Programming the Lynx for a living - blank carts

[+karri]

Just kidding.

 

Most Lynx programmers have a desire to create a real cart. Manufacturing is a bit expensive and getting into this area also has some risks included.

 

Today I looked in chip prices and thanks to automobile manufacturers certain chips have dropped a lot.

 

A 2k eeprom (SER. EEPROM 24AA16T-E/OT) is 40 cents and a 512k rom 1.5€ (IC, SST39SF040-70-4I-NHE).

 

The PBC from China would just cost around 4€ (top + bottom). This should bring the base cost close to 7€/cart.

 

Which brings me up to the subject. In large quantities the prices are considerable lower than the prices stated above. But the minimum amount is 100 carts to get the assembly done also.

 

I bought some chips and kicked up Eagle to create a new kind of very low-cost cart for homebrew releases. No fancy plastics i'm afraid due to cost.

 

The basic idea is to glue two PCB's together. The lower PCB is just 1mm thick and on that I assemble two low profile chips. The upper PCB has holes for the chips and brings up the thickness of the cart to 2.6mm. After the halves are glued together you also need to shape the connector so that it does not break the Lynx.

 

 

The upper PCB would also have a long hole in the style of Songbird carts for easier removal. The resulting cart would be completely flat. The hole with the chips would be covered by a game label. The cart would be green due to cost.

 

The cart would be blank after assembly. For programming you would need a Raspberry Pi that can erase the cart and program it. On top of Raspberry Pi comes a small breakoutboard that contains a Lynx cart socket.

 

This is kind of a Kickstarter project. I will order a small set of carts (20 pieces) and programmer boards (10 pieces).

 

If someone feels the urge to join the project drop me a line. The goal is to create a set of 100 carts (or more) after the prototype operates properly. My reality check tells me that my creations will be counted in tens of carts - not hundreds.

 

This cart is probably the release platform for Stardreamer (which is slowly going "where no man has gone before" - again).

 

--

Karri

 

 

 

Edited March 14, 2015 by karri

[LX.NET]

just a quick response: Very cool and I would love to participate and contribute. I'll send a PM later this weekend.

[sage]

The idea with the RPI add-on sounds good.

[BadPricey]

I like the sound of your idea Karri. Am I right in thinking this is similar to the flashcart? How much do you think the programmer boards will be?

I'd be interested in supporting this.

Rgds

BadPricey

[+karri]

The programmer boards are completely free to people who plan to produce a new title for the Lynx.

 

(or people who help with graphics, music, play testing etc.)

 

Perhaps the lines above were a bit too harsh. My plan was to have an affordable route for real homebrew games like "Shaken, not stirred".

 

The component price is just a few euros for the programmer.

 

I was planning to split the project in lots like 1 programmer board + 10 carts. If we get 9 guys plus me interested then we could make a run.

 

The cost for a lot would probably be somewhere between 40 - 70 euros.

 

--

Karri

Edited March 16, 2015 by karri

[+karri]

I feel a bit silly now. The chips arrived and the thickness was not what I planned for. That stupid 512k chip is 3mm high. The only low variants are for 3.6V.

 

Well... The design has to change to move the chips to the outer edge in order to fit in Lynx 1.

 

 

Goodbye to a stylish flat cart

 

We still need 2 PCB's. The maximum thickness at the shop I was looking at is 2mm. To obtain a proper thickness I believe I could use one 1.0mm and another 1.2mm (or 1.6mm) board on top of each other. I wonder how thick a glue layer is?

 

 

 

[LX.NET]

hi Karri, can you tell a little bit more about the programmer board. Is it a programmer like a batronix device or an adapter kind of thing?

[+karri]

The easiest would in my opinion be a daughter board to Raspberry Pi. It would contain two SPI->16 bit chips to have enough wires for address and data. The OE and WR lines would be directly from the GPIO pins.

 

As the Lynx sockets are impossible to find and nobody creates 1/16 inch stuff today I planned to build the socket pin-by-pin directly on the programmer board. The cart would lie flat on the board while being programmed.

 

 

The contacts would be "stolen" from some existing connector and manually soldered on the programmer board. So you don't need to kill a Lynx.

 

Basically the programmer could be used to read, write and erase the content of both chips. The commands are probably

- programrom cart.lnx

- verifyrom cart.lnx

- eraserom

- readrom > cart.lnx

 

And the same for the rw chip

 

- program2k data.dat

- verify2k data.dat

- erase2k

- read2k > data.dat

 

 

Everything can be changed of course. So if someone has great ideas I am all ears.

 

I did think about sending in the cart design to get the PCB's later today. The chips are here already.

 

Programmer hardware is still just in my head.

 

Perhaps people interested could say hi?

 

Me - definitely

LX.NET - perhaps?

BadPricey - perhaps?

 

at least seven more to go...

 

This may be a single run project. Once the carts are out ordering more is exactly as expensive as ordering them now.

--

Karri

Edited March 16, 2015 by karri

[LX.NET]

Cool. I wanted to get myself a Raspberry Pi 2 anyway.

And I am in. Not "perhaps". At least for one, maybe for two/three batches of 10.

[BadPricey]

Hi Karri,

I like the sound of the project. I'm not what you'd call a programmer, just a bit of a tinkerer. I did try LX.NET tutorial but struggled. However I still think I can support this so those with more skills can benefit the Lynx. I should be able help with some costs. Fingers crossed you can get more supporters.

 

On a side note I think most people on here have 2 Lynxs, my Lynx 1 stays in a bag, if I'm gaming I play my Lynx 2 as I see at as more of a rough and tumble machine. If my brother comes round for comlynx he gets Lynx 2, Lynx 1 is all mine. What I'm trying to say is leave the design for now so that you can concentrate on other aspects of the project.

 

On another note. You say this is like a mini kickstarter project. Why not set tiers for supporters.

Level 1 Offer a special edition 1st run cart with a free homebrew title programmed onto it.

Level 2 supporters get 2 carts with homebrew programmed onto it.

Level 3 supporters get the dev kit, programmer board and 5 carts.

 

This would help you pay for the project

 

Just my opinion of course.

Rgds

BadPricey

Edited March 16, 2015 by BadPricey

[+karri]

Thanks for the ideas. May be more work for the extra tiers than doing the project itself...

 

I sent in the design yesterday so now it is just to wait for a while.

 

512k is a huge amount of memory. It would allow for small in-game videos.

 

While waiting for the boards I could give a face lift to "Shaken, not stirred" to make it more playable.

[Irgendwer]

I thought also about a simple to build cart some time ago. Maybe my conclusions help:

 

I had also the idea to have cut-out in the PCB for the chip, but I would go for a single board:

* glueing isn't fun and likely smelly

* boards can delaminate

* material costs are increased by two boards

so the cutout should surround exactly the dimensions of the eprom. This can be 'dropped in' then and

soldered to the pins on the edges. (My old design took classic ICs into account, where the pins had to be bended away to fake a SMD design.)

 

Thanks to a software preparation step of the cart image addresses and data (A5 doesn't have to be necessarily A5 at the EPROM, shuffle them to optimize the layout), even a single sided layout should be possible - reducing the costs further (like here: http://atariage.com/forums/topic/120824-lynx-eprom-cart/page-2?do=findComment&comment=2061170).

 

 

[+karri]

Thanks for the comments. Currently it is hard to find manufacturers that make boards thick enough for a Lynx. You can easily get 0.4mm, 0.6mm, 0.8mm, 1.0mm, 1.2mm, 1.6mm and 2.0mm. But the Lynx requires 2.8mm.

 

The cheap ones are 1.0, 1.2 and 1.6.

 

The price of a Lynx PCB is not terribly high. The two boards would cost less than 3 USD's together.

 

This also opens up for interesting designs. For a few more cents we could make the bottom board white and the top board black.

 

Yeah! This would fit my Startrooper Lynx perfectly

[Irgendwer]

Currently it is hard to find manufacturers that make boards thick enough for a Lynx.

 

My usual manufacturer ( http://www.leiton.de ) goes up to 2.4 mm - which fits quite well (esp. with label on top).

Order there 500 pieces in this material (60 x 58 mm, green) cost € 1.35 per board (double sided € 1.71, single sided black € 1.52, (600 pieces, single sided black € 1.34, to give an impression)).

 

Time to change your supplyer?

Edited March 19, 2015 by Irgendwer

[+karri]

Lol. I am struggling with 100 pieces.

 

When TailChao released Zaku it was like a miracle. Original cart style. Perfect game. Quality.

 

After that I have been wondering if it would be possible to create a cart that would look and feel genuine. It does not sound so difficult. A very low-height 5V flash chip and some plastic.

 

Well. My proto is getting further away from the goal all the time. Time to look for flip chips.

[Irgendwer]

Lol. I am struggling with 100 pieces.

 

Ok, I thought you would go for a somewhat bigger batch and place an order including carts for other developers too.

 

Best of luck!

 

Just an additional hint:

Esp. when creating a double sided design, you could also direct the signals to another edge - allowing to program the cart via an recycled isa slot part. No need to fiddle arround with single narrow contacts and springs. Additional contacts could be hidden under the label anyway.

Edited March 20, 2015 by Irgendwer

[LX.NET]

Great idea! Getting the connector from a Lynx requires a desoldeting station or a lot of patience. This could be a very good solution to that. The ISA port i mean. Doe a cartridge fit in thickness?

[Irgendwer]

Great idea! Getting the connector from a Lynx requires a desoldeting station or a lot of patience. This could be a very good solution to that. The ISA port i mean. Doe a cartridge fit in thickness?

 

If karri does the two board approach, I see no problem here: Signal board fits to ISA thickness - you could else 'abuse' an IC test clip. If space for the connector is an issue, PCI or the AGP bus is a lot denser...

Edited March 20, 2015 by Irgendwer

[+karri]

Sorry. I don't get it.

 

Instead of soldering in 32 pins on a PCB I would do what? Make two sockets on the cart and still solder in 32 pins in the programmer? Or even 96 pins if I was supposed to route the signals between the sandwich.

 

Long working day. Perhaps I read this again after dinner.

 

--

 

Still don't get it. I don't plan to kill a Lynx to get pins. The idea is to pull out pins with pliers from a small connector. On the programmer pcb I will make holes for the copper strips. With a little luck I will find suitable single copper strips that may be bent and soldered in the programmer board.

Edited March 20, 2015 by karri

[Irgendwer]

Still don't get it. I don't plan to kill a Lynx to get pins. The idea is to pull out pins with pliers from a small connector. On the programmer pcb I will make holes for the copper strips. With a little luck I will find suitable single copper strips that may be bent and soldered in the programmer board.

 

Yes, this plus the mechanic to keep the cartridge contacted to the pins and 'flat' sounds more complicated than simply routing them additionally to another edge to match a std. connector which can be used out-of-the-box.

Edited March 20, 2015 by Irgendwer

[+karri]

Thanks for the explanation.

 

If everything else fails I have a dead Lynx in the closet.

--

Karri

[GadgetUK]

I am interested in this as well! Great idea!

[+karri]

Great! I am just making the programmers board. Already got all the chips.

 

 

Today on my way home I got an idea for the connector.

 

As you can see on the programmer board every Lynx cart pin has 2 holes.

 

If I tie a knot on a braided copper wire any push it through one hole it can then run over a silicon stick on the component side and back through the second hole. At this time I solder it gently in place and cut the wire.

 

Every cart pin would then have a braided copper wire in the right place resting on an elastic piece of silicon. When I put he cart on this assembly all pins should make contact as the silicon is elastic. Yay!

 

 

[LX.NET]

if it is any help and we have a small set of people: I do have at least 3 loose cartridge connectors and a whole bunch of sacrificable dead boards. Would that help?

[+karri]

I could do both. What if I add both? The connector could be at the bottom of the cart. In this way you don't have to kill a cat to get the board to work.

 

I was also thinking about putting in a 3-color LED. Would this be overkill? It could have a smooth triangular intensity from black to read while programming. Light up green when the cart is ready.

 

Actually I have been using Raspberry Pi2 in our theatre to control lights and sound. While playing around I found out that Raspberry Pi could easily have full control over LED strips as well. This would not require any new chips just a 4-pin connector. You could then connect a small number of leds to this connection and enjoy a psychedelic light show synced to some mp3 tunes while programming the cart