I have enabled stretched as well as the possibility of an acked strobe output on my PS/2 adapter. I’m calling this version, 3.0. This should make it easier to use the adapter with some vintage gear, such as an OSI system, which doesn’t latch strobe, but only polls for it. Read the updated doc on my PS/2 adapter page.
Archive for the ‘Vintage Computing’ Category
Updated PS/2 Adapter Firmware (version 3.0)
Thursday, April 17th, 2014Proposed PS/2 Keyboard Adapter Stretched Strobe Change
Friday, April 11th, 2014As I mentioned in a previous post, my PS/2 keyboard adapter requires some strobe stretching in order to work with some kinds of early microcomputers, such as some OSI models.
I think I have come up with a convent way to accomplish this, with only a small firmware change.
Here is what I have in mind.
As before, when the CR jumper is inserted, it disables recognition of the CNTRL-RST and CNTRL-CLR keyboard/RS232 sequences, which assert the reset or clear screen outputs, as appropriate. With my new design, in addition to disabling the aformentioned reset and clear sequences, inserting the jumper will change the clear screen output to an input. Now here is the key part of the change; whenever a character is typed and strobe is asserted (high true), if the CR jumper is inserted, strobe will remain asserted until:
To maintain current behavior (a 125uSec long pulse), you just need to tie the strobe output directly to the clear input.
In order to stretch strobe out a bit, you have a few options:
Behavior with the CR jumper disconnected is exactly the same as with the previous version of firmware.
The main assumption behind this change is that the clear output, is only required for the Apple 1. I have found little to no need to disable CNTRL-RST and CNTRL-CLR on the Apple 1 (or Apple II) and those features are probably of little value to other platforms. In any case, the PS/2 keyboard sequence of CNTRL-ALT-DEL will still assert reset no matter whether the CR jumper is in place or not. In fact, using the CNTRL-ALT-DEL combination from a PS/2 keyboard will be the only way to assert reset if a system needs both reset AND a stretched strobe output. Said another way, you will not be able to assert reset from a device connected with the RS232 port, if your system also needs a stretched strobe.
I’ll do some trials over the weekend and assuming that it works ok and I get no better input or ideas, I’ll have a new version of the firmware available very soon, possibly as soon as next week.
SwyftCards Now Generally Available
Friday, April 11th, 2014Swyftcard kits are now generally available for $55, which includes shipping.
Check out my SwyftCard page for more information about the SwyftCard and Information Appliance, Inc.
Or send email to:mike@willegal.net for ordering information.
PS/2 keyboard adapter used with Ohio Scientific Computers at VCF
Thursday, April 10th, 2014It took a modification, but Bill Dromgoole was using my PS/2 to parallel keyboard adapter in his display of OSI computers at VCF east 9.1. (The link to the image of Bill’s setup from the Vintage Volts blog). The modification was to stretch the strobe output with a 74123 one shot IC.
PS/2 to parallel keyboard adapter
I didn’t anticipate this when I made the design, but some vintage computers poll the strobe directly with software routines, while waiting for new input. My design assumed that strobe would be used to latch a flip flop (or equivalent), which would be reset by software when it actually read the character. I’m not sure how long Bill stretched the strobe, but with an 74123, it’s easy to adjust the pulse. Another person, told me at VCF that did the same thing to make the adapter to work with his vintage system. Unfortunately, I can’t remember who the second person was.
At some point, I’ll probably look into stretching the pulse in firmware, or at least adding a configuration option to stretch the pulse.
Problems with 8008 serial routines that posted
Saturday, March 22nd, 2014Just a heads up – the serial transmit routines that I have posted with the 8008 applications have serious issues. Though it was working for me, other were using serial ports that weren’t so forgiving and had problems with transmit. Finally Corey Cohen tracked down problems with start and stops bits used with the transmited data. As soon as I get a chance to check out the fixed routines, I’ll be updating all the applications on my 8008 application and BASIC pages. For those of you that can’t wait here is the code for a version that I think will probably work. This version is timed for 2400 baud.
My OS/X emulator was also broken and needs an update, as well.
OUTCHAR:
NDI 177 ; mask MSbit
CAL BITOUT ; STARTBIT
CAL BITOUT ; LSB
CAL BITOUT ; 1
CAL BITOUT ; 2
CAL BITOUT ; 3
CAL BITOUT ; 4
CAL BITOUT ; 5
CAL BITOUT ; 6
CAL BITOUT ; MSB
LDA ; timing delay
LAI 200
OUT OUTPORT ; stop bit
LDI 100 ; and extra for inter character spacing
JMP MORE2
BITOUT:
OUT OUTPORT
RRC ; shift to get next bit
LDI 3
MORE2:
LAA ; nop to delay 5 states to make timing work
DCD
JFZ MORE2
LDI 0 ; delay 8 states
RET
Reproduction SWYFT CARDs Arrived
Friday, March 14th, 2014I just got my reproduction SWYFT CARDs in and they look and work great.
Reproduction SWYFT CARD
I’ll be releasing them at VCF east in a few weeks. I’ll also be doing a soldering workshop in which you will be able to build one, even if you are a novice at soldering.
Reproduction and Original SWYFT CARDs
The SWYFT CARD was used as a prototype/proof of principal implementation by Jeff Raskin’s company, after he left Apple after the falling out with Steve Jobs. Later they came out with the SWYFT Computer and the Canon Cat.
Check out the VCF website for details on VCF east 9.1
http://www.vintage.org
regards,
Mike W.
Mimeo’s back in stock
Monday, February 24th, 2014Will be picking up the latest batch of Mimeos from the PCB fabricator this week (probably tomorrow). No changes from last batch.
At this point, I think I have some stock of all the PCBs that I have made since I started on this retro computer hobby.
Send an email to:mike@willegal.net if you have any questions or interest.
Kit and PCB status update and a question
Saturday, February 1st, 2014Now for the question – should I call my Apple II rev 0 reproductions Mimeo IIs, even though they came first?
reproduction SWYFT cards for the Apple IIe coming
Sunday, January 26th, 2014I plan on holding build your own SWYFT card workshop at VCF east in April
http://www.vintage.org/2014/east/
follow the sessions link to find information.
If you don’t know what a SWYFT card is. Check out this link.
http://willegal.net/superproto/index.php?title=Swyft_Card
regards,
Mike Willegal
Apple II Hedlock Fastener Repair
Saturday, January 18th, 2014The “Hedlock” connector on an Apple II is the fastener that holds the lid of the case to the case. It is similar to a 3M Dual Lock reclosable fastener. On the Apple II, it is attached to the enclosure and lid with double sided tape. Over time, the adhesive loses it’s effectiveness and the Hedlock connector comes loose from either the lid or the base of the computer’s enclosure.
Here is an example where the tape stayed with lid, but the fastener came loose.
Delaminated Hedlock Fastener
I used to use contact cement to reattach the fastener to the tape, but I found over time that the contact cement is less than an ideal solution. I think I found a better solution using the tape found in auto part stores that is intended to reattach loose trim pieces. This 1/2″ wide double sided tape has an extremely strong, long lasting adhesive on both sides, so I expect this solution to be an improvement over the contact cement approach.
Before attaching this new tape, you should remove the old tape and adhesive. This is, by far, the most difficult part of this job.
In this example the old tape, which was attached to the lid, simply pealed off.
Old Tape Removed From Top
The adhesive remaining on the Hedlock fastener was another matter. The bond is very strong, and in this case I eventually resorted to using a X-acto knife with a chisel blade to remove it. I left a few nicks in the surface, but since this is covered up with the new tape, I figured it wasn’t the end of the world. In a previous case, I managed to do it by rubbing with my thumb, but I ended up with a large blister on my thumb. I was temped to try a solvent like Goo Gone, but didn’t want to risk damaging the fastener. Here is the back of the fastener with the adhesive removed.
Old Glue Removed From Hedlock Fastener
Before reattaching the fastener, I cleaned up this area of the lid with Isopropyl Alcohol.
Next I pealed back a strip of the new tape and stuck it to the fastener and cut the tape off from the roll.
New Tape Applied to Hedlock Fastener
Then I took a sharp scissors and trimmed the ends of the tape to match the curve of the ends of the fastener.
New Tape Applied to Fastener and Trimmed
The last steps are to remove protective backing from the adhesive tape and carefully press the fastener back in place.
Hedlock Replaced
I don’t know if it makes a difference, but I would wait a day before putting any stress on the reattached fastener.