AU-999 Restoration and Mods

Sp33ls

AK Subscriber
Subscriber
Continuing from my previous thread, I moved on to beginning the recap and mods. Everyone has been very helpful, even before grabbing the solder. I much appreciate that.

I managed to get the main filter caps (went with 6800uF), head amp, tone blocks (with kev's preamp mod) accomplished this weekend.

I had a few of questions tho.

1. Do all the later revisions have this small, blue 1(M) 25V electrolytic cap in place of the mylars? See pic. I initially followed the service manual and swapped them for poly 0.47uF, but after observing #4 (below) I swapped those little guys back in. A 1uF film cap should be a suitable replacement, right?

2. After plugging everything in to ensure things were still working, I noticed one channel was dead. I looked over my work and realized that one of the KSA992's had a little tiny piece of solder bridging the leads. I removed it, checked it in diode mode (appeared to read fine, wasn't shorted), and now both channels work. Do you think I need to check anything else to ensure everything is alright? Any voltage checks?

3. I set the DC offset and went to try the bias, following the manual, and I think I turned the trim the wrong direction. One heatsink started to warm up pretty quick. I didn't see any smoke, and it still seems to work just fine, so that's good. But, when I went to try and read the voltage across the resistors on the filter boards (to set it at 13mV), I'm getting 0V at that location on both sides -- regardless of the bias trim.

4.The phono section is very very quiet. Max volume is silent. The only way to hear anything is to use a line-level device (I had it's output set to very low). And of course, if I turned it up a bit on the line level, it sounded both quiet and distorted. I'm guessing due to clipping. But, why is my head amp extremely quiet..? :(

I'm glad I decided to test things half way through.. makes troubleshooting a bit easier. The line level sections do sound pretty good, tho. Hard to tell in the limited time I had it on if there's any night and day improvements... The amp is probably out of bias and there could be something else going on with it, but the fact that it sounds as great as it does in this state is very promising.

Looking forward to any help!!
 

Attachments

  • MVIMG_20180714_195923.jpg
    MVIMG_20180714_195923.jpg
    38.5 KB · Views: 31
  • IMG_20180714_195907.jpg
    IMG_20180714_195907.jpg
    48.9 KB · Views: 30
  • IMG_20180714_190219.jpg
    IMG_20180714_190219.jpg
    53.9 KB · Views: 31
Regarding point #2, it's a good idea to keep a strict eye on each replacement - that is to take a picture of the traces before you begin swapping, that way you have a reference of where the traces can touch and where not. You can also have a print out or use a laptop or iPad screen with a picture of the print taken from the service manual.

As for point #3 - two things. The emitter resistor measurement can be tricky to get when the resistor legs are sleeved, you might want to use a blade of sorts to remove a few mm of sleeving. Alternatively just measure mA across the fuse holders per manual.

It's easy to turn bias pots the wrong way, or if replacing, starting up to high. In your case measure ohm on the "cool" bias pot by probing one of the outer legs and then center leg. That reading you dial in to the "hot" pot by measuringi on the same two legs - before power on. This should give you a working start up value. (Always do that if replacing pots - get the "working" values from the old pots, then dial in the new ones before start up.
 
3. I set the DC offset and went to try the bias, following the manual, and I think I turned the trim the wrong direction. One heatsink started to warm up pretty quick. I didn't see any smoke, and it still seems to work just fine, so that's good. But, when I went to try and read the voltage across the resistors on the filter boards (to set it at 13mV), I'm getting 0V at that location on both sides -- regardless of the bias trim.
Where are you measuring this? Across one of the emitter resistors?
Can you show us?
Also from memory, I think you wind that bias trimmer around quite a ways before anything happens, and then it all happens very quickly.
Always start with the timmers off, counter clockwise in this case. I never try to presume to "preset" them from where they were before, with new caps and transistors through the amp, it wont be the same anyway...
 
:crazy:...Ok if you are going to use the emitter resistor method, it's a good idea to just check the value of the white ceramic resistors, that they are pretty close to the .5 ohm. Here the 13mV is a calculated value, but in reality they can vary from .4 to. 6. since they are +/- 20%. I know Kev will say Iam overthinking this :crazy:

Well, take the trim pot pre-setting as a suggestion. If not for bias then for DC, where you don't want the pot all the way to one side or you'd instantly trigger protection. The other benefit, and especially if you are replacing both pots, is to measure the outside legs for the total value, to make sure you get the right pot in the right slot. (Someone I ..ahem...know messed up and bias and DC was out of control :dunno:)
 
:crazy:...Ok if you are going to use the emitter resistor method, it's a good idea to just check the value of the white ceramic resistors, that they are pretty close to the .5 ohm. Here the 13mV is a calculated value, but in reality they can vary from .4 to. 6. since they are +/- 20%. I know Kev will say Iam overthinking this :crazy:

Well, take the trim pot pre-setting as a suggestion. If not for bias then for DC, where you don't want the pot all the way to one side or you'd instantly trigger protection. The other benefit, and especially if you are replacing both pots, is to measure the outside legs for the total value, to make sure you get the right pot in the right slot. (Someone I ..ahem...know messed up and bias and DC was out of control :dunno:)

Oh, not at all Tom, you've got about 5mV to play with to stay in the sweet spot I believe.

always set the DC offset trimmer to 50%, this will be very close, especially if you matched the differential pair transistors
 
I got it figured out. Turns out the head amp block had the KSC1845FT's backwards :/ After flipping them around, the phono seems to work just great!

I gotta say, after recapping and swapping out the transistors, it's amazing how you can hear just how much more quiet the amp is when you crank the volume pot all the way up.

The amp has really come to life too. It's too early for me to really leave any real review, but so far so good. Also, I haven't completed the bass eq mod, trnsfrmr low-end mod, and Tom's driverboard mods yet. The transistors should be within 1%, though, just need to get the resistors in there, and order a few more C0G. I'm also going to swap out the 100k Ohm and up resistors as well just for that slight noise improvement.

Looking forward to hearing how it sounds after the bass mods!

It sounds pretty sweet as a pre-amp for my 300B SET, too.
 
That will do it, don't put those transistors in backwards!!

Glad your enjoying it...
 
Me too - that phono board is tricky as you have two different types of signal transistors with different pinouts + they are in some weird mirrorwise config.

It's easy to get "mod overwhelmed" with the 999. I'd start with kevs bass EQ mod...and then see If you get the lowend you seek before tagging on electrolytics on all your mylar coupling caps - in my rig it's more than plenty. The driver board mods will add resolution and definition from top to bottom. There is even another mod deviced by Smurfer77 in which you change the overall sound character by tweaking the value of the feedback resistor. He lap tested and graphed out 3 "values" -- lowest value gets you the 777a sound of low damping factor and looser/fatter bass. The middle value is the stock 999 and the highest is the stock value used in the near identical 888 - yielding an overall tighter sound.
 
Thanks guys! Yeah, actually most of the deep bass will be coming from my subwoofer in my setup, as I'll have smaller speakers that won't dig down deep anyhow.
From what I've read, it sounds like some of the full-range drivers (like Zu's) prefer to have a little less damping factor.

That being said, I am a sucker for a "full-bodied" signature. Not to be too poetic, but I like impact, presence, and still retain a great amount of texture. I love picking out little details in songs that I've never noticed before. In fact I've noticed it already with the AU-999. I'm very impressed with its detail! However, I don't like overly bright systems, and don't like to sacrifice that body for just a slight bit more of detail.

My first impressions of the AU-999 are that it's higher detailed than my Marantz 2220b, and a bit more tight. That's not to say it's in the bright camp. I think it seems to fall right in the sweet spot between smoothness and detail. I'm curious if others would agree with this observation?

I'm also wondering if anyone who has performed a complete system overhaul has noticed a "break-in" period, and if so, what sonic changes did you observe?

If I do end up primarily using the AU-999 as a pre-amp, how similar are the pre-amp sections (and thus the pre-amp tone) of all the AU-xxx's?

I've always heard others mention that most of an integrated amp's signature resides in the pre-amp section. In this case, though, it seems that the topology of the 777 & 999 are quite different in the amp section. But, what about the pre-amp section?
 
Last edited:
Back
Top Bottom