EconoSimm: Really cheap custom rom for your SE/30. Open source!

[croissantking]

I think -10 is 100ns - too slow. Needs to be -85 or better. -8 and -5 are 80ns and 50ns.

not sure if I should share it here.

Are you planning to buy up all the stock?

 

[Phipli]

I think -10 is 100ns - too slow. Needs to be -85 or better. -8 and -5 are 80ns and 50ns.


Are you planning to buy up all the stock?

Depending on the use, I wouldn't say 100ns is too slow. In a IIci the stock chips are 200ns.

 

[Phipli]

Oh, you're talking about RAM.

Please start a new thread, it is a very different topic and causes confusion.

@Cory5412 - is it worth moving the last few posts to a new thread? The topic seems to have pivoted from desktop ROM SIMMs to Powerbook RAM somehow.

 

[croissantking]

This has worked great on my first attempt. I used the guide as posted on GGLabs website here: https://gglabs.us/node/2324. I had to install ‘srecord package’ which I downloaded here: https://srecord.sourceforge.net/. I installed it on my bootcamp partition and then ran the commands in a command prompt to split the .bin file. Then flashed each chip with my T48.

 

[croissantking]

An observation, my EconoSimms work nicely in my Reloaded boards with NOS ROM sockets with metal clips, but are very floppy in an original board with plastic tabs, and won't work there. I ordered the PCBs with a 1.2mm thickness as mentioned in the GitHub link.

 

[croissantking]

I tried tinning the 'fingers' of the simm with solder. It helps to make it slightly thicker. It may just work.

 

[GRudolf94]

Theoretically SIMMs are to be 50thou thick - i.e. 1.27mm. Since no cheap PCB fab will do that, 1.2 is close enough, and shouldn't matter, since slot contacts are springloaded. Interesting that you encountered one case where the .07mm makes a difference.

 

[zigzagjoe]

Theoretically SIMMs are to be 50thou thick - i.e. 1.27mm. Since no cheap PCB fab will do that, 1.2 is close enough, and shouldn't matter, since slot contacts are springloaded. Interesting that you encountered one case where the .07mm makes a difference.

I was wondering: In theory, by going to 2 oz copper from 1 oz copper on the front/rear, this would add 0.0347mm * 2 thickness, brings a 1.2mm board up to 1.27mm.

Not certain if this thickness is compensated for elsewhere in the process, though.

 

[Phipli]

I was wondering: In theory, by going to 2 oz copper from 1 oz copper on the front/rear, this would add 0.0347mm * 2 thickness, brings a 1.2mm board up to 1.27mm.

Not certain if this thickness is compensated for elsewhere in the process, though.

Generally their tolerances include all the options, so picking everything will make it a bit thicker, but not much above 1.2mm.

Check your fabs tolerance specs.

Also, I've heard that putting stickers where the socket grasps the simm helps hold it in the correct position better. I've considered putting a block of silkscreen on both sides where the clips hold to help with this.

 

[croissantking]

I was wondering: In theory, by going to 2 oz copper from 1 oz copper on the front/rear, this would add 0.0347mm * 2 thickness, brings a 1.2mm board up to 1.27mm.

Not certain if this thickness is compensated for elsewhere in the process, though.

At least on JLCPCB, you can't select an Outer Copper Weight of 2 oz with a 1.2mm PCB thickness.

I can confirm that tinning the contacts is the correct 'econo' solution.

 

[zigzagjoe]

At least on JLCPCB, you can't select an Outer Copper Weight of 2 oz with a 1.2mm PCB thickness.

I can confirm that tinning the contacts is the correct 'econo' solution.

Good catch - looks like it is only an option on 4 layer+ boards. I may be trying that to see what happens.