Announcement

Collapse
No announcement yet.

Hiwatt design question

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

  • Hiwatt design question

    Hello, I.m wondering why fixed bias for LTP into Hiwatt designs. What do You think Mr. Reeves looking after doing like that, please ? Thanks.
    "If it measures good and sounds bad, it is bad. If it measures bad and sounds good, you are measuring the wrong things."

  • #2
    Originally posted by catalin gramada View Post
    Hello, I.m wondering why fixed bias for LTP into Hiwatt designs. What do You think Mr. Reeves looking after doing like that, please ? Thanks.
    For the extreme clean headroom, fixed bias allows it to be locked in where as a floating bias would vary depending on various things.

    nosaj
    soldering stuff that's broken, breaking stuff that works, Yeah!

    Comment


    • #3
      Thanks. ...well, how much it count to use a half a valve as low impedance reference and not to use a simple resistive divider instead ? From a ignorant perspective seems to be a waste but maybe I.m wrong...can You enlight me , please? Thanks
      ( we are talking about A.C. coupled version)
      "If it measures good and sounds bad, it is bad. If it measures bad and sounds good, you are measuring the wrong things."

      Comment


      • #4
        Above my knowledge level but you maybe able to get some info from Mark Huss's website https://hiwatt.org/
        nosaj
        soldering stuff that's broken, breaking stuff that works, Yeah!

        Comment


        • #5
          Originally posted by catalin gramada View Post
          ( we are talking about A.C. coupled version)
          The later PI version as described in the Mark Huss link from post #4 ?
          "The earlier circuit (PI 1) used a cathode-follower directly connected to the PI circuit to both set the DC level and buffer the signal.
          The middle and later units (PI 2) capacitively coupled the last preamp stage to the PI. The former cathode follower section now has its input connected to a DC voltage divider, and is used strictly as a low impedance voltage reference."

          There have been a few discussions previously that may give some ideas:
          https://music-electronics-forum.com/...ad.php?t=29728
          https://music-electronics-forum.com/...ad.php?t=39917
          https://music-electronics-forum.com/...ad.php?t=43403
          Originally posted by Enzo
          I have a sign in my shop that says, "Never think up reasons not to check something."


          Comment


          • #6
            Great! Thanks.
            "If it measures good and sounds bad, it is bad. If it measures bad and sounds good, you are measuring the wrong things."

            Comment


            • #7
              The direct coupled cathode follower follows previous gain stage plate voltage to a T (within 2V DC) , which may or may not be the best for maximum symmetrical signal output, you are fully dependent on tube characteristics which as known vary and drift a lot.
              And most probably best plate voltage on the first tube and best cathode and plate voltages on the second do not match, so Mr Reeves separated them, so he can optimize them in an independent way.
              Rather than a "waster", he showed to be a beyond average competent designer.
              Many small tweaks like that contribute to a significantly better final product, Hiwatt amps deserve their high reputation.
              Juan Manuel Fahey

              Comment


              • #8
                I guess the subject is the cathode follower used as a voltage regulator to feed the two 1M resistors that connect to the grids of the long tail phase splitter as in the DR504OL. Yes, seems like a waste of a tube section to me. Pull out the tube out, connect the 1M's to the voltage divider that was connected to the grid of that tube, and put a C to ground. I doubt that you could tell the difference.

                The only difference that I see is that the cathode follower holds the voltage a bit more constant when the two grids of the LTPS draw different amounts of current during overload with an asymmetrical waveform. It is a small difference, but if you think that you need a bit lower dc source impedance, lower the values of the resistors in that divider.

                Comment


                • #9
                  It is curious. If I were to use 1/2 of a 12ax7 as a voltage reference like that I would at least shunt the grid to ground with a cap. As it is, the bias on the LTP gets not just 1/5th of the DC of that particular B+ node, but 1/5th of the AC as well.

                  Honestly, I've never studied Hiwatt amps or their history. It wasn't until I looked a the Mark Huss page that I realized Dave Reeves worked at Mullard. I'm sure many of you are familiar with the book that Mullard published with a variety of hi-fi designs, from simple to extreme. One of the distinguishing features of the Mullard designs is the direct coupled LTP, which in those designs was always direct coupled to the previous stage. That was the whole point: to remove one reactive component (coupling capacitor) from the signal path. This made it easier to apply NFB and get wide frequency response.

                  Anyway, I'm not going to get into the head of Mark Reeves, but the fact that he had ties to Mullard is interesting to me. Who knows if he talked to the audio design engineers there, maybe even the same one that was responsible for all the designs in that book, but never got personal credit.

                  But, no, that cathode follower voltage reference makes no sense at all.

                  Comment


                  • #10
                    The point of the circuit is to precisely fix the operating point of the PI, regardless of tube characteristics.

                    The B+ voltage for the PI voltage reference (intentionally) comes from the furthest point down the RC chain - same as the input stage - so there's very little ripple to worry about. I think we can all agree that a low impedance voltage reference is better than a high impedance one (which would certainly be the case for a simple 1M/2M voltage divider). I suspect the mid-70s version, with the DC-only voltage reference design was an intentional improvement.

                    If Dave Reeves was a bean counter (instead of a 'best possible product' guy) he doubtlessly could have saved one ECC83. However, he was not a fool, so he wouldn't put an extra one in there unless it had some demonstrable engineering benefit.

                    Comment


                    • #11
                      Maybe the main benefit goes to the mass production to get with "one wire mod" a real dc CF. into circuit and get sort of compression with that for alternative models on same platform. One of extra hole in chassis Reeves claimed against Sound City circuit patent request can serve for a switch mode as well ...
                      "If it measures good and sounds bad, it is bad. If it measures bad and sounds good, you are measuring the wrong things."

                      Comment


                      • #12
                        Originally posted by mhuss View Post
                        The point of the circuit is to precisely fix the operating point of the PI, regardless of tube characteristics.
                        The long tail already does that. How does the cathode follower make that any better?

                        Comment


                        • #13
                          Meanwhile I bought this one today in original condition, meant Partridges and Mullards.Pretty much what Reeves put in his amps. It looks like original Hiwatt circuit, power stage at least. I will wire the CF in both circumstances to see what differences will do.

                          Click image for larger version

Name:	20181028_215449.jpg
Views:	1
Size:	2.26 MB
ID:	851946
                          Last edited by catalin gramada; 10-28-2018, 09:09 PM.
                          "If it measures good and sounds bad, it is bad. If it measures bad and sounds good, you are measuring the wrong things."

                          Comment


                          • #14
                            Originally posted by catalin gramada View Post
                            Meanwhile I bought this one today in original condition, meant Partridges and Mullards.Pretty much what Reeves put in his amps. It looks like original Hiwatt circuit, power stage at least. I will wire the CF in both circumstances to see what differences will do.

                            [ATTACH=CONFIG]50930[/ATTACH]
                            Nice score!
                            If I have a 50% chance of guessing the right answer, I guess wrong 80% of the time.

                            Comment


                            • #15
                              Originally posted by SoulFetish View Post
                              Nice score!
                              Thanks
                              And an other candidate to be butchered, bought today for little cost of parts

                              Click image for larger version

Name:	20181111_173049.jpg
Views:	1
Size:	2.65 MB
ID:	852045

                              Click image for larger version

Name:	20181111_173136.jpg
Views:	1
Size:	2.66 MB
ID:	852046

                              Power tubes was all ready replaced but full of Brimars inside.
                              Heavy as hell...
                              Last edited by catalin gramada; 11-11-2018, 07:41 PM.
                              "If it measures good and sounds bad, it is bad. If it measures bad and sounds good, you are measuring the wrong things."

                              Comment

                              Working...
                              X