Archive for February, 2010

Caps not arriving until late next week (at the earliest)

Friday, February 26th, 2010

Seems like my supplier has had a glitch and hasn’t yet shipped the caps. I want to try these caps, before I finalize decision on which caps to ship with Mimeo kits. This will set back putting kits up for sale until I get them and have a chance to try them out. Maybe I’ll be ready a week from Monday. The lower beta transistors I want to try on the video output, are also part of this order.

Meanwhile, I’ll see if I can solve the power up problems with the PS/2 to ASCII converter. I have changed the brownout detector setting, which helped, but that didn’t completely solve it. I have significantly rewritten the initialization code, closing some holes and adding some retry logic, but haven’t had a chance to try the new code yet.

PS2 to ASCII interface work

Sunday, February 21st, 2010

Since I didn’t get the new caps and transistors yet, I went to work on the PS/2-ASCII interface. Well except for an issue with booting up reliably, it seems to be working. I need to work on the AVR fuse settings for power up and/or add power on reset circuit.

PS2 to ASCII adaptor test setup

For the Apple-1 clear function I added a diode to the AVR’s output, similar to what is seen on the input to D15, pin 9. This is necessary because there is another TTL output that can drive this signal.

PS/2 to ASCII adaptor

The three jumpers on the right side of the board are to disable/enable the following functions.

  • enable automatic arrow mapping to Apple 2 – ESC-char cursor control sequences
  • enable inverted data – not needed for Apple 1 or Apple 2, but added in case some other vintage computer needs inverted data
  • enable/disable Control – RST and Control -CLR – reset and clear screen sequences
  • There are several options for mapping the wiring from 16 pin dip ribbon cable to the AVR.

  • It can be wire wrapped as I did on my prototype
  • The header can be left off and it can be done directly with jumper wires
  • You can build a plugable jumper block in a housing that plugs onto the header
  • Beta Version of Assembly and Build Manual Posted

    Wednesday, February 17th, 2010

    http://www.willegal.net/appleii/A1-assembly-v0.8.pdf
    Take and look and let me know what you think. If you find problems or mistakes let me know. Actual printed manual that comes with full kits will include full color image of the front of my prototype board on the covers.

    I have also put up an initial draft of my kit information page.
    http://www.willegal.net/appleii/apple1-kit.htm

    Video Imperfection

    Tuesday, February 16th, 2010

    There is a small flaw in the video section of the Apple 1, that causes spikes in the video signal. This results in extra pixels in the display. Here is a worst case display, with brightness turned way up, of those pixels, with no characters displayed ( the screen should be blank).

    I have been able to confirm with some help that this fault also occurs on the Obtronix and original Apple Is. I’ve tracked this down to cross talk between the video signal from C13-6 to the 3K resistor and CLK03, which feeds the 2504s clocks. These two signals share adjacent traces across the top of row D, just above the large ground trace.

    Wendell S. has done some further analysis and has determined that the fault is caused by the high impedance nature of the video output circuit. Wendell writes “The node driving the 3K can be very high impedance when both C13-6 and D1-13 are high so the network is very susceptible to pick-up as you described (nice catch!). The impedance of the node is set by the base current of Q5 as it pulls down the emitter followers in the TTL outputs.”

    I’ve ordered some transistors with lower Beta in order to see if I can eliminate this noise. I also did a quick test by pulling down the output C-13, Pin 6, with a 2.2K resistor and as expected, the noise completely disappeared. Another fix is to bypass the trace by lifting one leg of the 3K resistor and C-13, pin 6 and running a jumper wire, avoiding the crosstalk.

    Along with the aforementioned lower beta transistors, I’ve ordered some capacitors that I hope more closely mimic the originals. The assembly manual is also moving along, though I had a bit of a setback recently, when I accidentally put the whole manual into a bullet type format. I’m still undoing that. I’ll put a PDF version online once I finalize it.

    I’ve been getting a number of inquiries about bare boards and kits. Bare boards should be shipping in about a week and will cost $150. Kits should be ready about a week after that and will cost $666.66.

    Folks that are on the wait list will get first chance to order. I may have to order a new batch of boards, before I get completely through the wait list. Delivery of my first batch was pretty quick, about 2 weeks, so I don’t expect huge delays if I do run out.

    More ACI TroubleShooting

    Thursday, February 11th, 2010

    Workshop is now in the basement, which has a decent sized table that I’ve taken over. This allows me to spread out a little bit, but the O’scope still usually ends up on the floor.

    Ok, since wsander posted some AIFF files for download to the A1 from an ipod to the cassette interface on Applefritter, I had to try it. Well I barely could get it to work. Now I’m figuring this isn’t right and I’ll have to work some more on the Apple Cassette Interface. I’ve done much monkeying around including increasing hysterisis of the voltage comparator and adding a negative five regulator to it’s negative supply. All these changes made little to minimal improvement. I even put the circuit into a Spice simulator looking for a flaw. Nothing significant found. I even put the Apple II circuit into the simulator, and though there are differences, it’s not substantially different.

    Finally while looking at the input from the iPod I noticed some high frequency noise on the input signal. After some simulations I determined that adding .01UF cap between the input of the voltage comparator and ground might filter out that noise, at the cost of losing a bit of lower volume capability.

    After this last change I’ve been able to load wsanders iPOD files reliably. I’ve also been able to write and read cassette tapes, but the volume setting on my particular tape recorder, a vintage Realistic CTR-31, is very very critical. I’ve been able to save Apple 1 Basic, plus Hammurabi and restore it via tape. I wouldn’t call it reliable, but it can be done. I’ll probably revisit this in the future, but for now, I’m going to do some testing on CFFA-1 card, while waiting for a new batch of caps and finishing the assembly manual.

    Cassette Interface LED Issue Figured Out

    Monday, February 8th, 2010

    The cause of this problem is rather simple, just like the circuit. The base to emitter voltage drop of the MPS3704 transistor is .6 volts. In order to light the LED, you need an input voltage high enough to cover the base-emitter voltage drop of 2 transistors. This means the circuit is designed to light the led when an input signal of about 1.2 volts is seen. This is around the voltage of a normal line level signal, but the earphone output of my vintage cassette deck is nowhere near this, which is why the LED does not light. The shape of this signal is more important than the magnitude, which is why I can read data fine with my cassette deck, even at such low levels.

    More Mimeo-1 Testing About Wrapped Up

    Saturday, February 6th, 2010

    74161 investigation

    I did some further investigation into the problem with the 74161As. The problem is limited to location D-8, the other 74161 locations can be populated with 74161As without issue. Here is what I have found out about the problem with D-8 so far.

    The high level of the input to D15-9 (preload) is just at 2 volts. There is a small glitch below 2 volts just before the char rate clock(CLA) raises. D15 responds by lowering VINH and things go south from there. D15 should preset VINH to high, but apparently the floating preset input drops a bit at the same time, resulting in a preset to zero. Apparently a 74161A at D8 is fast enough to occasionally see VINH drop and counts on the nearby raising edge of CLA, putting the state machine into a mess. Apparently the 74161 doesn’t have enough setup time to count at this point. The time between early VINH drop and the next clock edge is about 20ns, but that is just about at the limit of my old, uncalibrated scopes resolving power, so consider this an estimate. This appears right on the edge of the 74161 setup time spec. Apparently enough for an A part, but not enough for a non-A part, which ignores this glitch and continues on nicely.

    Cassette Interface

    When I first tried the cassette interface, the input signal after the AC coupling cap looked rather nasty on the scope. I have learned through some efforts to build a regenerative RF receiver that AC coupling caps should be good quality. Testing this theory, I replaced the original cap, which came from a scrap box with a new monolithic cap. Sure enough the signal cleaned up considerably and I am now able to read from the tape correctly. The only thing not working, is the READ LED. I’m going to see if I can put this small circuit into my spice simulator and see how it should work, as there is nothing obviously wrong with it, though the input signal level is pretty small. It’s possible I just need a lower power LED. I have to run one additional test on the cassette interface, prior to declaring victory on this front. This is involves typing Basic into memory, WOZ style, then saving it and seeing if I can load it back in again.

    So What is Left

    I’ve got a few things left to do prior to taking orders.

  • Finish testing cassette interface
  • Obtain and test new decoupling caps
  • Finish the assembly manual
  • Figure out how I’m going to pack everything for shipment
  • I should be able to complete this effort before the end of Feburary. Those who are on the wait list will get the first shot at kits. If you asked to get on the wait list, I’ll send you an email when I’m ready.

    Interesting Evening with the Mimeo 1

    Wednesday, February 3rd, 2010

    As of this morning, the flakey video problem wasn’t completely solved. The current counter parts I’m using are 74161A parts, and I was hoping I would manage to find some without the A suffix locally, as they have slightly different specs. I stopped on the way home hoping to find some 74161’s at the only shop in the area likely to have such parts. I lucked out and we found enough parts in a close out bin.

    After I got home, Mike Fortuna (who works nearby) dropped by with a couple of cards to try on the Mimeo 1. First we tried out Rich Dreher’s CFFA 1 card. Not only did it work, we were able to load and briefly run a couple of classic programs. These included Apple 1 BASIC, Lunar Lander, and chess. I was impressed with this card.

    Second, we tried Vince Briel’s prototype cassette interface card, but we had less luck with that one. The behavior was similar to the Obtronix cassette card that I built up. Write appeared fine, but reading was unreliable. I’ve still more work to do on the cassette interface front. Mike says the card works with his replica-1, though I don’t think that Vince was ever able to get it to work.

    Mike and I also discussed the video problem, and we agreed that the symptoms are that of a video counter issue. It was great that Mike could stop by and help me check out the board and talk vintage computers.

    After Mike left, I swapped out the 74161As with the newly purchased 74161s and after an hour of testing, things are looking promising. It is starting to look like I might have to ask my supplier to do an exchange.

    Now I’ve got to take some time to fix my son’s Srt101 camera, which was recently dropped or at least roughly handed.

    Interesting Little Issue

    Tuesday, February 2nd, 2010

    I’ve had some intermittent issues with the video display getting out of whack. Turns out the known floating inputs to the video counters were not floating high. Floating inputs are supposed to float high on 7400 series parts and the Apple 1, is designed with this in mind. I suspect I need to find a different vendor for these chips. In the mean time I’ve pulled them up and so far, things are looking a lot more stable.