Image Map Image Map
Results 1 to 6 of 6

Thread: Model III Power Supply Problem

  1. #1
    Join Date
    Feb 2015
    Location
    Cleveland, OH, USA
    Posts
    1,186

    Default Model III Power Supply Problem

    Short description: After modifying a working Model III cassette machine I find it will not power the motherboard.

    More detail: Upgraded memory from 16 to 48k. Replaced the usual suspect capacitors on the power supply. Replaced the video potentiometers. Added an RS-232 and FDC board acquired from eBay. Added a 12VDC fan. Added DSDD and Gotek drives, custom cables. Added a second +5VDC / +12VDC PSU to power the fan, FDC and disk drives.

    Diagnostics: When I powered up the second PSU produced power to the FDC and FDs: the drive lights came on, drive :0 briefly, and drive :1 (Gotek) displaying on its LCD, and the fan spun. It looked like normal behavior from everything attached to the second PSU, in other words.

    There was no other activity: nothing displayed on the CRT even after turning both pots to either limit. As the lid was off I saw no glow from the CRT heater filament. I disconnected the second power supply to see if it made a difference (it did not) and switched PSU plugs (no change). I checked for a blown fuse on the TRS power supply; it has none.

    I verified 120VAC at both power supply cables. With 120VAC known at the Astec PSU input, I attempted to verify output of the supply. To do this I removed the CPU lead from the motherboard and inserted pins in +12V and GND in order to read voltage with a voltmeter. When I turned on power again I noticed the CRT heater glow. When I reconnected the power lead to the CPU and reapplied power the CRT heater did not glow.

    Here I am at the limits of my knowledge. Obviously something is connected or wired wrong; and probably the PSU is shutting itself down to protect itself. I would think the common strategy of minimal configuration would apply here: Can I run the PSU with the motherboard only connected - will that be enough load for the PSU? What else might I look for? Bad RAM? Reversed RAM? Wrong jumper setting?

    Thanks for your replies,

    -CH-

  2. #2

    Default

    Possible 12V short on the motherboard? The fact that the CRT glows only when you disconnect the motherboard power connector leads me to consider that. Check the PS output at the PS itself with the motherboard connected/disconnected. If 12V is shorted then try reversing any changes you made to the motherboard to see if that helps.

  3. #3
    Join Date
    Mar 2013
    Location
    Chaffee, MO
    Posts
    1,460

    Default

    One thing comes to my mind is with the memory upgrade. The Original RAM IC's used -5 VDC, from a supply
    on the Aluminum Metal Back plate that is behind the Motherboard. The -5 VDC goes to Pin 1 of each
    RAM IC Socket. The 416 RAM IC's used that -5 VDC Supply. The RAM IC's you have installed now
    are likely 64K RAM IC's and if you have used Pin 1 of the Sockets on the Motherboard that could be the
    issue. I don't know what RAM IC's you used or what Upgrade path you used. If a PCB trace was cut and
    modified for the existing RAM and the -5 VDC not used then that's not the problem.

    You should be able to use your Meter to check each feed of +12 and +5 VDC to VERIFY that nothing
    downstream is ZERO OHMS or about ZERO OHMS, causing the current limit to fold back the supply.

    Are you using two 38 Watt Supplies or a 38 and a 65 Watt?


    Larry

  4. #4
    Join Date
    Feb 2015
    Location
    Cleveland, OH, USA
    Posts
    1,186

    Default

    Thanks to Pete and Larry for their responses. After consideration I realize I have no choice but to go to square one and verify the build at each step. I will let you know how I fare.

    To answer Larry's questions: the second power supply is a Meanwell PD-65. Its output is 65 watt. Data sheet here: www.meanwell.nl/Images/pdf/datasheets/PD-65-spec.pdf. The TRS power supply is 35 watt.

    The OEM memory was 16K (one bank of 8 ICs) Mostek. I replaced it with three banks - 48K - of TI TMS-4116-15NL. Or so it says on the stamp. Whether it is genuine issue and operational is an open question. See picture attached.

    -CH-

    RAM.jpg

  5. #5
    Join Date
    Mar 2013
    Location
    Chaffee, MO
    Posts
    1,460

    Default

    CH,
    Just make sure the RAM has 7 Bit refresh rate versus 8 Bit refresh. It HAS to be 7 Bit Refresh rate.

    Just check each Conductor feeding it's load with your Meter (with Load(s) disconnected from Power Supply),
    to see what is a near Zero OHM load. It has to be something that is causing the Power Supply to fold
    over in Current Limit. I'd start by just powering up the Motherboard and maybe one Floppy Drive as the
    real load. With the RS-232 and FDC Power Disconnected. Then add maybe the RS-232, then the FDC, and
    the GOTEK.

    Larry

  6. #6
    Join Date
    Feb 2015
    Location
    Cleveland, OH, USA
    Posts
    1,186

    Default

    The 4116 has seven address lines, A0 - A6. In order to specify all 16,384 memory addresses it would have to receive the requisite address in two segments. It has inputs for -5V, +5V, +12V and ground. See attached screenshot of the datasheet. As far as I know, and I asked Ian, the 4116 is a drop-in replacement for the Tandy memory IC.

    I took everything apart; I re-soldered the AC inputs on the power supply just in case, and I detached and checked each item that was connected to the PSU independently of one another, looking for low-resistance situations. The video board read about 25 ohms between 12V and ground; I don't know if that is low or reasonable. But when I powered up the PSU with the video board only attached, the PSU put out 12V and 5V.

    The motherboard measured > 2K ohms on all lines, 5V being the lowest, but when I powered up the PSU with both the video board and the motherboard attached the PSU output voltage dropped to half on both 12V and 5V lines. It may be that the PSU is at fault, incapable of meeting the demand, but I don't have an extra to substitute.

    It may also be that while mounted in the chassis something else was causing a path to ground. I removed the motherboard and will replace the "new" RAM with the "old" and re-check the jumpers. I hear there is a nifty reference just for that purpose.

    'Tis a puzzlement.

    -CH-

    TMS4116.jpg

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
  •