not actually a fan of that reproduction - it looks exactly like what it is - a bunch of paper glued onto plastic.

The IV-9 Nixes are as close as we're going to get, but he didn't even use the right brand of orange capacitor (which is still made!) and as I said - it looks like paper. None of the depth and shimmer of a real board.

It will do for a Halloween pack, but I'm surprised he used real nixies - and didn't just go for fairy lights - concidering the rest of it.

It's all in the details

 

[exoray]

but he

She...

I've worked out I can produce a copy of the PCB for $10-25 (depending on if I can get more people interested in one). So the PCB would actually be simple.

On the correct type of phenolic PC Board? I have looked into it but in the end it just became too cost prohibitive to do "good" reproductions... I only paid $35 for one of my real calculators, paid a little more for the second but still within reason...

 

[abritinthebay]

She...

Whoops, my bad. Still, the rest of my point stands.

On the correct type of phenolic PC Board? I have looked into it but in the end it just became too cost prohibitive to do "good" reproductions... I only paid $35 for one of my real calculators, paid a little more for the second but still within reason...

The correct board material (from what I can find out by badgering people at Sanyo) isn't actually made any more. The closest thing possible in modern pcb plants is FR2 as a paper/epoxy combo.

However all small run places won't do FR2 as it is a) brittle as all hell, which is why no-one uses it anymore and b) expensive as all hell because no-one uses it any more.

FR4 is the modern alternative, but it doesn't quite have the same finish it is true - as its a glass/epoxy resin. This is a shame, but considering the alternative (screen printing, no copper finish, incorrect lacquer etc), it's an acceptable sacrifice.

I would prefer FR2 or a similar paper/epoxy resin for the board - and I am investigating if that is possible right now - but for a v1 of the replica FR4 will certainly do.

Plus, I have no idea where you are getting your calculators from.. but damn that is a cheap price! Most of them run $150-300 now - collectors have worked out that there is a cult need for them. You can find similar ones for much less, but... what would be the point in that?

 

[exoray]

FR4 is the modern alternative

Yeah I know, it's what the boards I have made are, and it's entirely the wrong look...

You can still source the phenolic boards, but you will probably have to etch them yourself as you said most PC board houses won't do them any longer... But, it's not out of scope to do it yourself...

Plus, I have no idea where you are getting your calculators from.. but damn that is a cheap price! Most of them run $150-300 now - collectors have worked out that there is a cult need for them.

Diligence... Most people simply don't stick to it, they do a quicky search here or there and then forget about it for 6 months or until someone brings it up, you need to devote time to constantly looking...

 

[abritinthebay]

Yeah I know, it's what the boards I have made are, and it's entirely the wrong look...

You can still source the phenolic boards, but you will probably have to etch them yourself as you said most PC board houses won't do them any longer... But, it's not out of scope to do it yourself...

Well... to do it to the required detail and have it not look like pickled ass... is pretty tough.

We're talking about tracks that are around 0.01" inches here - printing a solid consistent mask at that level needs really good gear. Plus once you factor in the etching chemicals, the extra time, testing copies, the copper board, the solder mask, the routing... well you've really not saved any money. It's actually a lot more expensive.

Prototype PCB makers will always produce a better quality board that you can at home, and for a lot cheaper (you can't do that for less than $20 a unit at home).

Diligence... Most people simply don't stick to it, they do a quicky search here or there and then forget about it for 6 months or until someone brings it up, you need to devote time to constantly looking...

I was looking for 12 months. I found 4 come to market. 2 of those were the LED type and the other two hit $350 and $280. There just aren't that many out there any more. Especially of the nixie type.

Plus as you have mentioned yourself elsewhere - unless you open it up before hand you may get one of the later revisions that aren't screen accurate.

So overall I think designing this board using FR4 is a good start - and then moving it to FR2 if possible is a better plan that will result in a more accurate, better quality, much more cost effective prop.

But the quest continues...

 

[ScooterinAB]

I think it's time the community looks to make replica parts. Nixies can be found on ebay for cheap (not exact, but no one but a diehard ghosthead will know anyways). Similar chips or styrene remakes/resin cast is possible and very inexpensive. As already mentioned, the other resistors, small capacitors and such and a few buck at most.

I'd also bee thinking of getting a board etched rather than doing the snick print idea (like the site Rik found. Good find btw). I think the difference between a passable prop and a good prop is to replicate as many parts as possible. A styrene board with a circuit "sticker" on it will pass, but an actual etched board which can be soldered to will be far better.

On that note, I'd be up for purchasing an etched board if someone made a run. I'm positive that there'd be a market for it (especially on GBFans). I'm doing some promos for an event that I work with, and I wanted to rebuild the belt gizmo as my slap together one has taken some battle damage. PM me if there is a "print" run planned.

 

[abritinthebay]

Agreed completely - especially about the part that the GB propping community needs to take charge. The people who ARE about prop making in GB are few and far between right now - and have mostly become isolated from the rest of the community due to being completely bored with the rest of the crap that goes on.

Leadership time

I'll be ordering a run of the PCB's in the next month or so (two tops, depends on cashflow) so I'll let you know when I do if you are interested.

It only gets cheaper (per board) if more people are interested. I'll post an interest thread on here and a couple of other places when I get to that stage. It would be nice to get down to the $10 a board level, but I'd need at least 25 people to get close to that. I don't know if enough people would be interested.

We'll see!

(random interesting point about the boards - no screen printing. It's standard on most boards now, but in the Sanyo the labels and text are instead in copper).

 

[ScooterinAB]

Interested.

Few questions. 1) Will they be cut into the circuit board shape, or will they come as a rectangle? 2) How much do they cost prior to 25 people (just as a comparison).

You should also post it over at GBFans. I'm sure that betwwen here and there you can get 25 people, or pretty close.

 

[abritinthebay]

Interested.

Few questions. 1) Will they be cut into the circuit board shape, or will they come as a rectangle? 2) How much do they cost prior to 25 people (just as a comparison).

You should also post it over at GBFans. I'm sure that betwwen here and there you can get 25 people, or pretty close.

I'll do that when I'm closer to ordering a batch. I want to get the pcb layout as close as I can first. But yes, certainly a good idea.

The price varies on quantity. For one unit its about $78, 3 are $34 each, 10 are around $23 each.. etc.

As for the shape - yes, they'd be the correct shape (including all the rounded corners the real board has). Pre-cut. Depending on the cost though I might not get the 4 holes in the PCB drilled (not the pads, they are drilled) as for some reason that seems to up the price a lot. That's about 10 seconds work with a dremel though - and I'd silk screen the positions on if that was the case.

 

[exoray]

Quick question what you are you using for the top side of the board artwork?

Are you simply using my bottom side of the board image flipped, or are you building with reference of a real board?

 

[abritinthebay]

Quick question what you are you using for the top side of the board artwork?

Are you simply using my bottom side of the board image flipped, or are you building with reference of a real board?

No I'm using photos of the correct board (ie - not the scanned image of the bottom you posted on gbfans) that doesn't have the black chips on the side but rather the black transistor and resistor clumps.

The scan is great for reference to see how they re-routed stuff with the chips though. Very useful.

Some of the top part under the nixies and the orange capacitors will be educated guess work, but I should be able to come up with something suitable for there - and it'll all be hidden anyhow in a final build.

Obviously as I don't have a real board in my hand the photos I have are all I have to go on. Though they are great. If I did have the correct board I'd bend a few components out of the way and take some more shots to make as much of it correct as I can.

 

[exoray]

No I'm using photos of the correct board (ie - not the scanned image of the bottom you posted on gbfans) that doesn't have the black chips on the side but rather the black transistor and resistor clumps.

Hmm where did you find these photos? To the best of my knowledge the door poster is the best image anyone has found, and it's grainier then all hell hardly any use... I have heard rumors of the ideal board being found but to this date have never seen any confirmation, Spooky has a very close one that might be the real deal but it's hard to confirm...

 

[abritinthebay]

Hmm where did you find these photos? To the best of my knowledge the door poster is the best image anyone has found, and it's grainier then all hell hardly any use... I have heard rumors of the ideal board being found but to this date have never seen any confirmation, Spooky has a very close one that might be the real deal but it's hard to confirm...

Well the photos aren't of a screen used board, but the poster image clearly shows that it's the black transistors. This is especially obvious if you do some image enhancement - sharpening, some color balancing.

It's grainy - but the obvious part comes from the way the light hits the components, it's a lower black section of multiple components, not the whole side of flat shiny ICC's that are on the later models. I'd post a picture, but it's really obvious on the gbfans photo of it for example - http://www.gbfans.com/equipment/uniform/calculator-circuit-board/

 

[exoray]

Well the photos aren't of a screen used board, but the poster image clearly shows that it's the black transistors. This is especially obvious if you do some image enhancement - sharpening, some color balancing.

It's grainy - but the obvious part comes from the way the light hits the components, it's a lower black section of multiple components, not the whole side of flat shiny ICC's that are on the later models. I'd post a picture, but it's really obvious on the gbfans photo of it for example - http://www.gbfans.com/equipment/uniform/calculator-circuit-board/

Yeah I know the differences...

As someone that was going down the road like you, I put it on hold because I couldn't get enough "good" reference to do it like I wanted...

For me, and I'm sure you are starting to find out, the cost to do it just kept going up, and up... And at least for me the only way I felt it would be a worth while venture was to nail it as close as possible to maximize sales or else I saw the project falling apart...

 

[abritinthebay]

Yeah I know the differences...

As someone that was going down the road like you, I put it on hold because I couldn't get enough "good" reference to do it like I wanted...

For me, and I'm sure you are starting to find out, the cost to do it just kept going up, and up... And at least for me the only way I felt it would be a worth while venture was to nail it as close as possible to maximize sales or else I saw the project falling apart...

Working in software (web development mainly) has taught me that things will never be perfect in practice on the first go - iterative development is key here. Versioning.

Unless some kind soul will lend me the correct type of board as a reference - this is as close as I can get for now. Version One.

If I get more reference material - or a better pcb material - then I'll revise the design and specifications. Version Two.

According to my pcb manufacturer FR2 and FR2-like materials are not possible for them to come by - and none of their suppliers even stock anything close. They're a big shop, and so are their suppliers, so it may be that paper/epoxy is a thing of the past.

Sad, but I won't cry over it.