tf_hh

I´ve a few pieces left. Look in your mailbox. Jurgen

Hi, a new firmware and flashtool is availible for Sys-Check V2 series. Sys-Check Files V2_2 Release 20160717.zip The new flashtool now supports both Sys-Check PCBs equipped with a SST39SF010A flash chip (V2.1 and V2.2). When you start the flashtool, it will ask you which version you have. To be absolutely sure, the flashtool shows a simplified graphic view of both versions. I think, everybody can handle it Of course the flashtool (and all other documents) are written in german and english. Also there´s a new firmware V1.4 out. This firmware can be used with ALL Sys-Check versions - also with the old V2.0 (without flash). Owners of the V2.0 must burn the 16 KB firmware image into a 27128 EPROM. All other please use the flashtool to update their Sys-Check. V1.4 has this minor changes: - Sys-Check can be used with the U1MB expansion. - The new version first checks the whole memory (to be more exact... the first 48 KByte) and then displays the defective bits (if there are any...) - When bad memory within the first 48 KB is found, the upper 16 KB will not be tested - If memory is fine, OS-ROM will be tested as before - If both tests are fine, then pressing START after this will set Sys-Check into endless loop test mode. Memory and OS-ROM are tested infinite - until any error occurs - During memory tests the background will be coloured red if defective bits were detected Jurgen

The 68B21 and 6520 are the same IC. They can interchange without issues. The BitWriter original PCB uses the 6821 (mostly A version, 1 MHz, the B version could be used with 2 MHz clock), but I used for 3/4 of the BitWriter Replica boards the 6521 - works fine. Jurgen

Hmmm... after repairing dozens of XLs I never had a defect LS158 before. I would guess, the source of trouble is elsewhere located. Especially at the 600XL PAL model I´ve had two times a bad delay line. When the machine boots up with Sys-Check enabled, there´s something wrong in the DRAM control circuit. First check the delay line, then the 74LS375 and 74LS51. Mostly they aren´t really bad, but when you pull these ICs, they have nearly black pins. Use a file and clean up BOTH sides of the pins, reseat the chips and with a good chance all is running again. Good luck, Jurgen

I wrote it in the "message to seller" form from Paypal... it´s me and 800 XL, as in your list. BR Jurgen

Hi Tim, I´ve seen many 800 XL with these DRAMs. They are from Hitachi and yes, far times ago Hitachi has let produced these DRAMs in a fab near Cologne, which later were acquired by Infineon, a Siemens spin-off. The Hitachi HM4864 are equivalent to any 4164 DRAM. And yes, these kind of DRAM could be get very hot if something inside is damaged - like the Micron MT4164, too. The CO61598B is the operating system ROM. It get´s a little bit warmer than usual, when the computer can´t start normally, because CPU is doing endless loops in the ROM sometimes. But as long you can put and holt your finger on it´s surface, all is fine BR Jurgen

Sys-Check doesn´t "need" really 64 KByte, but it expect it. So of course you can use it to test the first 16 KB, but memory test will always stop with errors at $4000.

Hi Eric, First I would suggest: Bad contacts at the PBI. If you haven´t done this until now, remove the whole XL mainboard from the case and clean both sides of the PBI contacts with a rubber, so that all contacts are clean again. Examine the PBI connector on Sys-Check. Is there any of the contact guide broken or bend? Garbage on the screen while the computer works further is often an issue around PHI2 signal. Sys-Check generates it´s own PHI2 for write accesss to prevent wrong data in the SRAM, but sometimes the timing at the XL mainboard is terrible. Especially the DRAM circuit at some XL mainboards has a lot of flaws. Mostly it works as long no expansions or mods are connected. If cleaning etc. doesn´t work, try to change (from another or something you´ve in the drawer) the following parts: - All DRAMs. Especially if they are from NEC with a "-3" marked (300 nS - norminal too slow for a 1,79/1,77 MHz machine!) - The 74LS08 chip - The 74LS375 chip - The 74LS51 chip Sometimes also the CPU itself has some strange timing. Also, if you had one, try another CPU. Best idea, of course, check another 800 XL Good luck, Jurgen

Hi, just my 2 cents to the power discussion. 1. Use your own parts. The difference in pricing is too small to earn the headache if something went wrong. I´ve bought some of these 1-2 USD regulator PCBs in china some time ago. 50% uses faked chips (500 mA version real, declared at 1 Amp and so on). The solder quality is bulls*it sometimes. The space around and so on is fixed to THAT one PCB you´ve ordered from china. Next order you get something totally different, even if the photo is the same before... 2. Don´t use open frame PSUs. I don´t know the legal situation and law in the U.S., but in europe YOU have to pay for all that can happen with such an open frame PSU. Fire, electric shock, death. As a private tinkerer or smallest business (wo-) men it´s better to use "ready-to-use" external PSUs. You can get 12V 5A stabilized external PSU for around 10-12 USD/Euros. Use such a thing... 12V input, step-down to 5V and 3,3V for your needs with own parts. Jurgen

Hi Steve, take a look here: http://www.realdos.net/UltraVideo%20xl.html Jurgen

I would suggest using the 16C550. Also an old, but good UART. Some manufacturers produce them until now, so they are availible. Like this IC, it´s easy to handle and need almost nothing special external components. Just use a oscillator for the baudrate generator, level shifter (1488/1489 or, if only 5V present, one or two MAX232), ready. Also good for "low speed" systems... the 16C550 supplies Auto-Hardware-Handshake, means than CTS/RTS must not set by the computer (if enabled), the UART will handle itself. And you can support also DTR, DCD, RI and so on... I like full serial ports, not only RxD/TxD. When I´m remembering right... somebody in the early 90´th years has publish such a solution with a Atari 850 compatible driver. Maybe anyone knows about this project? Jurgen

Sure. There´s a lot space! Look, this is one modded 600XL I´m working on for an ABBUC member. Ultimate 1 MB, VBXE, Stereo. Of course already 64 KB main memory upgrade, S-Video Upgrade. The stereo PCB is my own and includes an SIO2PC-USB, for that I will install an USB jack over the SIO jack. When I´m going totally nuts, I will try also to include a SIO2SD using one of the very small displays used by Panos for the external Micro-SIO2SD. As FJC already told, you´ve to use BASIC ROM´s place. I replace the 24 pin socket, solder a 28 pin socket into. Also you´ve the change the both "jumperfields" (look at the red jumpers upper right corner in picture 2) to the opposite places. And, it´s a good idea to remove the plastic mylar cable from keyboard to the adapter PCB used in the 600 XL and use some standard ribbon cable. Jurgen

Oh yes... he does! For example, in the BitWriter circuit the "R0" input from PIA are not as usually connected to "A0" from CPU (and "R1" to "A1"), they are exchanged... R0 is connected to A1 and R1 to A0... of course I miss that point and the first prototype PCB won´t work until I look more exactly...

Ja, drei Stück habe ich noch - yes, three Sys-Check left. BR Jurgen

At this time all boards are sold, but I´ve ordered 10 more PCBs and parts for it. In 3-4 weeks I´m ready and can deliver some more. Jurgen

Actual no boards are left, but two other users want one, so I´ve ordered 10 more PCBs and parts. Will take 3-4 weeks until they are finished. Normally all my projects are released fully to the public incl. schematics and so on. Because this project is mostly made of re-engineering (by permission) I´ve the ask the involved people. If everything is fine, surely I will publish all. No, both parts (Super Archiver and BitWriter) doesn´t use any programmable logic. The Super Archiver is only a 2 KByte SRAM, a 4 KByte EPROM and one 74LS00 glued in epoxy. The BitWriter itself has some 74LS00, 74LS193, 74LS251 and 74LS74 onboard, followed by a PIA (6520 or 6821) and 8 KByte SRAM (6264). The real power - and this is the amazing part for me - of both projects are packed in the software. The hardware is "only" to enable breakout the limits of the FDC-circuit in the 1050 original PCB. Not more. Kudos to Bob Puff for this great work of all. This is the original, full populated BitWriter board. The black block in the middle is the Super Archiver: The following both scans are the naked PCB after I´ve removing all parts to see what´s going on under there Jurgen

Hi James, The best way is to use the ABBUC website and the contact form. Go to http://www.abbuc.de/abbuc/kontakt (first switch to english language if you want, see upper right corner) and select "Wolfgang - 1st Chairman". The @abbuc.de mail addresses are not very good to use, because a lot of mails coming from outside germany (but also within) will be detected as spam and sorted out. Also some other email provider bounces mails to that domain, because it´s something strange forwarding, whatever... Jurgen

I think it works, but I can´t test it now. When selecting the whole flash (SST39SF040), then a message "Insufficient memory for flash data" appears. I´ve taken Sys-Check with external 512 KB RAMBO, because I´ve only one U1MB left at this time and it´s nothing built in now. Why UFLASH2 need more than 512 KB to flash 512 KB? Will check on weekend, sorry, no time earlier for that. Jurgen

Will test in the evening!

I use the SiDE programmer´s info found on Candle´s website: http://spiflash.org/block/17.html In this version $D5E0 is named, but I check out with a real SiDE2, and it´s $D5E1. So I take this one and use all bits same like at the real SiDE2 incl. external cart on/off. Dumping and flashing the SDX Supercart with UFLASH2 is no problem, works fine. UFLASH shows SiDE2 and of course - in the actual version - "only" the first 256 KByte will be read out or written, but it works like a charm. My logic decodes the complete address and only does action when writing to $D5E1 (SDX bank control) and writing/reading to $D5E2 - RTC control. All other adresses are ignored. So when UFLASH detects a real SiDE2 probing the ID string at $D5DF-$D5FF is simple way to differentiate should be possible. Jurgen

Hi, as already written here... these caps are not mandatory in such a simple PCB. Of course they have to be included in every circuit design, but it works also without. I would suggest to examine the PCB itself. The photo isn´t very detailed, but on both sides are scratches and damages visible. It´s not very hard to remove the solder resist, but some force is needed. I wonder how this can be happen within the cart... ? So check all uncoated, coppery traces - maybe there´s a cut trace anywhere, and then you got the reason why the cart won´t work Jurgen

Simple answer: To make the SDX Supercart work together with an U1MB expansion - even if there´s no really NEED for it, but it works. That´s the reason why I choose $D5E1 for SDX logic instead of using $D5E0.

This is the content Lotharek ships out. Lotharek U1MB.bin Jurgen

I´ve added some useful (hope so!) hints to the manual. At the last pages of the new version of the manual you will find some graphical explanations how to set up Sys-Check and use the common functions and features. Manual Sys-Check II English.pdf Jurgen

Yes, I got another feedback in the same direction. I will do it better and include tomorrow in the manual some simplified photos with the settings of the DIP-switches for each function. One picture told you more than 1000 words Second thing is related to the post I´ve done some days before, the flasher´s instructions are only suitable for Sys-Check V2.1. I will publish the new flasher version in the next days. Jurgen