Ed.—A WORD OF CAUTION: Any hardware project must be approached with extreme caution. SYNC cannot be responsible for any problem that may arise from attempting hardware projects. Obviously, any damage to your computer can be costly in time and money. If you do encounter a problem, write a clear description of the problem either to SYNC or to the author and include an SASE. We will make every effort to find a solution.
For our “Hardware Tips” department this time we have asked Robert Hartung to comment on some letters from our readers. Since he had only the information given in the letters to work with, he cannot guarantee his answers. Rather his responses should be regarded as suggestions to help look for the answers. In most cases more details would be required in order to give a fuller answer. We welcome comments from readers on these problems also.
ROM and RAM Problems
No Cursor
I added the 8K ROM and 16K RAM pack to my ZX80. With the 4K ROM/16K RAM pack, the cursor appears; however, with the 8K ROM/16K RAM pack, the cursor does not appear.
James Kinsella
[address omitted]
Comment:
1) Examine the 8K ROM carefully with a magnifying glass to make sure none of the pins are loose, broken, buckled, or bent under or to the side, and that all are entering properly into their respective positions in the IC socket, with the notched end of the ROM to the notched end of the socket as was the old ROM.
2) Guard against static damage to the ROM chips. If you do not have a grounding clip for ICs, it is good practice that the conductive foam wrapper be laid out next to the IC socket on the PC board before removing the old ROM from the socket and the new ROM from the foam. After gently lifting the old ROM with a very small screwdriver under each end, but with all the pins still touching the socket, place your little finger on the modulator shield, your middle finger on the foam, and lift the old IC out with your thumb and forefinger at each end of the IC, not touching any pins. Place it on the foam and pick up the new ROM the same way. It may be necessary to apply considerable pressure sideways, carefully and evenly, against all the pins at one side of the IC in order to line up all the pins on the other side for proper insertion.
A large (2″ or larger) jaw-type paper clip, such as those made by Esterbrook and sold for about $1 in larger stationery stores, makes a good substitute for a regular IC clip. Connect a grounding jumper between the clip and modulator shield and put the clip jaws over the IC module and on the pins on both sides before removing the IC from the conductive foam or prying the old ROM up from its socket. If properly positioned on the pins, the clip jaws will compress pin-rows toward each other for entry into the socket.
3) Herb Hornung’s suggestion in SYNC 2:4, p. 76, may apply.
4) If the computer still does not work and if the ROM has not been subjected to static damage in handling after removing it from the conductive foam, it is possible that you have received a defective ROM.
Unconnected Input on the RAM Pack
On the Sinclair RAM pack one of the inputs to IC 2 (7400) pin 10 is not connected to +5V or ground. This is an unused gate on the chip and could contribute to some RAM pack difficulties.
Rois R. Harder
[address omitted]
Comment:
Sinclair’s 1981 schematic for the 16K RAM shows pin 8 of the 74LS00 going to pin 1 of IC 3 and IC 4, and pins 9 and 10 of IC 2 both going to RFSH port of the edge connector.
RAM Pack Connections
The ZX81 performance with the Sinclair 16K RAM pack was so erratic that it was almost impossible to obtain any consistant results. Repeated cleaning of the contacts resulted in only temporary improvement. I have, however, improved performance immeasureably by a simple procedure. I inserted a short piece of copper wire about 1/4″ long under each of the 44 contact fingers on the RAM. The wire was from two or three one watt carbon resistors.
P. W. Andrew
[address omitted]
Comment:
Some later versions of the 16K RAM pack have greatly increased contact-finger pressure on the edgeboard connector compared to the earlier models. This is to accomplish what you seem to have achieved by wedging them more tightly to the edgeboard connecting strips.
RAM Pack Crashes
Until recently I have had no problem with my ZX81 with the RAM pack attached. Now the computer crashes while I am typing in a problem. The screen goes blank and the program is erased. After unplugging and plugging it in again, it works fine for 10-15 minutes. I have taped the RAM pack securely in place so it cannot move. I do not think the problem is overheating because sometimes it will work for an hour or two before crashing, but other times it will crach almost immediately after running.
Brent Helms
[address omitted]
Comment:
Should a Sinclair 16K RAM pack that works well on an ZX80 with the 8K ROM have any problems on a TS1000? Mine operates 5-6 minutes and then the screen fills with garbage and eventually flops over full of curved lines.
Jim Mahoney
[address omitted]
Comment:
Both of these situations may be caused by a combination of overheating and edgeboard connection problems.
The overheating can best be solved with an external voltage regulator (see SYNC 3:2, p. 68). Since the regulator requires nearly 2V offsetting voltage, some ZX/TS power supplies may not provide the 9V output under load required for SAVE mode on some ZX/TS computers.
However, by substituting a DPDT switch for the SPST shown in the SYNC article and connecting the regulator input to lug 1 of the first set of switch-poles and the regulator output to lug 2 (C) of this first set of poles, the external regulator is shunted out of the circuit for full voltage in this switch position which is used in SAVE mode. For the 7V output used for cooler operation in all other modes than SAVE, connect the regulator GND terminal (3) to lug 2 (C) of the second set of switch poles and connect a 330 resistor in series between lug 3 of this second set of poles and ground (-). The shunt resistor is omitted. Do not use this configuration with power sources greater than 13 WVDC.
The edgeboard connector problems can occur because of oxidation of the soldercoating on the connector strips even when mechanical stability has been secured. Swab both sides of the connector strip liberally with TV contact cleaner/lubricant and slide the RAM pack on and off several times. Insert the pack fully, then back it off just enough that the contact fingers are not pried open by pressure of the pack case against the computer case.
RAM pack wobble may be prevented in various ways: 1) Mount both the computer and the RAM pack on a flat mounting board (3/4″ chipboard is fine) with a bit of picture-mounting tac-dough under each corner of the computer and under the RAM pack legs will ensure they do not move relative to each other. 2) Some have had success also by putting extra-high pads under the computer so that the RAM pack hangs free from the connector. 3) Ribbon-cables with connectors on each end also work well. 4) Hard-wiring and placing both the RAM pack and computer inside a grounded metal cabinet as used with some keyboards is the best approach of all.
New Pads
My answer to RAM pack wobble was to get some sticky pads at Radio Shack (1/2″ in diameter by 1/8″ thick) to replace the standard pads. At the keyboard end I put 2 pads 1/2″ closer to the expansion port. At the port end, they are 1/2′ O.C. from the edge of the computer. On the other axis they are in line with the originals.
Andreas Rainwater
[address omitted]
Screen Display Problems
RAM Pack Buzzing Sounds
Raymond Fowkes in SYNC 2:4 suggested soldering two foil tabs on the underside of the PC board together with a short piece of wire to solve the problem of the buzzing sounds caused by the 16K RAM pack. Is this safe for the computer?
John Torrance
[address omitted]
Comment by Raymond Fowkes:
I learned that there was more to the story after I wrote to SYNC. My ZX81 was a kit, and it seems that Sinclair left out the instructions for installing the long thin metal grounding strap in the computer (not the RAM pack as a few thought). This raised the resistance in the 0 volt trace just enough to cause noise when the extra load from the RAM pack was added.
Therefore, anyone who does not have a long thin silver strip of metal running across the noncomponent side of the ZX81 may solder a wire to the two large pads of bare foil (labeled TB in Figure 2) on the underside of the PCB. One is next to the edge connector where the RAM pack is attached; the other is in the opposite corner next to the regulator (the 3-prong IC on the component side with the big metal heatsink). Factory assembled units should be OK in this respect, but those with hardware experience who are very careful could further reduce the noise by connecting wires in parallel with other traces, especially the one carrying 5V to the edge connector though this is not advised because of various risks.
Comment:
I referred the above question to Raymond Fowkes for clarification. I note from letters by owners of factory built ZX81s that some may have defective solder connections on this strip or else it may not be making proper contact with the metalized coating inside the case. The result is increase TVI, which the 16K RAM pack may raise to an even higher interference level. The solution is to make sure that all the solder connections are secure by reheating them and ensure that the strip is making contact with the case by arching it slightly higher above the board in the center. However, this will not cure the noise which originates in the voltage converter in the RAM pack itself. This feeds back transient noise into the computer power line as well as radiating RFI if all the grounding strips in the RAM pack are not making proper contact with the metalized coating in the case. Some 16K RAM packs, such as Memotech, do not use this power conversion and so produce less RFI.
The solution involves the adding of suppression around the Zener diodes which originate the most of this noise, but, unless the proper components are identified, the RAM could be disabled. A thin metal box formed to fit over the RAM case and grounded to the TV modulator case (but not touching the edgeboard connector) will alleviate the RFI which affects the TV display. Heavy aluminum foil may be used. (See Bruce Kirk’s letter.)
Dark Bands and Noise
The 16K RAM from Apropos Technology worked well on my ZX80 except for considerable noise in the TV display. By covering the ZX80 and module with aluminum foil grounded to the coax to the TV, I reduced the noise. However, when I used the 16K on the TS1000, two darker horizontal “bands” appeared in the TV display, moving slowly and evenly downward. The display itself (symbols and spacing) was not affected although these bands overloaded the TV sync and caused minor “tearing” of the picture.
Bruce P. Kirk
[address omitted]
Comment:
Your “fix” of the noise problem with the foil shows the need for more adequate shielding in the RAM pack case unless the noise source is corrected by suppression at the internal power converter of the RAM.
The moving horizontal bands in the TV display may be partially from transients going back into the computer along the 9V bus from this power converter noise. Usually such bands indicate inadequate filtering of the DC power supply. This may easily be corrected by connecting a 2200 uF 35WVDC capacitor between the DC power cord leads. However, the loss of TV sync and tearing may indicate a combination of RFI, noise transients on the power bus of the computer, and excessive ripple in the power input. You may have to work on these one at a time.
RFI Trash
The RFI trash on my portable TV screen is very annoying. However, on my 17″ TV the screen is crystal clear. What does my large TV have that my small one does not? Is there a circuit I can build for my small set to solve the problem?
Rick Goulian
[address omitted]
Comment:
The tuning section and selectivity of the 17″ TV may be better or have better shielding. Some TVs are designed for better RF harmonic rejection than others.
1) Be sure that the built-in antenna or rabbit-ears are completely disconnected from the TV input terminals and disconnect the CATV or other antenna leads. Relocate them and the TV power cord and the computer power cord as far as possible from each other.
2) Sometimes it helps to put 2 or 3 ferrite sleeves on both these power cords, respectively, as near the TV and the computer as possible. Wind the cord through the hole several times. (Radio Shack has assorted packages.)
3) The length of the cable between the computer and the TV may be critical with some tuners. The 48″ length supplied with the computer is designed to be one-fourth wave-length at the pix frequency of TV channel 3 for maximum signal-transfer to noise ratio. The TV/computer switch box supplied lengthens this enough to provide a match for TV channel 2, in length and in impedance. Clean the switch contacts by spraying TV contact cleaner/lubricant liberally inside the box and working the switch back and forth.
4) Some 16K RAM packs emit excessive RFI back to the computer and also as radiation. Use the computer only on a non-metallic desk or table, or place it on a sheet of heavy foil grounded to the outer conductor of the TV cable plug or to the TV modulator case. It may help to fold the foil up and around the RAM pack also. (See Bruce Kirk’s letter.)
5) When severe TV interference occurs, either from RFI from the computer/peripherals or from being in a strong signal area on the channel used by the VHF modulator or on an adjacent channel, the only real solution may be to replace the VHF modulator with a UHF modulator (see SYNC 3:1, p. 72). Try adding a short patch cable (4-6″) to the existing TV cable if the UHF signal to the TV seems to be down.
6) Another solution is get a well-shielded cord for connecting the computer to the TV.
A ZX81 and an 18 year old Sony
My ZX81 kit works only with my 18-year old Sony portable TV. It works best with the gain control turned down. With other TVs the ZX81 seems to be putting out too much signal and overdrives the screen resulting in an unusable, cross-hatched pictures. I have tried turning down the AGC on several TVs as well as a different TV/game box to no avail.
Ross A. Rainwater
[address omitted]
Comment:
The cross-hatching would indicate the problem is RFI rather than overdriving, which in effect is superimposing one or more spurious signals to the TV over the top of the desired one. The Sony gain control apparently reduces these other signals enough that only the legitimate one is visible in the display. Some TVs have better front-end shielding and selectivity and adjacent-signal-rejection than others, which may account in part for the difference in those you have tried.
Small coupling trim-pad capacitors placed in each side of the TV lead-pair in the TV/game switch box might help tune out and attenuate the unwanted signals. Drill small holes in the box directly over the trim-pads so they can be adjusted with the box closed, using a non-conducting tool.
A better solution, however, would be to prevent or shield off as much of the interference as possible at the source. See above for suggestions for RFI problems.