TS COMMUNIQUE: A forum for people having problems with their Timex Sinclair 1000, 1500, and 2068. If you have any questions, send them to:
TS COMMUNIQUE
c/o Time Designs
29722 Hult Rd.
Colton, OR 97017
I have a Larken cartridge which has a Spectrum ROM, which is activated by the command: OUT 244,3. Is it possible to program a 32K EPROM which contains both the Spectrum operating system and a 64 column operating system, and have them selected by a simple switch? Both operate in the 0-16384 memory region (switch selects one part of the EPROM, while the other part stays inactive).
Renato Zannese
Toronto, Canada
Dear Renato,
The answer is yes. To carry it even further, why not program a 64K EPROM to contain the 2068 operating system, the Spectrum operating system, and OS-64, each in its own 16K chunk. The advantages include putting the one EPROM inside the machine, freeing up the cartridge space for your own programs, and it would also cut down on “buss loading” by having fewer components loading down the data and address lines.
This has actually been on my “drawing board” for quite some time, but I have been waiting for the price of 64K EPROMs to come down. Because of the weakening dollar, they have actually gone up in price (about $15 from Jameco), but I guess I will have to go ahead and spring for one now that you have inspired me to look at it again.
The hardest part is programming the EPROM, but if you can handle that, it is mainly a matter of inserting the new EPROM into a modified socket inside the 2068, taking care of two extra control lines (because its an EPROM), and connecting address lines 14 and 15 of the EPROM thru a dip switch to select one of the four 16K areas:
address lines
14 15
0-16K low low 2068 system
16-32K high low your own system
32-48K low high OS-64
48K-64K high high Spectrum
Because you are selecting the upper two address lines, the 2068 will only see that one 16K chunk. Look for a project on doing this in a future issue of Time Designs. While you are at it, why not add 32K of RAM (one chip) to the system by piggy-backing one of the two ROMSs (or EPROMs) to a 43256 (32K X 8 Static RAM)? The pin-outs are the same! -Joe
I have an Amdek Amdisk III(B). Sometimes when first turned on and commanded to LOAD from drive #2, it will give a report: V RECORD NOT FOUND,0:2. Or it will load and after working on a program and then going to a line number in the program to save, the report given is, V RECORD NOT FOUND, 9999:3. This will also happen using a direct command to save. Is this caused by the drive or the disk? If it is the drive, can it be corrected?
Herbert Braun
Wickliffe, Ohio
Dear Herbert,
I also have an Amdisk III and have noticed the same problem, but only during the first minute or so after being turned on. You mentioned that it does it on drive #2. Are you powering your disk controller from the power supply of drive #2? If so, this may be where your problem is stemming from…as that was causing my problem, but only for a short while after power up.
You may want to try adding some tantalum capacitors to the 5 and 12 volt supply at the controller if it is being run from the Amdisk. Try 1 to 5 microFarads rated at better than 12 volts, and watch the polarity as they have been known to explode when connected backwards. If you don’t have the controller running off the Amdisk power supply, try cleaning the heads. If that doesn’t work, you may have to have the heads aligned, which would probably require you to send it off to Amdek. -Joe
I recently converted a surplus TI keyboard for use with my TS1000. Everything works OK except for the shifted functions on the “s” and “d” keys. Unfortunately, these include LPRINT and SLOW, which limits the usefulness of the mod. These functions work fine when the add-on keyboard is unplugged. The letters and unshifted functions work as advertised.
I read a piece in an old issue of SYNC, which indicated that the problem could be high resistance in the KB lines. However, I checked the KB1, KB2, and D4 lines with a multimeter and the resistance didn’t seem to be any higher than on any of the other lines. The problem in the SYNC article was not with an add-on keyboard, but with the keyboard resistor pack RP-3, and author Stephen Turner suggested solving it by wiring in parallel resistors (to RP-3) on the underside of the keyboard to reduce the total resistance. Would this be worth a try or is there a better solution?
Bob Johnson
Ypsilanti, Michigan
Dear Bob,
When using an external keyboard, make sure that the internal keyboard is disconnected. Also, the length of the cable that connects the external keyboard to the TS1000 may be too long…don’t go any longer than 8-10 inches. It really depends on the type of cable you are using. I would suggest ribbon cable using every other wire, with the ones in-between being grounded. Also try disconnecting any peripherals you have plugged in…see where the problem is actually occuring.
If all this doesn’t work, then try adding a 10K to 18K resistor between address line 8 and +5 volts, and address line 9 and +5 volts. If you have access to an oscilloscope, take a look at the address and keyboard lines going to the external keyboard. You should see a waveform swinging from about.5 volts to around 4.5 volts. By probing around, you can quickly determine where enough “signal” is or isn’t. If you still have problems, you may need to buffer the external keyboard (see Tim Stoddard‘s article in July/August ’86 issue of TDM), or make your own external keyboard interface which would plug onto the rear edge connector. Back issues of all the TS mags have had articles on this. -Joe
This concerns the inability of the TS2040 printer to print with linearity. The height is greater than the width. Is there a circuit modification that would alleviate this problem? Through the years, I have seen this question asked many times, but never answered.
Jim Powell
Napa, California
Dear Jim,
I have never seen any modification to cure that problem. I’m afraid that its more the mechanics of the printer (the way the motor “steps” through each bit line). Has anyone out there found a cure for it? -Joe
The cable on the TS2040 printer is extremely short, so I cut mine in half and wired on a male and female “D” connector to the ends. When plugged together, the printer works fine, but when I insert a three foot long cable with matching connectors that I made up, I am unable to even get the “K” cursor on the monitor screen. I have checked the shielded extension cable but found nothing wrong. I would apppreciate any suggestions you may have.
Arnold Nieuwenhoff
Sutton, Massachusettes
Dear Arnold,
The length of the cable is the problem. Too much stray capacitance and inductance is loading down the data and control lines. The 2040 printer is talked to through port 251 (FB hex), but only address lines A2, A7 and control line IORQ are used for decoding through the 74LS10 IC chip inside the 2040 interface. The strobe line out from pin 6 of the 74LS10 can be used to toggle a buffer IC such as a 74LS245 (octal bus transceiver). The other lines used are D0, D2, D6, D7, RD, and WR…plenty of room.
Some extra circuitry will be required to enable and control direction of the 74LS245 using IORQ, WR, and RD. With this buffer added, the TS1000 or 2068 bus is isolated from the long run to the printer and should thus cure your problem. Someone in my user’s group did this several years ago and had great results. -Joe
I have a 2068 and purchased the program called “TOMAHAWK“. It is a fun program. However, since I also have the RAMEX disc drive system, I’d like to convert it to disc. Here lies my problem: I’ve been able to save other purchased programs to disc, but this one’s got me short of POKEing #’s to addresses, I know very little about machine code. Could you help me out? I’m getting tired of growing a beard every time I want to LOAD this program from tape.
Regis Giacobbe
Ocean City, Maryland
Dear Regis,
The easiest way to save it is to have an NMI SAVE button on your interface. But since you don’t, you need to first be able to break into the program. I’m afraid that I don’t have a copy to try, but you may want to try turning on the 2068 and MERGEing in the loader part of the program so you can LIST it and see how the rest of it is loaded. There may be several loading sequences, so you may have to MERGE several times, but always make sure the 2068 is clear of any other programs.
If you can get into the loader program but can’t tell how long the code is that is ultimately LOADed in, and you don’t have any type of header reader, find out how much memory is FREE and subtract it from 65535. Then try loading the code followed by an immediate save using the result from above for the starting point, and the FREE amount for the length. You may SAVE more than you need, but you will have it all.
This is all provided that you can successfully list the program’s BASIC part. Some programs use their own LOADing routines which just simply can’t be broken into. I have “Ghost Busters” which is like that, and unless you have NMI SAVE, you can’t break into it! Good Luck! -Joe