A500 trapdoor memory: free design

[Mq]

Hello again guys. Today I built RTC, and it works good. I attach a picture:-)

[PeteAU]

Originally Posted by Mq666

Hello again guys. Today I built RTC, and it works good. I attach a picture:-)

Are you planning to make & sell these on an online store? Just my hunch
Use a RTC-72421 & remove the crystal, they're calibration free and more accurate.

[Mq]

PeteAU: I respect it's Your project (with Mathesar ofcourse).
I do not want to do big business on Your good job:-)
Since we already talk about it, I ordered 10pcs PCB (the smallest possible quantity), what is normal, I suppose.
By the way I would never find out why my extension does not work if I didn't build few of them parallely.
So I would like to finish all complete extensions and some of them I will keep for my purpose, and rest of them probably will go to my friends.

I would like to do one board with SOJ sockets to testing chips, because of my earlier issues with them.

[demolition]

Originally Posted by PeteAU

Use a RTC-72421 & remove the crystal, they're calibration free and more accurate.

Will they work as a drop-in replacement of the original RTC chip?
Edit: Looks to me like it should be compatible if you bend up pins 16+17 and connect them to pin 18 which will disconnect it from any external crystal.

[Daedalus]

Originally Posted by demolition

Will they work as a drop-in replacement of the original RTC chip?

Yep, they're pin and function compatible with the Oki M6242B found in the A500+ and on the A501. The system will pick it up just fine and you don't have to worry about tuning oscillators or any of that faff I've tested them myself in a Zorro-II RTC card and it "just works".

[demolition]

Originally Posted by Daedalus

Yep, they're pin and function compatible with the Oki M6242B found in the A500+ and on the A501. The system will pick it up just fine and you don't have to worry about tuning oscillators or any of that faff I've tested them myself in a Zorro-II RTC card and it "just works".

So even though the data sheet states that pins 16+17 should be at Vdd potential, it will work even if they are not?

[PeteAU]

Originally Posted by Mq666

PeteAU: I respect it's Your project (with Mathesar ofcourse).
I do not want to do big business on Your good job:-)

No worries. They can be sold for profit, as long as you release the modified design if somebody asks.

Originally Posted by demolition

Will they work as a drop-in replacement of the original RTC chip?

Kind of, you need to remove the crystal and 2 capacitors. Besides that, yes. RTC62421 is also good - they're made by EPSON and 100% compatible.
Edit: Make 16+17 pins NC, the chip connects them to Vdd, you don't have to.

[Daedalus]

Originally Posted by demolition

So even though the data sheet states that pins 16+17 should be at Vdd potential, it will work even if they are not?

Yeah, they work fine in my testing anyway without the crystal and associated capacitors populated. The datasheet is a little confusing, it lists those pins as Vdd, but they're actually outputs, possibly to do with making them useable in circuits that already have the crystal present? I don't know, but anyway, the key in the data sheet is a small line that says they are at Vdd potential and should not be connected externally (can't remember the exact wording...)

[demolition]

Originally Posted by Daedalus

Yeah, they work fine in my testing anyway without the crystal and associated capacitors populated. The datasheet is a little confusing, it lists those pins as Vdd, but they're actually outputs, possibly to do with making them useable in circuits that already have the crystal present? I don't know, but anyway, the key in the data sheet is a small line that says they are at Vdd potential and should not be connected externally (can't remember the exact wording...)

I wanted to use them in existing stuff without doing any other modifications, so I guess I'll just snip off those two pins to leave them NC.

[Mq]

Originally Posted by PeteAU

No worries. They can be sold for profit, as long as you release the modified design if somebody asks.

In that case, I have to think about it:-)
Actually I have some small ideas to improve PCB design, I might correct mistakes, and other odds and ends. Maybe I could to build extensions for people that can not do it themselves. But I have no time to do that, because of my own other projects I do... Maybe in the future...

[Daedalus]

Originally Posted by demolition

I wanted to use them in existing stuff without doing any other modifications, so I guess I'll just snip off those two pins to leave them NC.

Yeah, might be the safest approach since they don't seem to serve a real purpose. Just looked it up, from the datasheet:

The (VDD) pins are at the same electrical level as VDD. Do not connect these pins externally.
The N.C. pins are not connected internally. Ground them in order to prevent noise.

[sampedenawa]

Just for discussion's sake, I build three boards + Gary Rev.C up to now, with the following results:

A500 Rev6a modded for 1MB Chip

Board #1: 4 chips, so 1MB Chip + 1,5 slow: random fill errors when checking full range memory, no errors when checking single 512k regions one at a time.

Board #2: 3 chips, green screen on boot (probably a broken chip, have to find it out)

Board #3: 3 chips, everything's ok

I have now to test all of above on my Rev5 board, then find out why Board #2 goes green screen.

BTW: I ordered 10 main board pcbs + 10 Gary RevC, but I am not planning to use them all, so if someone could be interested within a few weeks (give me time to finish my tests) I would sell all the remainders at cost (say, 5€ each).

BTW#2: maybe it is OT here, but I just wanted your opinion on the following: I am now testing the final proto of (yet another)flashkick switcher. It, as many others, can switch KS by holding CTRL+A+A but mine is different (as far as I know) because it preserves the choice upon power off. So next time you power on you find the last KS you selected (it is done by an AVR micro). At the moment I have a fully working board, and I am trying to make it flashable from the Amiga (currently I flashed the chips externally).
Do you think it colud be of any interest (so I will maybe start a dedicated thread), or is it of no interest for anyone other than me ? Thanks

[Mq]

I think in board #1 You have errors because Your 6A is modded and because of it both chiprams are in the same address area. To use 4 memory chips, motherboard must not be modded. On motherboard You have to cut JP2 only.

[sampedenawa]

Originally Posted by Mq666

I think in board #1 You have errors because Your 6A is modded and because of it both chiprams are in the same address area. To use 4 memory chips, motherboard must not be modded. On motherboard You have to cut JP2 only.

Good point, I do not have the board inf front of me now, but I feel you could be right. I'll check and report to the forum. Thanks.

[Mq]

On board #2 it could be broken chip, but it could be short circuit between some pins on the chips too. Check it first.

Finally with board #3 I don't know what to do, I have no idea;-)

[katarakt]

Originally Posted by sampedenawa

BTW#2: maybe it is OT here, but I just wanted your opinion on the following: I am now testing the final proto of (yet another)flashkick switcher. It, as many others, can switch KS by holding CTRL+A+A but mine is different (as far as I know) because it preserves the choice upon power off. So next time you power on you find the last KS you selected (it is done by an AVR micro). At the moment I have a fully working board, and I am trying to make it flashable from the Amiga (currently I flashed the chips externally).
Do you think it colud be of any interest (so I will maybe start a dedicated thread), or is it of no interest for anyone other than me ? Thanks

Sure!! Very good, i have such a thing in my A600 but without possibility to flash . Do you have pictures? Maybe you can send me PN

[sampedenawa]

Originally Posted by Mq666

I think in board #1 You have errors because Your 6A is modded and because of it both chiprams are in the same address area. To use 4 memory chips, motherboard must not be modded. On motherboard You have to cut JP2 only.

Great man Mq666! It was it ! I shorted jp7 again and board #1 is 100% perfect in all tests !

[sampedenawa]

Originally Posted by katarakt

Sure!! Very good, i have such a thing in my A600 but without possibility to flash . Do you have pictures? Maybe you can send me PN

Sent you a PM with a picture of the actual prototype board. Please don't mind of scratches and bad solderings (I soldered and de-soldered several times on it, so it's quite a mess). It is anyway a fully working board (except that you have to flash the two 29F040 chips with an external chip programmer atm). I still have 9 pcbs atm, should you be interested in 1 of them to beta-test it for me. Details in PM, not to bother the forum here. If the stuff would get more interest by others, I will open a dedicated thread.

[PeteAU]

Hi,

I combined Mathesar's new gary adapter, redrawn by Mq666 into one board:

(links removed)

Changes:
- Added a jumper for RTC enable.
- Added a footprint for SMD RTC chip, so you can populate either one.
- Made jumper drills larger (0.8->0.9mm) per Mathesar's complaint.
- Gary adapter by Mathesar with flipflops for BLIT.
- Corrected JP4 error on gary adapter.

If you've never had PCBs made, it's dead easy:
- Visit your favourite cheap Chinese board house. e.g. https://www.elecrow.com/special-offe...pcb-10pcs.html
- Upload the gerbers ZIP file.
- Select 10x10cm, HASL and 1.2mm thickness.
Done!

[Mq]

I looked at the new revision, and it seems to be fine.
To know that everything is 100% ok, someone will have to risk ordering PCBs and build next one extension:-)

I think You should correct some elements values (another values on schematic and in description in the posts of thread) - especially C10 that is a resistor in fact:-)