SA-9800 Pwr Supply Assy problem

[vivace.ken]

Hi All...
My Sa-9800 died a couple weeks ago.. I've been meaning to take a look at it but life intervened... I got inspired to look at it after reading the recent threads discussing the power supply and the display brightness.. Thanks for the motivation..

I measured the voltages on the AWR-194 board and the values are way off from where they should be.. Here's what I measured.
pin 12: 75.5v, pin 13: 75v, pin 18 -.34v, pin 16 -.01v, pin 14: 0v, pin 17: -.15v, pin 15: -76v, pin 9: -76v, pins 1&2: -5.6, pins 10&11: -1v, pins 5&6&7 -3.8v, pins 3&4 4.6v

Help please... Any ideas on what's going on or where to look next ?

Thanks in advance..

 

[markthefixer]

Here's what you measured:

pin 12: 75.5v, ok
pin 13: 75v, ok
pin 18 -.34v, this is an AC voltage
pin 16 -.01v, this is an AC voltage
pin 14: 0v,
pin 17: -.15v, this is an AC voltage
pin 15: -76v, ok
pin 9: -76v, ok
pins 1&2: -5.6, should be -52v
pins 10&11: -1v, should be -32v
pins 5&6&7 -3.8v, should be +32v
pins 3&4 4.6v should be +52v

The negative voltages are a mirror image of the positive regulator, so everything keys off the +52v regulator. You have some transistors and a zener diode to replace - BUT the Q5 2sk34 is no longer made, nor are any replacements. There are a few members that have bought remaining replacements (20 for $20 from ampslab.com) that may make a deal, or you can try out one of their other 50 volt jfets (at 20 for $20). mouser.com parts numbers:

Q1 2sc1735 npn to-92 BCE 100v 0.5a 0.8w 50-300hfe 130mhz
512-KSC2690AYS to-126 ecb 120/a160 1.2a 20w 155mhz 35-320hfe $0.40

Q3 2sa733 pnp to-92 ecb 50v 0.1a 0.25w 90-600hfe 180mhz
512-KSB1116AGBU to-92 ecb 60v 1a 0.75w 160mhz 135-400hfe $0.09 (Y,G=200/400)

Q5 2sk34 to-92 sgd 50v jfet

Q7 2sc1914a npn to-92 bce 90v 50mA 0.2w 250-1200 hfe 150mhz
512-KSC2383YBU to-92L ecb 160v 1a .9w 50mhz 160-320hfe $0.17

D5 xz-060 6v 0.5w zener diode

 

[vivace.ken]

Thanks Mark..

I really appreciate the help..
My project here has taken an interesting twist.. After tracing through the circuits for a couple hours, I decided to take a break and left with the unit unintentionally powered up.. I came back a couple hours later and found everything working just fine.. The display was lit brilliant blue, the voltages on the AWR board were all fine, and the pre-amp/amp all worked great..

This lead me to think the problem was somehow thermally related. I unplugged the unit and let it sit for a couple hours. Turned it on and Bingo it worked fine.. Next I let it sit unplugged overnight.. This morning when I turned it on all was working well ??

Now I'm thinking it must be something mechanical ? Are there any switches, pots, relays, or other mechanical devices that could cause this sort of behavior and the power supply to act the way it did ? If so what is the recommended way to clean them. Could there be a loose solder joint somewhere causing this ?

Also is it possible that the load to the power supply (aka the amp circuitry) was somehow disabled causing the power supply to loose regulation since there wasnt enough current draw ?

Now that its working, I'm not quite sure what to do ? Any and all advice is most appreciated !!

Thanks, Ken

 

[restorer-john]

Ken, the AWR914 PSU board is notorious for dry joints in and around the hot regulator transistors and all those TO-92s- due to the poor air circulation. The UK and AUS models have They are wire wrapped down and it is a major committment to remove it unless you have a wire wrap gun and even then, no fun. You can unsolder/cut the side panel side and hinge the board up, but the force you will put on the pins and thick wire on the other side isn't worth the dry joints you may create. When I last did one, I took every wire off, took the board out and essentially removed, cleaned and rebuilt the entire regulator section. Because Pioneer hangs everything off one zener I'd replace that one as a matter of course. The FETs in my experience are ok, but dodgy joints around them don't help.
You've got the AWR 194 board which has a second set of regulator transistors so make sure you check all joints on the 4 heatsinked T)220s

 

[vivace.ken]

Thanks John..

Well it looks like refurbing the PSU is the path I will be going down since my SA9800 has quit and come back three times in the past day...

Its been a while since I've done much wire wrapping.. The pins are certainly larger than I'm used to.. Do you have any info on a tool to use for these pins ?

Also while I'm at it I guess I should replace the caps on the PSU as well.. Would someone have a list of of the mouser.com part numbers for the caps ??

Thanks much all.. I really appreciate the help..

Ken

 

[markthefixer]

The wire wraps should not be disturbed. This is experience talking. The solid wire gets an intentional nick on each and every corner of every turn, and it WILL break off short either on removal or attempted installation. There is NO slack wire inside of these units. It will be anything from a royal pita to an absolute disaster.

It may seem like a catch 22, but to get access at times I unsolder the pin from the board, leaving the wrap undisturbed.

It is FAR better to disassemble the frame around the boards to get access. I have done that uncounted times.

If I could do an entire sa-9500 (with all it's shields and etc) without disconnecting wraps, the sa-9800 is far more wide open, there IS a graceful way to be found..

 

[Backlight]

Well it looks like refurbing the PSU is the path I will be going down since my SA9800 has quit and come back three times in the past day...

I suggest you also replace all electrolyte capacitors while having the board out or flipped to re-solder all joints. I have an order list for you (all but one Nichicon from Mouser.com) if you want.

One question: do you have four transistors with heat sinks? That would mean you have the newer board version. I have only two transistors while the PCB is designed to have four installed. The reason Pioneer installed 2 extra transistors is to reduce the load on each of them. The two-transistor design caused a lot of failing transistors due to over heating.

And a request: When you are successfully done with the PSU, could you please make some voltage measurements on your indicator board for me please? I am troubleshooting my display and I need some readings from a normal and correct working display.

Good luck!

 

[vivace.ken]

I've successfully flipped the board over by only disconnecting a couple wires.. I think I should to be able to do everything I need to do without totally removing the board... so far so good.

If I could get a list of the Mouser p/n's to replace the electrolytic capacitors that would be great !! The PSU board only has 2 transistors with heat sinks. The silkscreen says AWR-191A. Is it possible to add the two additional transistors ? You are right, the board seems to be laid out for the second pair..

Once I got the board out it was clear to see the heat damage. The board is burned black where the heatsink is mounted to the board. Additionally the solder side is covered in a resin. Is this flux that got "cooked" out of the solder ?

If nothing else, it might make sense to get larger heat sinks for the T0220s..

I would be more than happy to measure some indicator board voltages for you.. Too bad, I could have gotten them yesterday, during one of the brief working moments

Thanks all..
Ken

 

[markthefixer]

Make the list, the format has been discussed recently in threads:

example:
awg-018 c1 1uf 25v cssa
awg-018 c2 1uf 25v cssa
awg-018 c7 47uf 16v cea
awg-018 c8 47uf 16v cea
awg-018 c17 10uf 25v csza
awg-018 c18 10uf 25v csza

and I'll add the parts numbers.

 

[Backlight]

I will send you my order list tonight. Check the values of all caps physically as well, not just the values from the Service Manual.

I don't know if it is easy to add the two transistors. I shouldn't be I think. But we would need to have two additional heat sinks too. I have not (yet) come across a SA-9800 PSU schematic that includes the two additional transistors, so we will have to figure out ourselves how it should be modified.

 

[vivace.ken]

Hi Mark..
Attached is a list of the electrolytic (and other) caps on the AWR-191 board. I verified the info in the service manual with what's on the board..

Thanks for the help crossing them into Mouser p/n's..

Also if you have p/n's for the two power transistors I would appreciate it.. I should probably replace them while I am at it..

They are:

Q11: 2SD712 (2SD313P)

Q12: 2SB682 (2SB507P)

Thanks ..
Ken

 

[Backlight]

Hi ken,

That corresponds (logically) to my list. No need to replace the non-electrolytic caps. Mark will probably help you with a nice part number list. You will end up with Nichicon UPW caps for all but C2 and C3. These two are axial types and I chose for Vishay replacements, part number 594-2222-118-19221. By the way, you won't be finding 220uF/80V, so replace them with 100V, as these vishay's are.

The vishay's are extremely long life, a bit thick, but will fit comfortably.

You can also look them up yourself by searching on Mouser choosing Nichicon as manufacturer and filter for "low impedance" types. This will result in UPW series caps which are excellent quality for this hot environment. (I have this from Mark )

If you are not so experienced in soldering, be aware to de-solder properly with a solder suction pump or a de-solder iron and remove ALL solder before pulling out the cap. It's easy to damage the copper lanes on the PCB when you need too much force to get the caps out.

It's best to glue C2 and C3 to the board (as the originals are) for stability. I use RTV162 silicon glue for that, also available through Mouser. Clean the PCB with alcohol before you apply the glue for long lasting "stickyness".

Keep us updated with your progress!

 

[markthefixer]

I have a rally to prepare for and help at tonight, so that list will be worked on tomorrow.

Cap rules of thumb (nichicon)

I use upw, uhe for general caps including signal coupling in that order depending upon stock "cea"
I use ues, uep & uvp in that order for bipolar depending on stock
I use ukl for low leakage caps (in sensitive areas they reduce the noise floor)"ceanl", "cssa", "csza"
0.1uf to 2 uf I use stacked film where they fit: DFS and ECQ

solid aluminum and tantalum caps get stacked film when possible, or ukl's otherwise. "cssa" "csza"

c?B means it's an axial cap = one lead on each end
c?a = radial = both leads on one end


searching at mouser:

easy..... need a 100uf cap? search on: 100uf UPW

then find your voltage and if stocked / in stock and possibly a size choice. go UP in voltage if your voltage is not available.

voila....

sub uhe, ukl, dfs, ecq etc for upw for the different type...

 

[vivace.ken]

Working with the Cap rules of thumb, I was able to put together an order for the caps with Mouser. Had to bump the voltage up on a couple devices so they make it here before summer... but I think it should be ok..

Are there similar rules for figuring out semiconductor substitutions ? How could I go about figuring out suitable replacements ?

Thanks agn.. I appreciate your patience and helping lead me through this..

Ken

 

[vivace.ken]

Hello again...

I was reading on another thread that the power devices (2SD712 AND 2SB682) are particularly problematic due to the heat. Someone on the thread suggested replacing them with a 2SD525 and a 2SB595 since they "run cooler".

Any thought on this ? Is it worth considering ?

Thanks, Ken

 

[markthefixer]

Hello again...

I was reading on another thread that the power devices (2SD712 AND 2SB682) are particularly problematic due to the heat. Someone on the thread suggested replacing them with a 2SD525 and a 2SB595 since they "run cooler".

Any thought on this ? Is it worth considering ?

Thanks, Ken

please quote that post so I know whom to watch out for....

NO, they will not run cooler. They will run the SAME temperature. I haven't looked them up to see if they are even remotely compatible voltage, current and gain wise...

Heat IS a problem. To cool them, you need to add more surface area to their heat sinks, to allow more effective transfer of heat to the ambient air.

 

[Backlight]

Hi ken, any successes on your SA-9800 worth mentioning here?

If you need a hand just let "us" know...

 

[vivace.ken]

Yes, just today I was able to finish reworking the board. I replaced the caps and the power transistors; adding larger heatsinks.
Powered it up and viola... everything seems to be working just fine..
Still needs to burn in a bit to make sure since the previous problem was intermittent... but so far so good.
Thanks to everyone for the help... I'm a happy camper..

 

[vivace.ken]

Also I forgot to mention. I was able to find some 2sk34 to-92 sgd 50v jfets in case anyone is interested..
Sorry for the double post...
Ken

 

[thartsell]

I'd be interested in SA-9800 parts...

You mentioned in 2009 that you might have 2sk34 to-92 sgd 50v jfets in case anyone is interested.. If you've still got one or a couple I'm interested...

Just found you as someone gave me a SA9800 and I started searching for info. My voltage is 70v instead of 58v ... thought I'd replace all the transistors and FETS if I could. Appreciate all the threads... very informative. I did download the service manual...

I think I also saw an alternative circuit in my multiple thread search but I lost the link... The zener is hard to find also right?

Thanks in advance...