Announcement

Collapse
No announcement yet.

Mosfet substitute questions

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

  • Mosfet substitute questions

    I need a sub for the 2SK1300 which is too hard to find at a decent price. I think I can use a FQP13N10L but figured I'd double check if any concerns. Both datasheets attached.

    This is an odd circuit, for the power amp, think of a Twin with transistors.
    Output transformer, single positive supply rail feeding OT. Two pairs of 2SC type bi-polar power transistors getting around 35VDC via the OT. Output transistors driven by these 2SK1300 Fet's, one for push, one for pull side. The push vs pull are too imbalanced, I think one was damaged when one side of the outputs blew.
    Now with the original Fet's, I get 'push' output xistors idling around 40mA each, and 'pull' xistors running around 180mA each. So too far unmatched as far as I can tell, plus I think it's fair to assume the driver was damaged or stressed when the outputs blew.

    The 13N10L I think looks like a good match, but does not have the Vgs clamping internal zeners. Do I need to add them for this application, and if so, what voltage?
    Attached Files
    Originally posted by Enzo
    I have a sign in my shop that says, "Never think up reasons not to check something."



  • #2
    I jumped to say the original ones would be lateral while FQP-anything would be verticals, but reading datshaeet carefully they just look like very good very sensitive verticals 8switching type).

    IF original amp worked with that type Mosfets then your replacements do not sound that far off.

    Try to lift and post at least the power transistor area, plus biasing and drivers, so we have a better idea.
    Juan Manuel Fahey

    Comment


    • #3
      Originally posted by J M Fahey View Post
      Try to lift and post at least the power transistor area, plus biasing and drivers, so we have a better idea.
      This is one of those 'locked up' schematics that I can't get. (Atlas AA120).
      It's a modern build, so not easy to draw up. I was hoping if I described the circuit someone might be able to post a similar amp schematic that I could alter with any differences.
      They're using an OT because it's designed for 70V or 100V line distribution system. Although, most designs I've seen for those use auto-transformers rather than this standard style OT. I'm used to split supplies, this is the first SS amp I've ever seen like this.
      Originally posted by Enzo
      I have a sign in my shop that says, "Never think up reasons not to check something."


      Comment


      • #4
        I wasnīt asking you to crop the Factory schematic but to grab pencil, paper and good light and draw just OT, power transistors, drivers if any and whatīs biasing them.
        Less than 10 components in total.
        Juan Manuel Fahey

        Comment


        • #5
          Well here's some quick and dirty, bottom is just mirror of top. Secondary of OT is just regular but with 70V and 100V line taps as well as an 8 ohm tap.
          There is a lot of other stuff on this power board, if it were just 10 parts I would have done it long ago. Most of it goes back to another board which has a trim pot on it, bias adjust I would think.

          Click image for larger version

Name:	aa120pwr.jpg
Views:	1
Size:	34.7 KB
ID:	853116
          Originally posted by Enzo
          I have a sign in my shop that says, "Never think up reasons not to check something."


          Comment


          • #6
            I know, I know.....electronic parts from Walmart? I just found this and will pass it along. Do what you will with the info.

            https://www.walmart.com/ip/STP16N10L...SABEgKU8vD_BwE
            "I took a photo of my ohm meter... It didn't help." Enzo 8/20/22

            Comment


            • #7
              Originally posted by The Dude View Post
              I know, I know.....electronic parts from Walmart? I just found this and will pass it along. Do what you will with the info.

              https://www.walmart.com/ip/STP16N10L...SABEgKU8vD_BwE
              Walmart using Amazon's model where businesses sign up and use Walmarts infrastructure .
              I've used it and it works
              nosaj
              soldering stuff that's broken, breaking stuff that works, Yeah!

              Comment


              • #8
                Originally posted by g1 View Post
                Well here's some quick and dirty, bottom is just mirror of top. Secondary of OT is just regular but with 70V and 100V line taps as well as an 8 ohm tap.
                There is a lot of other stuff on this power board, if it were just 10 parts I would have done it long ago. Most of it goes back to another board which has a trim pot on it, bias adjust I would think.

                [ATTACH=CONFIG]52589[/ATTACH]
                What I meant was not the full amp but the output section, which you did.
                Some 16 parts so far
                We just need bias.

                Ok, the point is, and GLAD for this drawing, that the MosFet is NOT the power transistor which so far I had assumed, but a plain "driver buffer", a simple unity source follower.

                Weird choice, maybe designer wanted to simplify drive, so he had the couple Amperes needed to drive 2SC5200 base but itself was a "no base current" transistor.

                So he can drive it with "anything" , from a humble Op Amp to a low current (a couple mA) TO92 transistor to even a cathodyne.

                My point being that unless very damaged, your problem does not lie in the MosFet but in the real muscle: the 2SC5200 quad.
                You might have one of them open (or associated resistors) so half semicycle can now supply half the needed current ... not a MosFet fault .

                OR you have a bad/unbalanced drive signal, which power stage will copy faithfully.

                so:

                first check DC voltages; Vbe at each power transistor; V across each emitter resistor, bias trimmer should let you get, say, 10mV across each or somewhat less, the point being that all show at least "something" ... failure to do so makes me suspect an open device.

                second: AC: drive amp to clipping , where I suspect you will find both halves asymmetrical (per your earlier description), scope all emitters, "should" be about same; Bipolar bases to ground (horrible waveforms but again we are searching for similarities/differences) and the cherry on the cake, both MosFet gates.

                If signal is already unsymmetrical there, power pack is fine, must look earlier up the chain.

                Ok, test and post

                EDIT: I had assumed it was a Guitar amp (given the Twin comparison) but now I notice itīs a typical Walmart/school/train station PA amp.
                Attached Files
                Last edited by J M Fahey; 02-19-2019, 01:19 PM.
                Juan Manuel Fahey

                Comment


                • #9
                  Originally posted by g1 View Post
                  Well here's some quick and dirty, bottom is just mirror of top. Secondary of OT is just regular but with 70V and 100V line taps as well as an 8 ohm tap.
                  There is a lot of other stuff on this power board, if it were just 10 parts I would have done it long ago. Most of it goes back to another board which has a trim pot on it, bias adjust I would think.

                  [ATTACH=CONFIG]52583[/ATTACH]
                  Are you sure the Mosfet drains connect to the power transistor collectors and not to B+?
                  - Own Opinions Only -

                  Comment


                  • #10
                    Originally posted by J M Fahey View Post
                    My point being that unless very damaged, your problem does not lie in the MosFet but in the real muscle: the 2SC5200 quad.
                    You might have one of them open (or associated resistors) so half semicycle can now supply half the needed current ... not a MosFet fault .

                    OR you have a bad/unbalanced drive signal, which power stage will copy faithfully.
                    Sorry, I did not give a complete explanation in the first post.
                    I replaced the burnt parts, pair of 2SC's and 1K resistor on the bad side. The Mosfet gave an 'ok' reading on the meter and was not available so I thought I would try without replacing it first. Brought up to voltage slow with variac and all was ok except imbalance of idle currents. Signal looks good on scope right up to full power into load.
                    One side running hotter than the other, approx. 18mV per emitter resistor vs. 4mV on the other side (idle). Running at load I can feel quite a difference at the heatsink.
                    So I swapped the Mosfets around and the higher current stays with the mosfet (the one who's source resistor burnt).
                    This confirms the Mosfet as the issue so I just need to find out if FQP13N10L sub appropriate and whether the Vgs clamp zeners were needed for the design or just coincidental to the part they chose.

                    Originally posted by Helmholtz View Post
                    Are you sure the Mosfet drains connect to the power transistor collectors and not to B+?
                    Yes. The Mosfet drains and power transistors collectors all connect to the OT primary as shown. There is no B+ connection to this board at all.
                    Originally posted by Enzo
                    I have a sign in my shop that says, "Never think up reasons not to check something."


                    Comment


                    • #11
                      Yes. The Mosfet drains and power transistors collectors all connect to the OT primary as shown.
                      Thanks. That would make the configuration a kind of hybrid "Darlington".
                      - Own Opinions Only -

                      Comment


                      • #12
                        https://www.westfloridacomponents.co...ransistor.html

                        Comment


                        • #13
                          Thanks guys. The FQP13N10L I mentioned in the first post are my cheapest option due to shipping to my locale.
                          I was hoping to get some input on the Vgs clamps I asked about so I could include any zeners in the part order that I need to put in this afternoon.
                          Originally posted by Enzo
                          I have a sign in my shop that says, "Never think up reasons not to check something."


                          Comment


                          • #14
                            1) replace **both** Mosfets

                            2) you didnīt post needed DC voltages, besides those across emitter resistors.
                            Still need Vbe at bipolars and Vds at MosFets to have a better picture of whatīs happening and blame either MosFets or unsymmetrical biasing.

                            3) since you get them cheap, buy extra Mosfets, so if necessary matched pairs can be selected.
                            Those are general purpose Mosfets so you are not "wasting" anything, may be useful elsewhere; postage is more expensive than parts themselves.

                            Iīm designing a new Fet compressor since LDRs are getting hard to find, "should" have bought 10 to experiment or 100 for initial production stock but since I bought from Mouser USA *and* had to go through Argentine Customs, besides paying 1 kg air freight as "minimum package" and hiring a registered "customs handler" , I straight ordered 1000 of them, go figure.
                            So now freight+tax+customs expenses add up to 50% surcharge ... instead of 500% in a smaller order

                            Side note: since problem "follows Mosfets" we "should" find some voltage difference around them.

                            4) without reredading datasheet , in general Vgs has no problem until 20V o so, while they can be passing lots of current (not even needed here) with +10Vgs, I think a safe seat-of-the-pants value should be series back to back 10V or 12V Zeners across GS so voltage reaches 10V unhindered (all the current you want) but canīt go higher than 11 to 13V. (safe). Just in case recheck datasheet values.
                            Juan Manuel Fahey

                            Comment


                            • #15
                              Yes, getting some extras and matching was the plan.
                              Base voltages at bi-polars were in keeping with currents, from memory approx. .55V for the ones idling lower current, and approx. .6V for the ones idling hotter.
                              Again, voltages and idle currents followed the Fet's when I switched them around.
                              I'll recheck and post all fet voltages before and after replacement, thanks.
                              Originally posted by Enzo
                              I have a sign in my shop that says, "Never think up reasons not to check something."


                              Comment

                              Working...
                              X