The Kenwood KA-3500 restoration

[curtpenn]

Bought one of these and its companion KT-5300 + rack handles around '77 - my first "real" stereo. I mail-ordered them because I was a poor college student in those days and that was the cheapest way to get them. I see you are in Dallas. Used to travel from Waco to Dallas on the weekends haunting all of the stereo shops around that time. Ended up with a pair of Cerwin Vegas from Sound Climax which was around Lover's Ln and Inwood. Well made, very efficient, and really suitable for the type of music I was mostly listening to in the '70s. Thanks for triggering a walk down memory lane.

 

[aatallah]

I know this is an oldish. thread, but thank you for putting this together. I own this amp, and it was an incredible guide to fix it up and get it back in top-shape.

One comment: I tried the capacitor substitutions you suggest, and it made the sound of my amplifier a bit muffled. I went back to the original values (new capacitors, with your suggested voltage upgrades), and it is now back to the original sound. Just want to let know any people who follow this guide. Maybe it's due to my transistors (kept all original), and I know in your case you changed some of them.

Still, incredibly detailed and useful tutorial. Thank you!!

 

[hopjohn]

Still, incredibly detailed and useful tutorial. Thank you!!

Thanks, I'm glad it helped. Cheers.

 

[Mayank]

hopjohn:

Reading your amazing thread inspired me to rebuild/resurrect a KA-3500 whose main PCB had been lying in my "can scavenge for useful stuff" pile for over 5+ years. The board was missing several transistors including, you guessed it, the unobtanium 2SD588/2SB618 pairs in both channels.

Prior to ordering replacement components suggested in your thread, I felt prudent to check if I could at least repopulate one channel using components from my parts bin and get it to work.
The bare bone KA-3500 is alive and working on the test bench



Happy to report that in addition to the NJW0281G/NJW0302G pair suggested by you as replacements for 2SD588/2SB618, one can also use MJL4281AG/MJL4302AG. These devices are TO-264 with insulated hole which makes fixing easier with no need to insulate the mounting bolt.

Also, one can replace Re15/Re16 with 150 ohms resistors to bring down DC offset to very respectable single digit levels.

Thank you for taking the time to document & share your efforts so methodically!

Rgds
Mayank

 

[Mayank]

Here's a big shout out of thanks to @hopjohn !!

I did replace the missing 2SD588/2SB618 output pairs with new Onsemi NJW0281G/NJW0302G. These are TO-3P devices with wider legs which needed (a) one hole drilled per device centered above the original bat-wing mounting holes and (b) delicately lengthening the PCB slot with a dremel for the wider transistor legs. Installation competed with SilPads.

The KA-3500 is alive. After about 3 hours of settling in the new capacitors, I am happy to report that it sounds very clean and has a much more pronounced bottom end as compared to the other KA-3500's with the NEC outputs.

Rgds
Mayank

 

[JRH99]

Great job on documenting your work on the KA3500! It has been very helpful to me as I plan my first full restoration. I have a KA3500 in near mint cosmetic condition that I plan to do a full recap on. It had a problem with distorted bass in both channels with volume above 3. I was able to diagnose and fix by replacing the power supply caps and its now back in perfect working order and sounds great. However, I am going to do a full restoration on it and have been working on the part list to try to put in one order for all the parts to save on shipping.

Here is where I could use some help if anyone can give some advice on what you would do: I have all the parts in a cart on the mouser web site and all are available except one… It is MKS2B051001N00JSSD Mfr.:WIMA Customer #:Cd23,2C24 Film Capacitors 10.0 uF 50 VDC 5%. So my options are to buy this part from digikey who has it in stock and pay the extra for shipping, or select an alternate part that mouser has in stock. Mouser has the MKS2B051001N00KO00 Mfr.:WIMA Customer #: Cd23,2C24 Description: Film Capacitors 10.0 uF 50 VDC 10%. Can I use the 10% vs 5% tolerance? Or should I just move to a standard 10uF electrolytic cap?

 

[rickl]

Can I use the 10% vs 5% tolerance? Or should I just move to a standard 10uF electrolytic cap?

i think the 10% WIMA will be fine. I used a Fine Gold electrolytic when I did mine.

This might work too (5%, 5mm lead spacing, 10uF, 50v, ...)

These are nice amps.

 

[JRH99]

[QUOTE="JRH99, post: 14227205, member: 307568"Can I use the 10% vs 5% tolerance? Or should I just move to a standard 10uF electrolytic cap?

i think the 10% WIMA will be fine. I used a Fine Gold electrolytic when I did mine.

This might work too (5%, 5mm lead spacing, 10uF, 50v, ...)

These are nice amps.[/QUOTE]
Thanks! Looking forward to restoring it.

 

[JRH99]

Working to restore my KA-3500 as my first restoration project. Replaced all the caps on the preamp board with many stops to test my work on my DBT and all is going good so far.


I need some help related to resistor replacements. @hopjohn recommended 10K > 10K 0.5W Yageo metal film Matched Signal Path for Rd 51 / 52. My question is; does the size (length / diameter) matter if values are correct? I did not consider the size when I ordered them and am now concerned that could be a problem. Did some searching on the topic but found no definitive answers. I found that Metal Film resistors can be smaller than Carbon with same value, but also found comments about surface area and heat dissipation. Can anyone give me some direction on this?

 

[Semanon]

Modern components are often smaller that vintage counterparts because of advances in design and manufacturing. If the resistor has the same values for resistance, heat dissipation, and any specialty features (I.e., flame proof), it’s a drop in replacement.

 

[JRH99]

Modern components are often smaller that vintage counterparts because of advances in design and manufacturing. If the resistor has the same values for resistance, heat dissipation, and any specialty features (I.e., flame proof), it’s a drop in replacement.

Thanks for the help to confirm that for me! I learned a lot about resistors looking for an answer to that question.

 

[Semanon]

I learned a lot about resistors reading the Pioneer ‘Tuning Fork’ tech papers. Very informative. They especially give context to what was available at the time and why decisions were made for one component over another. Resistors are in Volume 1, I think.

 

[JRH99]

Completed my KA-3500 restoration project. Big thanks to @hopjohn for the nice guide, that was a huge help in getting be through my first full restoration project. Unit is back up and running and sounds great.

I am attaching my BOM worksheet that shows the parts Hopjohn recommended and the ones I actually used. This includes the p/n and a link to them on Mouser. Hopefully this could help some future newbie tackle this restoration. https://docs.google.com/spreadsheets/d/1STxtiiORratVAdV_3YjveIc_sozHsMJco4q5JwZIv_4/edit?usp=sharing

Some notes and questions I had from doing this project, feel free to comment:

1. Although I based much of what I did on Hopjohns guide, I stuck to the capacitors that more closely matched the uF of the originals. I am sure Hopjohn had good reason for using the higher uF, but being new to restorations it was not clear to me what the benefits would be to the higher values. Also, I had already replaced the capacitors in the power supply with originals values months ago as they had failed and caused distortion. I was not sure if they all needed to be upgraded in the same way. In the end I decided I would be happy to have it sound as good as original and did not want to make too many changes.


2. For gain matching the input differential pairs Qe1/Qe3, this became a bit of a rabbit hole for me… I bought the parts and purchased a multifunction tester to gain match them. The more I looked in to this, the more opinions I found on how to do it correctly. I quickly found from my tester how temp sensitive they are. I tried to isolate them but results were inconsistent. I found that some build special testers boards just for the gain matching and found many differing opinions on those. Finally, I decided to move forward to replace all the other parts I have listed and that I would then test the DC Offset and decide from there. So, my results are: 9.6mV right channel and 11.2mV left channel. I decided not to touch it. Would love to hear any comments or suggestions on this.


3. Reapplying the thermal paste on the output transistors, this again turned in to a bit of a rabbit hole for me. I purchased the Wakefield 120 compound and was planning to do that last after everything was tested and working. Lots of opinions on if you should replace it or not, but it makes sense to me since you’re in there just do it. However, the more I looked in to this I became concerned that I might break the mica insulators trying to remove and clean them. I decided I would buy replacements just to be safe. Turns out insulators just like the output transistors themselves are unobtanium. So, I decided not to take that risk as all appears to be working as it should. I have a temperature probe on my DMM and have checked after running the am hard for hours and the transistors stay about the same temperature as the heat sink with in 1-2 degrees C. Again, would love to get input / feedback on this.


4. The Bias Trimmer was the last thing I replaced and I had a lot of concerns about setting them low and getting them in correctly so that low was not high if you know what I mean. I did all this work and did not want to blow up the amp now.. I’m sure this is just a newbie concern as it was not clear for me and I spent a lot of time to be sure I understood this before I installed them. You can see from all the pictures I had at the end of my BOM that I probably spent a little too much time on this. Maybe it will help someone else work it out faster first time around.

5. Question related to measuring Bias: Should bias be set when amp is cold? What is the proper procedure for measuring the Bias? I read that you should set it from a cold start, but does that mean within +/- 10min? Should both channels be tested from a new cold star? This is what I did: Cold start amp with DMM connected and let it run for 1 min for each channel. Is this the right way to do it? I see that it will continue to drift as the amp warms up, is this normal?

 

[Semanon]

I hope you can repost this to a new thread. It can get linked in the restoration sticky to help future restorers.

I probably spent a little too much time on this.

As a newbie, I don’t think there is such a thing unless you hated it, learned nothing, and the amp sounds worse than when you started. For me this is a hobby. I work on a project to reset my mind and do something for me. When I was young it was rebuilding cars. This is smaller, cleaner, and less expensive.

 

[JRH99]

5.) Bias should always be checked/set when warm. After the resto, set the bias a few millivolts low and let it idle for 15 minutes, then recheck and set it. After that, pump a signal through it and run it hard for a half hour and check/set it again.

Thanks for your comments! I will go back and check / set the bias with your process.

 

[nicholas2381]

I'm currently working on a KA-3500 and it's mostly done. But I have a couple questions:

1. The capacitors for the Phono/Tuner inputs (C3 and C4) are twisted together but not attached to ground - where should these be grounded? Looks like maybe somebody had them attached to the phono ground screw previously but they came loose.
2. The phono switch causes a pop in the speaker when engaged - louder if you push harder; softer is you push lighter. If the volume is down, no pop. I need to investigate further, but any starting points to consider?

 

[rickl]

1. The capacitors for the Phono/Tuner inputs (C3 and C4) are twisted together but not attached to ground - where should these be grounded?

When I look at the schematic, it looks like ground to me. When you look at the schematic, what do you see?

I haven't worked on a 3500 in a while but do have a 5700. I checked the ka-5700 schematic and it has the similar C4. My guess this is goes to a chassis ground not the phono ground.

2. The phono switch causes a pop in the speaker when engaged - louder if you push harder; softer is you push lighter. If the volume is down, no pop. I need to investigate further, but any starting points to consider?

a pop might mean voltages are in the input to the selector switch. I'd see if you are getting any DC on the output of the phono switch. Leaky transistor or cap might cause this. check around Rd23/Rd24. did you replace Cd11/Cd12?

 

[nicholas2381]

Thanks, rickl - yes, the schematic definitely shows ground. I'll see about getting them connected to chassis ground.

And yes, I recapped everything, but I'll double check those and the resistors mentioned. Appreciate the input!

 

[rickl]

check ... the resistors mentioned.

I'd measure the voltage at those resistors to ground to see if there is any DC. If so, you start backing up to find the leaky part.

I had a Toshiba receiver that did the same thing. pop'd when selecting the phono. I had to replace the blown opamp in the phono section. it was dumping 14v to the selector switch. I know the Kenwood doesn't have opamps in the phono but the idea is the same.

 

[nicholas2381]

I'd measure the voltage at those resistors to ground to see if there is any DC. If so, you start backing up to find the leaky part.

I had a Toshiba receiver that did the same thing. pop'd when selecting the phono. I had to replace the blown opamp in the phono section. it was dumping 14v to the selector switch. I know the Kenwood doesn't have opamps in the phono but the idea is the same.

A quick poke around the board turned up no DC voltage to be found. I'll do a more thorough check when I can.