I noticed a strange 1541 clone for sale on TradeMe over Xmas which according to the seller started smoking when he powered it up to test. I put a bid on it and it went for a song.
It duely arrived and I opened it up. There are no identifying markings on this drive other than PC-606. Googling that just shows the Trademe auction. The floppy drive itself is a Gold King GYE55A also give no results when googled.
First thing I noticed was a discoloured tantalum cap on the main pcb. No surprise there as they tend to short and since the PSU is a linear type it feeds more amps into the circuit until something fries.
Googling typical original 1541 faults showed that a 10uf in the 12v line usually shorts. I checked with a digital meter on continuity test and it indeed was shorted. After replacing it with a electrolytic and powering up the drive with the floppy cable disconnected got the board booting, with the drive light coming on red for a few seconds as expected.
I connected the floppy drive to the main pcb and powered it up, expecting the repair to be done. Low and behold more magic smoke appeared. This time it was a tantalum cap on the floppy drive itself.
I removed the melting cap and checked the 12 volt rail which still showed as shorted. Checked the rest of the tantalum caps on the pcb with the meter and found another 4.7uf cap was shorted. Removing that cap resulted in no more shorts. There were no shorts on the 5v rail fortunately. Replaced both caps with electrolytics I had to hand, and after re-assembling, the drive power up without loosing any magic smoke.
Strangely the power LED stopped working. After checking that the LED was getting power I noticed the green wire wasn’t even soldered properly as it wasn’t stripped properly. Must have only just been connected since it was made in 1985. Resoldered the wire and the power LED worked fine.
Put it all back together and connected it to my C64. Tried a couple of disks and its all working fine now.
It even comes up with the normal 1541 version on boot. I’ll dump the eprom to see if its different from a real CBM drive.
[Edit] I dumped the eprom. Its 1541-325301-01 joined with 1541-901229-05.
Seems there are still some questions regarding the extra pins (Konami-1 is 42 pins rather than 40 like a 6809e) and interrupts. Maybe some examining of boards with builtin Konami-1 replacements like Junofirst and Gyrus is necessary.
As far as I know Circus Chalie and Ropeman use this daughter board too.
Swapped the switching FET to an open collector 7406 (I didn’t have an LS handy) and I was able to reduce the delay in the read considerably.
Reading the directory from the SD card works which is cool.
As the SD/FAT code needs to run in the System24 bios, it all needs converting to 68k assembler. After a few evenings work, the code to interface to the sd card is written but still needs debugging on real hardware. FAT16/FAT32 (or maybe just FAT32) code to be converted next.
The beauty of having the code working in C is that I can just replace each function with it’s assembler version and test each one. (after realising I needed to create some trampoline functions to convert the stack passing method to a more typical assembly call method.. unless there’s a compiler function to use register passing.. I wonder..)
After spending a ton of time researching sdcard interfacing, perils of voltage shifting and with the help of Charles MacDonald getting the Genesis C compiler working on System 24, I have successfully read a file from a SD Card. whoo hoo.
This is only proof of concept as the hardware is very slow receiving data and all the C sdcard and fat16/32 code will need rewriting in assembler. Hopefully a change from a FET to a 74ls05 will improve the speed.
After googling on and off for a week or so (every time she complains about it), today I finally came across mention of Apples two-step verification causing issues (Original post) . Bing the light came on.
Basically, the email on Android phone doesn’t handle two-step verification so you have to create an App specific password to use instead of the normal iCloud password as discussed here Using app-specific passwords.
So, go through all the steps as usual to add the email as shown in the mail settings link, except use the super long and annoying to type App specific Generated Password instead of the iCloud one.
I brought an Outrun pcb which looked original from the auction picture. It even had the Sega licensed holo sticker.
But but when it arrived, it looks like a bootleg.
Back in the 80’s it was expensive to import electronic equipment (tv,etc) to New Zealand so many companies just imported the parts and manufactured them locally. In other words Taito NZ most likely brought the custom chips (including the ck5205 chips), from Sega and built them. Hence the ‘Manufactured under license sticker”. Oh, and thanks guys for soldering in all the eproms.
Original pcb for comparison
Edge connectors and interconnect cables are better though.
Very low volume was the first issue. Checked the volume pot as it had been changed. Was the correct value. Replaced the 1000uf capacitor that was visibly bulging and the sound came back very loud!
When it is first powered up all the ROMs and RAM 7 show as bad, but powering the game off and on again reports RAM OK ROM OK.
Left for a few minutes it doesn’t again.
Having already replaced the capacitor which was bulging, the rest of the capacitors were suspect.
M72 games have a reset chip at ic7, with two timing capacitors. Shorting out both when the board was powered off produced the same error message every time. Replaced both capacitors at C8 and C10 got the board reliably booting.
This little pcb was created by Radich Electronics back in the day to convert a bootleg defender boardset into a Robotron boardset.
But where is the Special chip 1 pair you say? All the TTL logic above the roms implement the blitter chips. In a attempt to prevent copying they have sanded off all the identifying labels.. even underneath the chips!
Fortunately nowadays we have chip testers that can identify any TTL chips, so I am currently labourously desoldering and identifying each chip. Next step is to scan the pcb and make a net list to create a schematic. There are tools out here to assist with reverse engineering but they are all expensive, or require paying someone else to do the work.
The natural progression for Sega after the popular system 16 and system 18 games, was system 24. Higher resolution graphics meant using a medium res display and needed more memory to store the graphics data. Sega’s solution was to use a cheap medium instead of expensive eproms. Over 8 megabits of data…. Floppy disks! How can it fail. But wait, floppy disks can be easily copied, so Sega continued the usage of their FD1094 security processor.
Move forward 20 years and arcade collectors have a major pain. FD1094 batteries failing and disks either getting lost, or damaged from overuse.
A cheap repair/fix solution is needed.
Putting together a System 24 USB adapter, modified System 24 bios and a disk server program running on my PC and voilà, nice fresh disk images can be written using the System 24 hardware itself.
Modified Bonanza Bros disk running on a Dynamic Country Club rom board
My System 24 Scramble Spirits cabinet buried under stuff