POWER4.101 file by R. V. Getsla 74405,1177 Last update: 19 MAY 85 Rev. 1.0 ----------------------------------------------- First of all, I received a request for more detail on the drawing for connecting solar cells and a rechargeable battery to the model 100, so here it is. +++++++++++++++++++++++++++++++++++++++++++++++++++++++++(positive side) (sleeve) | | Model 100 battery solar panel (center) | | ----------------------------diode here-------------------(negative side) (band on solar side) Be sure you solar array has the proper voltage and current ratings to charge the battery you plan on using. See the POWER4.100 file in DL 4. ----------------------------------------------- This is the second in a, hopefully, continuing series of articles on the hardware inside and outside of the TRS-80 Model 100. Presently geared to the user who is curious about how it all fits together and can manage to wield a soldering iron with aplomb, if not skill, though at this time I do not contemplate you needing one, unless you are as brave as I am ... a qualitative decision on your part. The first item of interest is the internal NiCd battery which keeps the RAM and clock chips alive while you change the AA cells or otherwise deprive the main circuits of a power source. The memory power internal battery is a small nickel-cadmium bank of 3 cells. It seems that its only function in life is to supply the power for the RAM and internal clock chips in the event that it is not coming from one of the other two sources available. It "floats" across the internal power bus all of the time that the memory power switch is ON through a diode to prevent it discharging through the main power supply, and a current limiting resistor. If a power source is available, it is continuously being trickle charged, current limited by the resistor mentioned above. The internal power bus can get its power from either the internal AA cells OR from the external DC connection that the AC adapter plugs into. When you plug the AC adapter into the side, there is a switch which feels for the plug and disconnects the AA cells and connects the plug to the bus. This switch has been known to cause some people grief because the springiness can go out of it and it will fail to make contact with the AA cells when the AC adapter plug is removed (Thanks to Janet McC & others). There are two ways to fix this. First, replace it, and second, carefully bend it so it works. Bending it will not be permanent as it will eventually repeat the same type of failure. If you are a bit queasy about surgery on your Model 100, then arrange for RS to do it at probably a pretty steep cost out of pocket, otherwise have them order RS part# AJ-7338, manufacturers part# YJB03S001Z and install it yourself, being careful to arrange the wiring the same way, of course. In over a year of plugging and unplugging mine lots and lots of times, it has not failed me ..... yet. Anyway, there is a momentary loss of power when you plug in the adapter, since the switch first breaks the connection to the AA cells, then makes the connection to the plug as you push it in. The opposite happens when you unplug it. It is done this way to avoid shorting either the AA cells or the AC adapter. Since the AC adapter plug has disconnected the AA cells, it is therefore important to plug the adapter into the wall! If the adapter IS plugged into the side connection, and NOT plugged into the wall, or 110AC power is not being supplied, then, you are running the memory only from those little NiCd cells! The technical reference manual gives the RAM life under internal NiCd power only, as 40 days if only 8K is installed, and 10 days if 32K is installed. Referring to the memory effect mentioned in the previous file, it seems to me to be a good idea to exercise these little beasties to ensure they will retain thier rated capacity. What this means is that you would have to wait a LONG time without AA cells installed for it to drain the internal NiCds. From a practical point of view, this is not practical. Instead, I suggest running the thing till it drops off the line on its own. This hopefully will exercise the internal NiCds enough to retain some or most of thier capacity. If all else fails, have RS order RS part# ACS-0100, manufacturers part# ZBN036102Y, and solder it in place of the existing cells. The 3-cell pack is located on the component side of the main board about in the center left to right, and near the top. OK, I hear you ask, how do I get the thing apart? Well, before doing ANYTHING inside, I wholeheartedly recommend getting the "TRS-80 Model 100 Technical Reference Manual", RS catalog # 26-3810. For you more adventurous types, grab your phillips head screwdriver and follow me. Turn the M100 over so it is laying face down on a soft surface. I used a towel on the table. There are 4 screws at the corners. Remove them and put them where you can find them again. Turn the Model 100 over so it is again facing you with keyboard up. Now comes the hard part, separating the two halves of the case. Gently, but firmly (as the tech manual sez), pry the covers apart using ONLY your fingers! Careful! There are things inside which can be damaged. There are plastic clips which you have to separate here which can be a bit difficult. Once the case is open a little, be careful. There are wires connecting the two halves together. They will be on the right side as you lift the top, so turn the top over and to the right so as to not pull on the cables. Now what you should be looking at is the CPU board fastened to the back (the dark brown side), some cables and wires connecting the two halves, and the keyboard, LCD, and buzzer fastened to the top half (the light beige side). Remove the connectors from the main board. Wanna clean the keyboard while we're here? Sure, you do. There are 5 screws holding it to the top. One at each corner, and one in the center at the top. Remove them, and lift the keyboard from the case. Pull the keytops off the switch posts, and use the vacuum cleaner you already had ready to get the dust and such out of the works. Please try to put the keytops back in the same place! Draw yourself a map if need be, or you will cause yourself untold grief. If you haven't done it yet, take a break and go down to the drugstore and pick up dental rubber bands, a lot, each, at least one for each switch post anyway, you know, the kind which are used for holding braces on teeth and that friend of yours shot at you in grammar school. These are little rubber bands about 1/4 inch in diameter. Put one around each of the switch posts that the keytops go on. They will make the keys MUCH quieter and easier on the fingers as well. (Thanks to Alan Trick for this) No, you cannot do anything about the function keys. Those little devils are click-type diaphraghm switches commonly found in calculator keyboards (like that Bowmar Brain gathering dust in your closet). If one of these babies goes bad on you, replace it, as there is no way to clean them which I know of that works. Order RS part# AS-7570, manufacturers part# SK0101X10A. It is extremely difficult to remove, so be EXTRA careful. I recommend a good "solder sucker", a small wattage soldering iron, and a LOT of patience, else, to paraphrase SYSOP ^Dave^, "lack of patience begets patients", i.e. you and your dear departed Model 100. Remember, RS is going to charge you an arm and a leg, if not other parts of your body, to fix your mistakes, that is, if they even will touch it after you've fooled with it! Now for some more information on the power supply and how it behaves under the trying circumstances that you are putting it through. The internal power supply is essentially a DC-DC converter which takes the single supply, +5 or so, and synthetically produces two supplies, a +5 and a -5 supply. It is a neat trick how they do this, but the discussion is not appropriate for here, more geared to the hardware guru as it were. So, it takes the single supply voltage, and it makes two out of it. Power is then applied to the various parts of the main board to get the thing to do something i.e., "wake up", so to speak. What is important here is what happens when you turn it "off". You may have noticed that the "low battery" LED winks at you after you "hit the switch". This is perfectly normal. Here is what is happening. You turn off the switch, there is a capacitor across the input terminal of the converter. The converter now starts using this capacitor as a temporary power supply, like a real sort term battery. Of course, it will not last very long. The sense circuit looks at the voltage on the power bus. At 4.1 volts +/- 0.1 volts, it triggers a transistor to turn on the low battery LED. At 3.7 volts +/- 0.1 volts, the sense circuit turns off the oscillator circuit which is driving the DC-DC converter. At this point, RAM is placed in a "minimum power" state, analagous to tri-state devices being put into the high-impedence state. In other words, tell the chips to turn themselves "off" by setting the chip select line to a logic zero, which is the turn off condition anyway. Ok, so what this all means is that on a loss of power, for whatever reason, the low battery LED lights, the RAM is placed in an "off" state, and the rest of the machine dies, including the LED, and, well, everything, almost. Placing the RAM in the "off" state really means that the chips disregard ALL inputs, except for the CS (chip select) line, and "you guys are on your own. Keep the faith. So long, and farewell. We'll be in touch." The beautiful part of all this will be revealed shortly. Put in the following program and run it. 100 x=x+1 110 print x 120 for y=1 to 1000 130 next y 140 goto 100 This is a BAD example of programming as it places the machine in an "infinite loop". It counts, pauses, and counts, forever or until the machine dies of its own accord. Anyway, while you are running this, look at the number in the display, turn the on/off switch to "off" and watch it die. Then, sometime later, turn the switch back "on" and watch it resume EGG-ZACT-LY where it was when you hit the switch. Now that the demo is over, hit the BREAK, and break out of the loop, or it will go on forever, on until your Model 100 dies of natural causes. You wouldn't want that, now would you? Ok, that about does it for this installment. I have received suggestions about including programming tips here as well. Watch for forthcoming issues of POWER4.xxx and see what develops. Let me know of your experiences and/or your inputs with and for the Model 100. I will incorporate them if I can, and give you the credit in the text as I have done for the ones who contributed so far, though Janet McC did so unwittingly. Thanks, Janet! For putting up the problem you had, and those who responded to a damsel in distress. So, now, I have a problem ... what is to be the next topic of POWER4.xxx? I am at a loss. I have now covered between POWER4.100 and this file, all of the suggested items for this type of file. Where now? Feedback, I need it! Ralph V. Getsla [74405,1177]