help me with a kenwood part number decode...

quaddriver

120 What?'s per channel
Subscriber
the part is C53 (.47uF-50V polarized) found between the collector and base of Q23 - part of the ASO in a KR-4070

(SM here on the AK)

I pulled a little aluminum electrolyte (this is on the 4070 in another thread that got seriously shorted out - it has all new silicon, caps and resistors) I replaced with the same, but 5-10 minutes into setting bias, a hum showed up and it shut down both channels.

trouble shooting found that .01 of a volt at a time the voltage for the differential inputs dropped and dropped (14v spec due to D1 Zener) and about 5.2v is goes fuzzy and by 5.1v its dark.

further trouble shooting, found that the base of Q23 was +rail which would turn it on, but the collector of Q24 was negative. This cap apparently went direct shorted, and its brand spanking new. granted, stuff can break, but before I put another one in (with ASO disabled it has been running outputs disabled, on drivers for hours to proof the system) I want to check the part number. consulting page 14 of the hymnal (aka SM) we find that C53 is kenwood part CE04AW1HR47M.

In the tuner for example, there are many of the same, but the part number (page 13 of same hymnal) is CE04AW1HR47CC.

I see that in the SM, the M suffix tends to mean 'mylar' but alas C31,32 are 1Uf 50V polarized and are C04AW1H1R0MCC. WTH?

am I dealing with typos or or there something special about this part?

also, WTH does C53 even do? In a similar location, on the screen print, is C52, which is a) not present and b) not in the parts list.

so obviously there are multiple revs in here (like R31-34 are used 220ohm, 5%, 1/4W) but I challenge you to find them in the SM or parts list. I wanted to know ifthey could be regular or fireproof, not finding any info, I erred on the side of safety...

thanks in advance....
 
There's nothing special about the part. If it was orange originally replace it with a polyester film cap or a Nichicon KL. If originally black, or gray, or maybe blue use a GP cap. If it's been inserted with the correct polarity and damaged you've got issues elsewhere.

A photo would help identify whether the resistors were something other than basic carbon or metal films.
 
small update, I soldered in another one, same maker, same mouser order and it worked. the addition of this cap when working, holds the 'its all ok' voltage on the base of Q23 to about .515v (starts out at about .540 and falls steadily) - damn close to firing the switch! the original was black btw. I have the value or close to it in non polarized film of some sort for my tube stuff, but it would never fit without jumpers, this particular part is squoze in.

let me detail how it appeared to a human as it failed: I had it playing on headphones and was setting the idles, 15 minute burn in when all of a sudden I started hearing a hum about 120hz and it got louder but never loud on the left, then the left got distorted and faded out, then the right followed. after pulling bias and emitter resistors to isolate the outputs I probed and tested and as I worked, off, on, off on etc, it came along faster and faster. I studied the SCM and figured it had to be Q23 slowly applying the brakes, but the ASO aint set up to do that, thinking Q23 went bad I pulled it and of course it came back to life. thats when I discovered the cap had 39v across it and it was a 50v cap...Im ready to toss this up to manufacturing defect. one that causes other stuff to go out and throws you off the trail.

and in debugging I found something out, due to backfeed, unintended consequences whatever, if the ASO does trip...it works backwards a 1.something, nil current voltage on Q19/20 (over current detectors) - it almost holds itself in shutoff until you power down and remove stuff. every ASO I have ever worked on have been bad when they fire, either they got raced to destruction, or Q23 aint big enuf to sink all the current we are asking of it?

at any rate, she works, all hooked up, setting idles/burn in for the rest of the day.

Im pretty sure the pulled resistors were carbons, and a clue that I would have been ok not flameproof is that kenwood puts all the danger resistors in the air, not on the board. these were tight to the phenol. I think all my stackpoles (flame proofs) are carbon or metal film...need to read the packages.

also of note for posterity, not knowing WHERE in the rev stream this unit was, R89 (SM, SCM, parts list) is listed as flame proof carbon, 56 ohm, 5%, 1/4 watt yet I found 33ohm in the same configuration. I used the old 'put in what you found' method given that this ran for a number of decades in the harshest environment short of shooting it into space with Musks tesla...

one last thing for posterity...leading up to the voltage splits for the input pairs, R88 (4.7Kohm, 1/2 watt 5% carbon) is on the rail and feeds a split to each input pair of 10K. so basically it burns roughly 47% of the 39volts at the break point. without D1 (14v, 1/2w zener) in place, the voltage is 20-ish.

The 4070 I developed the sub list for, had no D1 in it, it was on the screen print, but there was no evidence it ever was in there. its in the SM, parts list and SCM. so there seems to be 2 schools of thought...older models are ok with 20 volts for the input pair, new ones (with other added fun) want it to be 14. I wish there was a consensus on which way to rebuild em...but unless I suck at searching, the only 4070's in blowed up state have been ones I got - they just dont seem to fail. either that or people tossed them...
 
Back
Top Bottom