Announcement

Collapse
No announcement yet.

is LT1085 current SOP

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • is LT1085 current SOP

    Been a few years, and I assume the LM317 is OBE. Is the LT1085 the current simple 3 terminal regulator?
    Need to whip up a nice clean linear supply for my crossover boards.

  • #2
    Quick drawing. It has been years since I played with LTSpice and with updates, I have lost some of my library consolidations. Have to remember all that stuff from a long time ago.
    Click image for larger version

Name:	Xkits ps.JPG
Views:	138
Size:	78.4 KB
ID:	1454448
    Attached Files

    Comment


    • #3
      The LT1083_4_5 have lower dropout than the LM317. People still use the LM317 though, it's dirt cheap.
      Francis

      Comment


      • #4
        LT's are claimed to be a bit better regulator, just as the 317 is a bit better than the old fixed ones.

        Diodes are probably not right. I want at least 500mA. Need to research the current thinking of fast-recovery vs Schottky. Pennies difference in price. Back in the day, it was 1N400x for everything. We used snubbers across the diodes too, but not sure needed with fast diodes.

        Lucky I have a handful of old 24V transformers. I'll see how they actually perform, may need to lower the output a little. Going for a best reasonable practice. Not cascading them, etc. Amazing what you forget in 10 years.

        Comment


        • #5
          Originally posted by tvrgeek View Post
          LT's are claimed to be a bit better regulator, just as the 317 is a bit better than the old fixed ones.

          Diodes are probably not right. I want at least 500mA. Need to research the current thinking of fast-recovery vs Schottky. Pennies difference in price. Back in the day, it was 1N400x for everything. We used snubbers across the diodes too, but not sure needed with fast diodes.

          Lucky I have a handful of old 24V transformers. I'll see how they actually perform, may need to lower the output a little. Going for a best reasonable practice. Not cascading them, etc. Amazing what you forget in 10 years.
          There's a good thread on DIYAudio about snubbers. The conclusion was the best results used almost any diodes, but put a good snubber across the secondary winding. I think the guy ended up with a 0.15uF cap in series with a 100 ohm resistor as snubber.
          Francis

          Comment


          • #6
            https://www.diyaudio.com/forums/powe...-test-jig.html
            Francis

            Comment


            • #7
              UF1002, 50 ns, 100V, 1A looks like a modern choice. Any favorites out there?

              Comment


              • #8
                Hmm, not in the mood to build a one-off test fixture. I may just use the horse shoe method. Something close. Sure with reputable transformer suppliers gave their capacitance and leakage specs.

                Good to see a thread without any magic properties and claims of perfection, when we are trying to reduce power line emissions as well as spurious RF that can be picked up. Looking inside my MUSE power supply I am re-boxing, I see they used a half wave with caps in parallel to the diodes. Looks like I have two supplies to fix. Horse shoe approach seems to be around 200nF and around 10 Ohms.

                Comment


                • #9
                  ESP says fast diodes are BS. Pass even suggests the parallel caps to slow down generic caps, 1N5401, 1N4002 etc. I wonder if different viewpoints are looking at different directions. Snubber to protect the power line, others to protect the load. Load should be protected by the regulator for low current, so the best practice may differ if low power vs power amp.
                  Nakamichi has a snubber on the primary side in their old preamp. That seems subject to a harsh environment.

                  Comment


                  • #10
                    LT3045 leaves the other regulators in the dust. Check it out!

                    Comment


                    • #11
                      I see some boards using it for sale. Does look very good. Current project needs only 300mA, single supply so might order one.

                      Also found TPS7A4700 /3301 and ULN-LD78 module 3-pin compatible and some rather expensive discrete modules. .Things sure have progressed since my last builds.

                      Been thinking what use some of the big coils I have left over from crossovers might be in a pi filter. I have some as large as 30mH. Calculator time. I tossed a LT1083 as a prefilter for my old DAC and dropped the power harmonics, but still 10dB above the floor.




                      Comment


                      • #12
                        Originally posted by tvrgeek View Post
                        I see some boards using it for sale. Does look very good. Current project needs only 300mA, single supply so might order one.

                        Also found TPS7A4700 /3301 and ULN-LD78 module 3-pin compatible and some rather expensive discrete modules. .Things sure have progressed since my last builds.

                        Been thinking what use some of the big coils I have left over from crossovers might be in a pi filter. I have some as large as 30mH. Calculator time. I tossed a LT1083 as a prefilter for my old DAC and dropped the power harmonics, but still 10dB above the floor.



                        Careful with big inductors ringing if you suddenly disconnect a load. Sometimes you need a good sized cap like 220uf in series with a 1 ohm resistor (typical values) to ground as damping. You'll see what I mean if you model a step load.
                        Francis

                        Comment


                        • #13
                          Even in a pi-filter? I'll toss it into LT Spice first of course. Bleeders across the big caps too.

                          Comment


                          • #14
                            Originally posted by tvrgeek View Post
                            Even in a pi-filter? I'll toss it into LT Spice first of course. Bleeders across the big caps too.
                            Be sure to check it with a step load. Yes, the L-C network can be high Q and ring like crazy when the load is removed, or switched.
                            Francis

                            Comment


                            • #15
                              Slowly remembering how to run Spice. Been about 10 years. For the coil sizes I have, up to 2mH, not a problem, but I found a 1 Ohm resister to be far more effective. Splitting into three sections should pretty much put my ripple below the floor. Then switch to a lower noise regulator and I will be pushing Johnson limits. That's hard to imagine from back in "olden times" when I was a lab rat working on DASD read recovery.

                              Comment

                              Working...
                              X