Image Map Image Map
Page 1 of 6 12345 ... LastLast
Results 1 to 10 of 56

Thread: Quest COSMAC ELF

  1. #1
    Join Date
    Jan 2011
    Location
    Vancouver, BC
    Posts
    1,995
    Blog Entries
    1

    Default Quest COSMAC ELF

    I won this piece quite a while ago but had left it at a freight forwarder while tending to my ill father for the last month. Anyway, I got home a few days ago and had them ship everything to me, and here it is.

    20170908_191531.jpg 20170908_192339.jpg20170908_192347.jpg20170908_192402.jpg20170908_192423.jpg 20170908_192649.jpg

    I've not found documentation for the Quest variant of the ELF yet (as opposed to the Super ELF, there is lots of documentation on that). General reading lead me to conclude I needed a 9V power supply connected to the + and - 'terminals' marked out on the ELF board. So I took a chance and hooked said up, and sure enough it lights up. For some reason, the builder opted not to go for the hex displays but LEDs instead. I've no idea how to 'read' what's going on there and can't really tell functionally if it's working. I did try playing around with switches and got lights to change. I'm confused about the data, LOAD, MP, etc switches being three position.. in videos it looks like they are only two position - on or off.. am I wrong on that? There's also an additional PROM switch, my understanding is the Quest ELF had a prom for some purpose.

    If anyone has any info on this particular variant or any very, very simple instructions on how to test functionality, I'd be most grateful. I think it's at least partially functioning but maybe not all the way.
    Last edited by falter; September 8th, 2017 at 06:28 PM.

  2. #2
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    1,308

    Default

    The circuitry on the little daughter card (with the LED displays) effectively make two binary hex displays. The LEDs are arranged as two columns of four LEDs (each column belonging to one of the HEX 7-segment displays they replace). The for individual LEDs (forming one effective display) are just the binary equivalent of a Hexadecimal character.

    The switches should be two position (single pole changeover). I suspect that your guy just used what he had lying around at the time. I suspect your switches may be ON-OFF-ON type (i.e. You will need to switch through the middle position quickly). You should be able to check the switch arrangement out with a multimeter.

    Dave

  3. #3
    Join Date
    Jan 2011
    Location
    Vancouver, BC
    Posts
    1,995
    Blog Entries
    1

    Default

    Quote Originally Posted by daver2 View Post
    The circuitry on the little daughter card (with the LED displays) effectively make two binary hex displays. The LEDs are arranged as two columns of four LEDs (each column belonging to one of the HEX 7-segment displays they replace). The for individual LEDs (forming one effective display) are just the binary equivalent of a Hexadecimal character.

    The switches should be two position (single pole changeover). I suspect that your guy just used what he had lying around at the time. I suspect your switches may be ON-OFF-ON type (i.e. You will need to switch through the middle position quickly). You should be able to check the switch arrangement out with a multimeter.

    Dave
    Thanks Dave for your (as always) invaluable insight. So if I understand you correctly, if I power up the machine and the top row has the two right most LEDs on and the bottom nothing, I read that as 20 ? (if I sort of remember how binary is read and added). I wonder with the LEDs why someone would go to the trouble of rigging that up? Seems like a lot more work to build that daughter board and interpret the LEDs vs installing two hex displays..

    I found the Quest documentation on the yahoogroup file section. I did try a couple of steps of their test program but the machine doesn't seem to react the way they expect. I'm really hoping I've got power connected correctly - in their instructions you are meant to connect a 12VAC adapter to the 'AC' pads on the board, and I think a fusible link. However it's clear looking at this board it was never set up that way. There's just four sets of posts where two capacitor should be, and the top most are marked + and -. Having seen other Quest ELFs using a 12VDC power supply I assumed I hooked up there. Hope I'm not doing it wrong.

    I'll have to figure out these swtiches too.. I'm pretty sure all the way up is ON.. but haven't quite figured out what the deal is with middle and down.
    Last edited by falter; September 9th, 2017 at 08:52 AM.

  4. #4
    Join Date
    Jan 2011
    Location
    Vancouver, BC
    Posts
    1,995
    Blog Entries
    1

    Default

    I did a bit of testing with my multimeter. I'm not really understanding it yet but I'll try to explain what I've found.

    Let's take Data switch 7 as an example. It has three positions, and three poles on the bottom that are soldered to different things. The top and bottom poles of all the data switches are chained together. The middle pole of each switch is soldered to the middle pad of where they're supposed to mount on the PCB. If I connect my DMM probes to the middle pad for switch 7, and the top pole of switch 7 and switch the switch into the down position, I get continuity between them. If it's in middle, nothing, if it's up, nothing. If I flip the switch to the up position, there is no continuity between either the top of bottom pole of the switch to the middle pad on the PCB. So it seems like the only way a circuit is made is if the switch is in the down position. Does that make sense?

  5. #5
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    1,308

    Default

    Yes, the LED 'values' will be a 'power of 2', so the LEDs will correspond to 8, 4, 2 and 1 respectively. Now which 'end' of the block of 4 LEDs is 8 and which is 1 you will have to work out for yourself by experiment.

    Can you point me at the schematic you have found please. I may be able to help you further if we are both using the correct schematic.

    Your switches do sound like ON-OFF-ON switches. Centre position being OFF and either end being ON. By wiring both ONs together, makes the switch OFF in the middle position, and ON when it is up OR down. That would be my guess anyhow. Most of the schematics I have seen for the ELF have the data switches switch between Vcc and Gnd. For your switches to work, there needs to be a pull-up resistor somewhere (if I am understanding what you are saying).

    Dave

  6. #6
    Join Date
    Jan 2011
    Location
    Vancouver, BC
    Posts
    1,995
    Blog Entries
    1

    Default

    Here's the schematic and some construction notes:

    https://drive.google.com/file/d/0B4p...w?usp=drivesdk

    I should note that that doc I think is for a version 2 board.. mine is either a version 1.0 or 10 depending on how you read it. It's possible I'm being hampered by bad RAM or something. The machine definitely isn't operating consistently. Sometimes when I try loading in data I can see blinks as the In button is pressed but the LEDs do not change. I did notice if I set the switches all to middle I consistenly get that 2 leds up top lit. But if I put the Load switch up all the way they go out. Don't think I have any spare 2101s around unfortunately to swap in.. that'd probably be a good suspect. I checked voltages and we have 5.0v on the nose at the ICs so that appears to be all happy.

  7. #7
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    1,308

    Default

    The original ELF had a full-wave bridge rectifier and a protection device fitted. You are correct in that yours doesn't - so only feed DC (not AC) to your unit!

    Your card appears to have 256 bytes of RAM and 32 bytes of ROM. It will be interesting to find out what is in the ROM when you get it working properly...

    With the ROM disabled, you will have RAM configured from addresses 0x00 through 0xFF. When the ROM is enabled, the ROM will appear between 0x00 and 0x1F. RAM will be at 0x20 upwards. This would be my first guess at looking at the schematics.

    There is a switch to enable the ROM.

    I see that the TIL311 displays have been removed. U11 is the least significant nibble with U9 being the most significant nibble of the address/data. Each display has four data lines on pins 12, 13, 2 and 3. I guess these have been wired directly to the discrete LEDs on the 'bodge board'. From looking up the data sheet, pin 12 = +8, pin 13 = +4, pin 2 = +2 and pin 3 = +1 when trying to convert from hex to binary. You will have to chase around from my above description (or the TIL 311 datasheet) to which discrete LED corresponds to what hex weighting.

    As for the switches, my first thought was correct. The 'middle' connection is the signal to the logic. Each other pole should be either connected to +5V or GND. I suspect the 'middle' setting is actually illegal...

    U15 pin 1 = switch S14 = D0.

    With the switch down = OFF = U15 pin 1 GND.

    With the switch up = ON = U15 pin 1 +5V.

    Hope this makes sense?

    Don't forget if you are measuring voltages with your multimeter, open circuit and GND will appear to be the same. To measure GND, put the red lead of your multimeter on +5V and measure with the black lead.

    Some of the manual was missing. Any chance of uploading the rest (with the checkout procedure). I may be able to come up with some simple tests to test the switches and LEDs for you?

    Dave
    Last edited by daver2; September 10th, 2017 at 04:55 AM.

  8. #8
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    1,308

    Default

    I assume your 'black' wire is actually positive and your 'red' wire is negative !

    I see that you don't have the parts for the speaker driver or the RAM battery backup nor the bridge rectifier etc. So (as you correctly said) you need to feed DC to the points marked '+' and '-' on the board. Anything above 7.5V DC should suffice for a 7405 voltage regulator. Anything above that just increases the temperature of the voltage regulator and costs more in electricity...

    Dave

  9. #9
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    1,308

    Default

    I have just started to trace out the PCB (from http://www.astrorat.com/images/Quest...al_Rev_2_1.pdf). I am now not sure whether UP/OFF/GND/DOWN/ON/+5V are correct as I stated back in post #7...

    The 'centre' wire of the switches should be the 'signals' to the logic. One side of most of the switches (let's just take the eight data switches for the moment) should all be connected to 0V/GND. The other side of the switches should be connected to +5V. For the data switches, one side should be connected to U15 pin 7 (0V/GND) and the other side of the switches should be connected to U15 pin 14 (+5V). You should be able to test this unpowered with a multimeter set to a low resistance setting.

    Next, the 'centre' pin of the switches (let's take S14 = D0 for example) should be connected to U15 pin 1. The ON/OFF setting for this switch should set U15 pin 1 to 0V and +5V when operated. If you have a 3-position switch; I would like to guess that the centre position does not connect U15 pin 1 to either the 0V or +5V rail. If this is the case - DO NOT USE THE CENTRE POSITION OF THE SWITCH FOR DATA ENTRY.

    You can check the other switches similarly.

    Note that the 'IN' switch should be a pushbutton. Is that the case in your set-up - or is it another 3-way switch?

    I think I have worked out what is going on with the RUN/LOAD/MP and ROM switches - but it will depend upon what you find out with your multimeter regarding the three positions of the switch. I suspect the original maker either bought the wrong parts - or this is what he/she had in their 'bits' box.

    Randomly operating the RUN, LOAD, MP and ROM switches will just cause confusion! They have to be operated in a specific order for things to work correctly.

    It would also be useful to power up the card and measure the voltage on U1 pin 3 whilst operating the RUN switch and U1 pin 2 whilst operating the LOAD switch. These two pins are key to the start-up operation of the CPU.

    If you can let me have some feedback regarding your multimeter measurements, that will help me with identifying the next stage.

    If you can find the second half of that PDF it would be helpful - as I think it goes on to describe the checkout procedure.

    Regards,

    Dave
    Last edited by daver2; September 10th, 2017 at 09:03 AM.

  10. #10
    Join Date
    Jan 2011
    Location
    Vancouver, BC
    Posts
    1,995
    Blog Entries
    1

    Default

    Here's the second half, Dave.

    https://drive.google.com/file/d/0B4p...ew?usp=sharing

    I will get out the DMM and do some testing and let you know. My 'in' switch is a momentary - you push it down to activate it and it springs back up. Only two positions it can be in.

    What I tried to do was follow the directions they have in the guide for testing out the ROM functionality, but I did not get the anticipated code back on the LEDs on the third or fourth step. Now again, I was assuming that the switches were wired in a way that when the instructions said 'flip up' I flipped it all the way up. Certainly flipping the PROM switch all the way up does cause an LED change. It's just the rest of what they tell you should be happening isn't happening. But perhaps we're not understanding how these switches are set up.

    And yes, my lead colours are backwards. At the time I was thinking only of how to differentiate them so picked two random pieces of wire not realizing I put the polarity opposite to the way those colors normally are.

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •