Loading ...
Sorry, an error occurred while loading the content.

Crow Bar circuits

Expand Messages
  • joshbensadon
    Has anyone ever had damage done to their computer due to a voltage spike from a failed power supply? I keep thinking, it might not take much to have a
    Message 1 of 10 , Sep 6, 2012
    • 0 Attachment
      Has anyone ever had damage done to their computer due to a voltage spike from a failed power supply? I keep thinking, it might not take much to have a regulator or output transistor short and deliver the full unregulated voltage to a computer. Or if there's a variable resistor for voltage adjustmen, that could go open?

      I'm just wondering if it's worth my while to add crowbar circuits to some of my computers?

      Thanks for your input.

      Cheers,
      Josh
    • David Gesswein
      ... I have not had it happen. My 8/E does have a crowbar on the +5 as part of the power supply. Linears are the more likely to have this happen. Short of
      Message 2 of 10 , Sep 7, 2012
      • 0 Attachment
        On Fri, Sep 07, 2012 at 04:33:28AM -0000, joshbensadon wrote:
        > Has anyone ever had damage done to their computer due to a voltage
        > spike from a failed power supply? I keep thinking, it might not take
        > much to have a regulator or output transistor short and deliver the full
        > unregulated voltage to a computer.
        >
        I have not had it happen. My 8/E does have a crowbar on the +5 as part of
        the power supply. Linears are the more likely to have this happen. Short
        of switching transistors normally don't cause an overvoltage. Some
        switchers and linears have a separate over voltage shutdown circuit for
        regulator failure such as the trim pot going open. That doesn't protect
        against pass transistor failures on linears.

        It might not be a bad idea as long as you can put it in such that it
        doesn't cause problems. With linears you normally have some significant
        capacitors that will deliver high current that the crowbar has to handle.
        Also what will the power supply do with a short on the output? A false
        trip due to noise spike causing the crowbar to burn up the power supply
        will also be irritating.
      • Systems Glitch
        ... Not with an assembled computer, but I ve had a bench power supply fry a whole board of TTL and a really nice LED display. There was an intermittent ground
        Message 3 of 10 , Sep 7, 2012
        • 0 Attachment
          > Has anyone ever had damage done to their computer due to a voltage spike from a failed power supply? I keep thinking, it might not take much to have a regulator or output transistor short and deliver the full unregulated voltage to a computer. Or if there's a variable resistor for voltage adjustmen, that could go open?

          Not with an assembled computer, but I've had a bench power supply fry a whole board of TTL and a really nice LED display. There was an intermittent ground connection to the ground supply of a 3-terminal 5 Volt regulator. When it momentarily went open or Hi-Z, output voltage would jump to something above 10 Volts on the +5 rail. It was intermittent, so that supply took out a few things before I realized it was something with the supply. A crowbar circuit would have prevented that.

          On the other hand, I've got a 5 Volt modular supply that smoked a power resistor because the crowbar circuit falsely triggered. Most of them are composed of a trigger circuit and SCR, so the crowbar stays on until power is completely removed.

          Thanks,
          Jonathan
        • Bill Sudbrink
          ... Are we talking vintage or modern? If modern, yes. I had an Antec 450 watt supply fail and take the entire computer... motherboard, memory, CPU, video
          Message 4 of 10 , Sep 7, 2012
          • 0 Attachment
            Joshbensadon wrote:
            > Has anyone ever had damage done to their computer due to a voltage
            > spike from a failed power supply?

            Are we talking vintage or modern? If modern, yes. I had an Antec
            450 watt supply fail and take the entire computer... motherboard,
            memory, CPU, video card, hard disk and DVD player with it. After
            the "event" the supply was completely dead and made a delightful
            stink. I didn't have time or inclination to do a "post mortem" so
            I have no real idea of the details of the failure but the only usable
            electronics left were the case LEDs and the case fans.

            Bill S.
          • joshbensadon
            ... Ouch, I guess that power supply said If I m going down, I m taking everyone with me! . I m talking about a Mark-8 computer. It has a linear power supply
            Message 5 of 10 , Sep 7, 2012
            • 0 Attachment
              --- In midatlanticretro@yahoogroups.com, "Bill Sudbrink" <wh.sudbrink@...> wrote:
              >
              > Joshbensadon wrote:
              > > Has anyone ever had damage done to their computer due to a voltage
              > > spike from a failed power supply?
              >
              > Are we talking vintage or modern? If modern, yes. I had an Antec
              > 450 watt supply fail and take the entire computer... motherboard,
              > memory, CPU, video card, hard disk and DVD player with it. After
              > the "event" the supply was completely dead and made a delightful
              > stink. I didn't have time or inclination to do a "post mortem" so
              > I have no real idea of the details of the failure but the only usable
              > electronics left were the case LEDs and the case fans.

              Ouch, I guess that power supply said "If I'm going down, I'm taking everyone with me!".

              I'm talking about a Mark-8 computer. It has a linear power supply from Condor with the typical 723 regulator and what I thought was a 2N3055 but after removing the cover I found they are SDM4004 which look like darlington (hFE=1000) equivlents of the 3055.
              I haven't finished building the computer yet and was deciding if it's worth my while to install Crowbars on the 5V and -9V supply. Which after reading your story it's a definite yes.

              Jonathon and David talked about the Crowbar frying the power supply, but a proper Crowbar circuit should include a fuse right before the SCR.

              PS. I'm also helping a friend with an Analog Computer and I think it should have crowbar circuits installed too. There are no chips in it, only transistors, but I imagine these are all fragile transistors.
            • Bill Sudbrink
              ... On reflection, I should revise that a little. The keyboard, mouse and LCD display also survived. The repair ended up taking almost an entire
              Message 6 of 10 , Sep 7, 2012
              • 0 Attachment
                Joshbensadon wrote:
                > --- In midatlanticretro@yahoogroups.com, "Bill Sudbrink"
                > <wh.sudbrink@...> wrote:
                > >
                > > Joshbensadon wrote:
                > > > Has anyone ever had damage done to their computer due to a voltage
                > > > spike from a failed power supply?
                > >
                > > Are we talking vintage or modern? If modern, yes. I had an Antec
                > > 450 watt supply fail and take the entire computer... motherboard,
                > > memory, CPU, video card, hard disk and DVD player with it. After
                > > the "event" the supply was completely dead and made a delightful
                > > stink. I didn't have time or inclination to do a "post mortem" so
                > > I have no real idea of the details of the failure but the only usable
                > > electronics left were the case LEDs and the case fans.
                >
                > Ouch, I guess that power supply said "If I'm going down, I'm taking
                > everyone with me!".

                On reflection, I should revise that a little. The keyboard, mouse and
                LCD display also survived. The "repair" ended up taking almost an entire
                frustrating weekend. It was my wife's computer. Initial evaluation:
                "Oh, the PS died. I'll just pop out to MicroCenter and get a new one.
                You'll be back up and running in an hour." One new PS later (tested
                independently) and the machine still dead: "Well damn. The PS took the
                motherboard with it. It's going to take a little longer than I thought."
                I went ahead and bought a new motherboard, CPU and memory. Later I
                confirmed that the old memory and CPU were also dead. I put everything
                back together and got "beep codes" (Oh, I guess the speaker survived too)
                for no video adapter. I should have checked the hard disk and DVD drive
                right then but didn't. Back to MC for a new video card. Installed it
                and promptly got "No boot device". The HD wasn't spinning. Checked the
                DVD drive and found it dead as well. It should have occurred to me to at
                this point to be surprised that the keyboard, mouse and display were
                still working but I was so angry that I just stomped back to MC for a new
                HD and DVD. Then spent the rest of the weekend reinstalling the OS and
                all of my wife's applications.

                Bill S.
              • David Gesswein
                ... That was the purpose of my comment to make sure you thought about proper design of it. Commercial designs didn t always get it right. The current/energy
                Message 7 of 10 , Sep 7, 2012
                • 0 Attachment
                  On Fri, Sep 07, 2012 at 02:26:22PM -0000, joshbensadon wrote:
                  > Jonathon and David talked about the Crowbar frying the power supply, but
                  > a proper Crowbar circuit should include a fuse right before the SCR.
                  >
                  That was the purpose of my comment to make sure you thought about proper
                  design of it. Commercial designs didn't always get it right. The
                  current/energy the fuse can deliver before it blows is also an issue
                  on picking the SCR.

                  > PS. I'm also helping a friend with an Analog Computer and I think it
                  > should have crowbar circuits installed too. There are no chips in it,
                  > only transistors, but I imagine these are all fragile transistors.
                  >
                  You can look up a couple of the part #'s to see the breakdown voltage.
                  Since the analog computers weren't normally trying to use the fastest
                  transistors they may have a good margin on breakdown voltage.
                • joshbensadon
                  ... Ouch, that must have been very frustrating! The guys down at the store must be thinking You again? . Well, this would be one of those things in 20 years
                  Message 8 of 10 , Sep 7, 2012
                  • 0 Attachment
                    --- In midatlanticretro@yahoogroups.com, "Bill Sudbrink" <wh.sudbrink@...> wrote:
                    >
                    > LCD display also survived. The "repair" ended up taking almost an entire
                    > frustrating weekend. It was my wife's computer. Initial evaluation:

                    Ouch, that must have been very frustrating! The guys down at the store must be thinking "You again?". Well, this would be one of those things in 20 years you can look back at and laugh (if you're not laughing, add another 20 years).

                    > (Oh, I guess the speaker survived too)

                    Oh, that's good, that saved you $0.50 for a new speaker. (Sarcasm never comes through in email)
                    I guess that was a weekend of "what can go wrong next?". I've had my share of bad repairs, but not that bad. Well, chaulk it up to experience.

                    Cheers,
                    Josh
                  • joshbensadon
                    ... Are you saying the transistors in the old analog computers are robust? I haven t looked at part numbers yet, but they look like the old germanium TO-05
                    Message 9 of 10 , Sep 7, 2012
                    • 0 Attachment
                      --- In midatlanticretro@yahoogroups.com, David Gesswein <djg@...> wrote:
                      >
                      > You can look up a couple of the part #'s to see the breakdown voltage.
                      > Since the analog computers weren't normally trying to use the fastest
                      > transistors they may have a good margin on breakdown voltage.
                      >

                      Are you saying the transistors in the old analog computers are robust? I haven't looked at part numbers yet, but they look like the old germanium TO-05 case style. I have zero experience in Analog computers, I would love to pick someone's brains on them. New thread?

                      Cheers,
                      Josh
                    • s100doctor
                      Bill described what appears to be a catastrophic failure of a 21st century personal computer s power supply, that apparently took all the other components with
                      Message 10 of 10 , Sep 8, 2012
                      • 0 Attachment
                        Bill described what appears to be a catastrophic failure of a 21st century personal computer's power supply, that apparently took all the other components with it, drives and video cards and of course the "motherboard".

                        David Gesswein addressed some issues of analog computers which used discrete transistors (as opposed to vacuum tubes, for example).

                        But the original question was about crow-bar circuits on power supplies, or tantalum capacitor failures.

                        My point in calling this sequence out, is that I caution against trying to draw broad conclusions from very specific kinds of repair problems, component failures, and types of design - conclusions across different eras of personal computers. I think broad conclusions and comparisons are not productive. and even harmful.

                        Without posting a lecture about it......I think it's more reasonable to focus a discussion of computer repairs, to a fairly specific class of computers - specific time and place, models, designs - than to go off to talk about a five-year old computer failure in the same thread as 40-year-old transistors in an analog computer. It's like talking about an engine breakdown from an oil leak in a 2009 Toyota Camery, in the same discussion as a head gasket leak in a Model T Ford engine. Those are good discussions, but not the SAME discussion, in my informed opinion; what you learn from one won't help much with the other.

                        Herb Johnson
                        retrotechnology.com
                      Your message has been successfully submitted and would be delivered to recipients shortly.