The KA-9100 restoration.

With all the replacement parts now on hand, a new to me Fluke 87V multimeter, and some Deoxit Gold G100L for the switches it's time to move forward.

The front end is disassembled leaving all its guts spewing out. Initially I put bubble wrap around the power meters to protect them from damage, but after moving the amp few times I felt the best way to proceed was to remove them altogether. I did this by desoldering the wire wrap pins that hold them to the small meter pc board.

8yt825N.jpg



The first board to receive work is the control board X11-1410-00 A/4 or as I refer to it the "top control board" since the second control board is sandwiched just beneath it. This tightly packed board houses twenty-six electrolytic caps. Photos of spare boards owned by gort69 and other stock boards found on the WWW prove that the type of some of the capacitors used on this board were not consistently the same. Kenwood seemed to play fast and loose with whether or not to use low leak or general purpose caps in positions Ci9, Ci10, Ci33, Ci34, Ci39 and Ci40. Ci9 and Ci10 are at the input pairing and should always have been orange 47uf 16v, but because Ci9 was such a tight fit against one of the switch bodies my guess is they may have eventually moved away from using them and instead opted for what my board had installed, a smaller bodied general purpose 47uf 10v (shown below just underneath the furthest right switch).

g21bGEw.jpg


A depopulation of the board is shown below. I've removed the four mylar caps at Ci35-38 and replaced them with smaller Panasonic ECQ equivalents in order to make room for the Wima MKS2 1uf 63v caps at Ci31 & Ci32.

oMwbn7b.jpg
 
Work continues on the control board.

The 1/4 watt, 5% tolerance carbon film resistors are being replaced with 1/2W, 1% metal film. It's not something I usually do, but I've got an excess of resistors on hand and nothing better to do with them. In addition, if any amp deserves the extra attention to detail the KA-9100 does.

FOqdV3S.jpg
 
Nice work John. From your previous post, wondering about the increase from 1uf 50V to 1uf 63V for Ci31,32?
 
Nice work John. From your previous post, wondering about the increase from 1uf 50V to 1uf 63V for Ci31,32?

I'd prefer to use a 50v, they are smaller. At the time I needed to re-order 1uf caps Mouser was limited on their 50v caps availability and charging more than double what the 63v was going for. Just a case of bad timing. If you're following along and have the 50v, use those.
 
Here's a follow-up on the capacitor type discrepancies on the control board I mentioned earlier in post #22 now with some photo evidence. (Thanks to gort69 for the photo he provided)

"Photos of spare boards owned by gort69 and other stock boards found on the WWW prove that the type of some of the capacitors used on this board were not consistently the same. Kenwood seemed to play fast and loose with whether or not to use low leak or general purpose caps in positions Ci9, Ci10, Ci33, Ci34, Ci39 and Ci40. Ci9 and Ci10 are at the input pairing and should always have been orange 47uf 16v, but because Ci9 was such a tight fit against one of the switch bodies my guess is they may have eventually moved away from using them and instead opted for what my board had installed, a smaller bodied general purpose 47uf 10v."

Two control boards with varying capacitor types. The top has low leak, and the bottom general purpose.
Ci9, and Ci10 are my main gripe which are just above and below the the furthest left switch in the photos below.
ReYkjpc.jpg


A board photo taken from ebay (thanks to the original photo taker) that shows low leak caps being used.
biQIVIi.jpg
 
I don't think I'd lose any sleep over your cap choices.
I point it out more as a fascination with the historical process of creating these wonderful things than a concern of what replacements to use. I'd love to be able to talk to someone who was designing these amps and involved in the production process to ascertain why they did some of the things they did.

Concerning the ECQ-P ...I wish I had stocked up on the 50v varieties before they disappeared, I only have just a few values.
 
Progress report.

The installation of the metal film resistors is going nicely. There's around ninety of them on this board alone. The treble and bass potentiometers were removed from the board for better access to the many resistors underneath. The wire dress of the jumpers is pretty sloppy in stock form so I've tidied up most of it.

OJghAGF.jpg


The rear of one of the two pots (they are identical). 20Kohm B taper x2
skgMHEU.jpg

 
JOHN:
So... to put this project in prospective, how many hours have you spent on this so far? Curious
because I did only one like that a few hours a day so it took a long time as far as rebuilds go.
This is the worst part of the unit to do and seems you are tackling it quite well John.

Regards
John M
 
So... to put this project in prospective, how many hours have you spent on this so far?
Good question.

I've probably got a solid 7-8 hours into this board alone. I've been going in small stints so it's a little hard to say precisely. The main thing with the resistors is that I've been pulling one at a time so as not to screw up and that makes it' a timely procedure. I didn't want to check every installed resistor against the schematic. If I'd done a mass depopulation it'd gone faster, but I probably would have made a mistake along the way because it gets pretty monotonous after a few hours and I tend to lose a little focus.

Here's how I'm going about it.

I used the schematic and made a resistor list starting with the value in ascending order followed by the part id #. I worked across the board left to right using a soldapullt to remove the resistor and some wick if I needed it. Once pulled, I read the color code and I check the value on the multimeter, then check that value against the list, then pull the replacement from the big box o' resistors, check the value of the new part on the multiimeter as well as read the color code. When I was happy everything matched, I formed the leads, installed the insulator if need be, cleaned the pads, installed it, then marked it off the list. Rinse, repeat X90. Quite the procedure, but it keeps you from screwing up.
 
Last edited:
The top control board is complete with the exception of Ci9/10, which gort69 is helping provide since they were errantly left off my Mouser order. Thanks buddy, I owe you one. I upped some of the voltages of the 10V and 16V caps as well increasing the local filter values of Ci41/42 from 220uf to 330uf. I had originally planned to replace the four 220uf 10v caps with Nichicon HE 220uf 35V, but they measured low in capacitance. I ended up going with the 220uf 25V Nichicon FG caps which measured right at nominal. The remainder of the jumpers were made to look nice. Once giving them a thorough Fader Lube and DeOxit Gold bath, the treble and bass tone pots were carefully reinstalled. All of the other switches were cleaned thoroughly as well with a treament of CRC contact cleaner followed by D5 and finally some needle applied DeOxit Gold.

jBeHLYS.jpg


Work on the bottom control board begins by removing all the electrolytic caps while checking the cap positions against the silkscreened polarity markings. Access to this board is made easier by desoldering the wire wrapped pins numbers 6 through 13. Pin 12 remains because the ground wire here had already broken free from the chassis during disassembly. This board, I believe, has only 36 resistors so it should go much quicker than the first.

nPQGYCm.jpg
 
I've just about completed all the resistors on the second control board. Desoldering those pins I mentioned last post really makes this board so much easier to work on. There's been no problems at all. The straight lead component insertions enables you to use a Soldapullt with great success. I've only been forced to use wick on the occasion when a solder pad has been on a large ground portion of the foil pattern.

rrfD4WF.jpg
 
That looks fabulous. Do you use one of those lead bender contraptions or do you have a bionic eye? - those things are perfectly centered.
 
That looks fabulous. Do you use one of those lead bender contraptions or do you have a bionic eye? - those things are perfectly centered.

Thanks. I have fond memories of Lee Majors and even fonder of Lindsay Wagner.

I have a set of those plastic, red, triangular shaped lead benders I think you're speaking about. For small components I find them useless as I can never get them to work as well as I'd like....too much slop. What I use are bent nose pliers. Just some cheapies I got from Harbor Freight many moons ago. They were part of a set. Looks like they still have a set or you can get these individually. Anyway, I just eyeball them. Once you get a handle on the lead spacing just memorize where the bend needs to be. No bionics required, though I will make the sound effect while I bend these from now on thanks to you.
 
Back at it again after taking care of some priorities. The remaining metal film resistors were installed as well as all of the new electrolytics.

The 10uf 25V back to back capacitor pairs at Ci119/121 and Ci120/122 were each replaced with a single 4.7uf Wima stacked poly and a jumper. The small ceramic 100pf caps Ci101/102 were traded for C0G type MLCC ceramics of the same value. The .033uf Mylar caps at Ci115/116 were upgraded to Wima polypropylene. Ci103/104 47uf 16V got low leak 47uf 35V Nichicon KL. The four 100uf 35V local filter caps at Ci123-126 were bumped to 220uf 35V Nichicon PM. Ci107/108 and Ci117/118 220uf 10V were replaced with 220uf 25V Nichicon FG.

XPIKrd2.jpg
 
Last edited:
Just curious, are you changing the transistors?

Elsewhere absolutely, but not on these two boards, EQ and Tone amp.
Here are the transistors on these boards:
(10) 2SK68 JFETs
(2) 2SC1328 (2SC1345 in others)
(4) 2SA847
(4) 2SA640
(2) 2SA850 (2SA777 in others)
(2) 2SC1904
(2) 2SA899

I could maybe see replacing the 640 differentials with KSA992 or ZTX795A, but meh. 640s are in nearly every Kenwood I work on, and they don't cause trouble. Everything else I'd prefer to leave alone.

I may however replace the diodes Di1-4 1S0276 with 1N4148, haven't decided yet.
 
Back
Top Bottom