Batwing transistor questions (KA-3500)

Discussion in 'Kenwood-Trio/Kensonic-Accuphase' started by mikepick, May 1, 2018.

  1. mikepick

    mikepick New Member

    Messages:
    29
    Hi folks:

    I've acquired a KA-3500 for not much money, it's a nice looking little amplifier and the casing / faceplate is in pretty good shape.

    It came to me with a blown left channel, so I'm working through trying to restore it. I made an initial shotgun pass to recap, and that did not help. So I've gone back to examine it more closely by checking voltages and so forth, and I have a batch of transistors to sub in as well, aside from the output transistors. There is a good signal coming out of the preamp into the power board, and I've already found out-of-spec voltage at R7 / Q1.

    It's sort of hilarious what I did not spot on my first pass through, you can see here that the top of the transistor Q9 is sheared off, and there are apparent cracks in the left channel output transistors. (The right side outputs are not cracked.)


    I haven't pulled the outputs yet for testing, but given the cracks the outlook is not good (in my limited knowledge), and so I've been doing the research into replacing them. I've read a few threads here where people have used the old transistors to hold down the new ones.

    If necessary, I'm looking at using the FJP5200RTU / FLP1943RTU to replace the 2SB618 / 2SD588.

    My first question is about pinout: according to what I find the 2SB618 / 2SD588 have a collector on the batwings as well, and the FJP5200RTU / FLP1943RTU do not seem to, if I am reading it right. If so, do I need to rig a connection from the collector to the heatsink?

    My second question is about mounting: are the batwings absolutely necessary? The FJP5200RTU / FLP1943RTU look to have a heat sink tab with a hole already, would I theoretically be able to drill and tap new holes to use to mount the transistors?

    Finally, I'm looking for good info on replacing outputs as well: thermal paste or mats, connections, etc. If any threads come to mind, I'm looking to read up on it.

    Thanks for reading this long-winded post. Any comments are appreciated.
     

     

    Please register to disable this ad.

  2. transmaster

    transmaster Addicted Member

    Messages:
    7,634
    Location:
    Cheyenne, Wyoming
    Here is a very good book mark for you.
    https://alltransistors.com/
    First click on the transistor type the you enter the spec’s for the transistor, or enter the part number and it will give you substitutions for it.
     
  3. quaddriver

    quaddriver 120 What's per channel Subscriber

    I replaced exactly those outputs in a 4070. the collector tabs were 100% isolated from everything, so keep that the same...use a film and grease and a plastic shoulder washer.

    btw - you can get for like $10 a trans mounting kit - many m3 screws, nuts, washers, micas etc, from amazon

    ps: the thread I did this in was '4070 went silent' or words to that effect. I worked out all the subs for all the components of the amp silicon, if you have similar - work done for ya

    eta: cracked or blown apart silicon = dead silicon.

    safe bet they have long since left the scene AND the low value resistors on the amp should go as well. dial up your mouser acct..

    eta2: the SM is on hifi-e. the SCM part and parts list ok, the pictures - crappy. Its similar power and age as the 4070, but different topology and no protection. you are going to be able to sub in inputs, vas, bias, but the drivers and pre-drivers need to be worked out.

    for the cost differential, whatever you do to side 1, do to the other as well...
     
    Last edited: May 1, 2018
  4. mikepick

    mikepick New Member

    Messages:
    29
  5. mikepick

    mikepick New Member

    Messages:
    29
    Aha! Thanks a bunch, @quaddriver ! The pics of your output substitutions are exactly what I was looking for, a drill and tap seems to be the way to go. Looks like I'm pulling all the pins on that board again...
     
  6. quaddriver

    quaddriver 120 What's per channel Subscriber

    2SA620 and 2SA640 can both be replaced with KSA992F and is appropriate for that amp

    AND, that sub list has you heading to KSA1220/KSC2690 pairs for the drivers, and the modern 1220/2690s are real good in that position as well
     
    UncleBingo likes this.

     

    Please register to disable this ad.

  7. mikepick

    mikepick New Member

    Messages:
    29
    Thanks again @quaddriver !

    Reading through your thread, I think I will do a taped-on test for the outputs before drilling. The way the 3500 heatsink is constructed, its not going to work to saw some of it off if the alignment is not right. :D
     
  8. transmaster

    transmaster Addicted Member

    Messages:
    7,634
    Location:
    Cheyenne, Wyoming
  9. mikepick

    mikepick New Member

    Messages:
    29
    Update! I managed to find some time in the last couple of days to work on this amp.

    Over the last few weeks I amassed all the transistors, including the output transistors I listed above as replacements, and yesterday I went through and replaced them all.

    fullsizeoutput_54d.jpeg

    I also replaced C1 and C2 with new film caps and while I was at it, I noticed that resistors R41 and R43 were looking crispy, so I pulled them and sure enough, they were open circuit, so I replaced them with metal film resistors I had around.

    I drilled and tapped the heat sink to attach the new FJP1943 and FJP5200 out transistors and used a transistor mount kit from Amazon to attach them. (Had to buy a #4-40 tap as well, which I used with a 3/32 bit.)

    fullsizeoutput_54b.jpeg

    I ended up placing the holes 20mm from the bottom, and centered between the prior batwing holes. This gave me more than enough pin below the board to bend them back against the contacts.

    Here it is reassembled, you can see the batwing detritus behind:

    fullsizeoutput_54c.jpeg

    Finally, this morning I reattached the wiring and mounted the board in the case. Fired it up with a dim bulb, which went as hoped, and then proceeded to measure voltages on the power board to compare to the schematic. All seemed ok, so I tried a headphone test. Sound in both channels!

    I tested the DC voltage on the speaker outputs and it was ~1.5-2mV, so I hooked up speakers and gave it a whirl. Sound in both channels, and it sounds great!

    The only issue I am having now is in trying to set the bias. I'm fairly certain I am measuring in the right connections, but I get a steady 3.5mV across the resistors on both channels and it does not respond to turning the bias controls pots, so I'm not sure what is happening there...
     
  10. rjsalvi

    rjsalvi Active Member

    Messages:
    487
    Location:
    San Diego, CA
    Nice job!
     
  11. mikepick

    mikepick New Member

    Messages:
    29
    Thanks! I'm not so sure yet, the bias issue has me wary and I am thinking I am hearing some distortion from it at times... I need to do some more test listening and see what is going on with it, but at least both channels are up.
     

     

    Please register to disable this ad.

  12. rjsalvi

    rjsalvi Active Member

    Messages:
    487
    Location:
    San Diego, CA
    Distortion can easily be switch contacts, dirty pots, bad solder joints, or a dicey transistor. You'd mentioned you'd replaced all the trans, so I'd be first inclined to check the solder joints/solder pads/traces on them. In the case of switches, sometimes the only way to remedy them is to take them apart and manually clean the contacts. Symptoms can range from intermittent distortions to full blown dropout. Pots are generally self-evident when turning them and are you sure it's the amp distorting and not something else?
     
  13. mikepick

    mikepick New Member

    Messages:
    29
    Since I've got it reassembled now, I think I'll start from the outside in, so pots and switches first. There is some static in the volume knob. Then on to inspecting joints and testing other components. Unfortunately the schematic is slim on voltage test values...
     
  14. rjsalvi

    rjsalvi Active Member

    Messages:
    487
    Location:
    San Diego, CA
    When you have a situation in which the SM/schematic has no values at strategic points, then compare the bad channel values against the good channel ones. If there's a glaring mismatch, you'll spot it.
     
  15. mikepick

    mikepick New Member

    Messages:
    29
    Good call. Gave the pots a clean, switches are harder to get at. No more static when turning the volume knob, but distortion is still there when listening. To clarify, the distortion I'm getting is a light fuzz, not especially harsh or crackling. It gets worse when the Loudness switch is engaged. Seems to be in the headphones as well as the speakers.

    I'm testing by subbing it into a working setup with a Marantz 1030 and PSB Alpha speakers, with the signal coming from an iMac, so it's definitely the amp that is fuzzy.

    One other oddity is that the speakers switch works for A or B, but not at the A+B position.

    I think I need to test the preamp some more as well. Before repairs I did some listening to the signal coming from the pre-amp to verify I had both channels coming through, but not for signal quality.
     
  16. gort69

    gort69 AK Subscriber Subscriber

    Messages:
    1,008
    Location:
    Vancouver, WA
    Some Kenwoods from that era (not all) wired the A+B switch in series, so unless you have 2 pairs of speakers connected it's an open circuit.
     

     

    Please register to disable this ad.

  17. rjsalvi

    rjsalvi Active Member

    Messages:
    487
    Location:
    San Diego, CA
  18. mikepick

    mikepick New Member

    Messages:
    29
    After some research I came across this thread in which the poster notes distortion and inability to set bias and finds that transistor Q17 was in backwards.

    I pored over the schematic and figured out that I had done the same thing with Q17 and Q18. It might be an error in the schematic / manual. I oriented the replacements based on a photo I took before removing the old ones, and used the manual notes to determine the pinout...

    Anyway, the good news is that the reversal worked, and I now have clean sound out of both channels, and the bias pots responded properly and are now set at ~40mv for each channel. The DC offset is ~50mV on the speaker posts, which can't be changed but seems reasonable. Thanks for the help all!

    IMG_0196.JPG
     
    rjsalvi likes this.
  19. rjsalvi

    rjsalvi Active Member

    Messages:
    487
    Location:
    San Diego, CA
    Errors in schematics are not uncommon. Nice work!
     
  20. quaddriver

    quaddriver 120 What's per channel Subscriber

    the 50mv offset is in the 'fix it, but it wont exactly blow up today' range....did you match up and any all pairs of inputs? and matchup I mean hfe 1-2 digits apart?
     

Share This Page