• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Help With Pioneer Sa-9800 Please

H

hifileslie

Member
With no sources connected I have a fairly loud hum on both channels which does not change with volume setting. Occasionally there is also a hiccup/thump on both channels whether source connected or not. Apart from this the amp sounds fantastic and plays perfectly. The sound recovers immediatly after the interruption. With the pre/power amps links removed there is no thump and very little hum.

Grateful for any advice please.
 
Register to hide this ad
I'm willing to help, do you have a DMM and soldering equipment?

Are you handy?

If so download the sa-9800 hi res service manual from the database, so we can both be on the same page as the power supplies are checked.

Also, most importantly, the sa-9800 has to be cleaned/deoxited (it can make a HUGE difference). Otherwise we could be chasing (unrelated) symptoms.

The power supply voltages can be quickly checked for problems when you start the cleaning.
 
Checked the PSU voltages. Outputs 3 and 11 are 32.7 and 32.9v so the Control Amp is getting the corect voltages. The differences with the design voltages are:

Output 3 is 65.9v and 1 is 70.5 v.
Q12 is 55v
Q8 is .3v and 64v

The reminder are close to design figs.
 
hifileslie said:
Checked the PSU voltages. Outputs 3 and 11 are 32.7 and 32.9v so the Control Amp is getting the corect voltages. The differences with the design voltages are:

Output 3 is 65.9v and 1 is 70.5 v.
Q12 is 55v
Q8 is .3v and 64v

The reminder are close to design figs.
Click to expand...

A DMM on DC will average out the ac signal on a dc reading. Look for the AC voltages on the DC supply lines. Set the DMM to AC and read all the PSU voltages.

inductor said:
regulators need checking that feed the preamp section for ripple or dry solder joints, and\or dried out capacitors
Click to expand...

I'm confused:
1 -52v regulated
2 -52v regulated
3 +52v regulated
4 +52v regulated
5 +32v regulated
6 +32v regulated
7 +18v regulated
8 0v ground
9 raw -48v UNregulated
10 -32v regulated
11 -32v regulated
12 raw +50v UNregulated
13 raw +68v UNregulated
14 0v ground
15 Raw -68v UNregulated

Q8 E= 0.0v B= -0.6v C= -51.4v

Q12 E= -32v B= -31.4v C= -48v (it is acceptable for the Collector voltage to be different)

If I make some assumptions, the ripple from shorted + / - 52v regulators (you say 65.9 & 70.5) will not be as noticable in the amps, but a fault could affect the master zener diode reference in the +52v regulator and with it ALL the other voltages.
 
I much appreciate your help-I could not find all of those voltages in the Service Manual!! I have measured all output voltage now:

Pin 1&2 71.2 v dc 0v ac
Pin 3&4 66.6 v dc 146v ac
Pin 5&6 32.7 v dc 72v ac
pin 7 18.4 v dc 40v ac
Pin 9 55.2 v dc 0v ac
Pin 10&11 32.7 v dc 0v ac
Pin 12 52.6 v dc 116v ac
Pin 13 70.7 v dc 157v ac
pin 15 72.6 v dc 0v ac
 
hifileslie said:
I much appreciate your help-I could not find all of those voltages in the Service Manual!! I have measured all output voltage now:

Pin 1&2 71.2 v dc 0v ac
Pin 3&4 66.6 v dc 146v ac
Pin 5&6 +/-?32.7 v dc 72v ac
pin 7 +/-?18.4 v dc 40v ac
Pin 9 +/-?55.2 v dc 0v ac
Pin +/-?10&11 32.7 v dc 0v ac
Pin 12 +/-?52.6 v dc 116v ac
Pin 13 +/-?70.7 v dc 157v ac
pin 15 +/-?72.6 v dc 0v ac
Click to expand...

First, polarity is important, a missing + is acceptable, but a missing - can cause big problems.
Second, were ALL these measurements taken with the negative (black) lead of the meter connected to a metal chassis ground?

The voltage on pins 1&2 is BAD, too high
The voltage on pins 3&4 is BAD, too high
The voltage on pins 5&6 is GOOD
The voltage on pins 10&11 is GOOD

IF you are not in the USA, I don't know where to tell you to get replacement transistors, nor what type to use for each.
IF you ARE in the USA, I'll make a list.

It will be safest to replace Q1,Q3,Q5,Q7 for the +52v regulator and Q2,Q4,Q6,Q8 for the -52v regulator, unless you can test each of these out of circuit and identify any that have survived.

Did you download the 34 meg hi-res service manual from the database?
 
I have downloaded the hi-res SM.
With the -ve lead connected to chassis ground all polarities are correct(just the values are wrong!!)
I can see the transistor types from the SM but if you know the modern equivalents that would be great-I live in UK.
You mentioned possible damage to master diodes. Would it be worth changing these? If so are these D5 and D6? Do you know what type they are?

Could the AC on the outputs be due to these components having failed?
 
hifileslie said:
I have downloaded the hi-res SM.
With the -ve lead connected to chassis ground all polarities are correct(just the values are wrong!!)
I can see the transistor types from the SM but if you know the modern equivalents that would be great-I live in UK.
You mentioned possible damage to master diodes. Would it be worth changing these? If so are these D5 and D6? Do you know what type they are?

Could the AC on the outputs be due to these components having failed?
Click to expand...

D5 is the only reference and IT is OK. The correct voltage of +32.7 volts on pins 5&6 and the -32.7 volts on pins 10&11 means D5 is correct. D6 is a conventional diode and is just used to prevent disasters during fault conditions.

Since you are in the UK, perhaps you could talk to Ak'ER inductor about the parts, IIAC he has a shop and has been contributing to the active sa-9100 thread on this (Pioneer) forum that I am also working with.

Once we have the voltages correct, then we can re-check for AC on the correctly operating DC lines. IF you are feeling adventureous, you could replace the electrolytic caps on the power supply. I'm not sure over there what that cost would be, but when I did an sx-3800 recently, the caps cost between $0.06 to $0.60 each. BE VERY CAREFUL to get EACH AND EVERY capacitor's polarity CORRECT before applying power. ( I goofed on one and it cost me 2 transistors, a capacitor and about an hour, not counting the sweat.)
 
hifileslie said:
Many thanks.
Do you know the modern equivalents of the transistors/JFETs?
Click to expand...

I know American equivalents from Fairchild, and where to order them in the USA (mouser.com, Dallas- Ft. Worth Texas USA), but that won't help you over there...

THAT'S why I suggested you PM AK'er inductor for help, he is in the UK!! (and has a shop??)

And a few of the older ones (like the 2sc1735) may have a thicker collector lead which was intended in these old rare transistors to conduct heat OUT of the transistor, dissipating more heat than modern equivalents can, through their cases alone. Thus I choose a different case, and transistor that should be very robust, there should be enough room to fit it in.

The ksa910 and ksc2310 are complementary transistors (opposite in polarity, but matching in specifications), and in a few cases are safely over rated for the application (a good thing) while the ksc2682 AK'er inductor knows as his 2sc2911 replacement for the fuji 2sc1451.

The jfets may be the hard ones to find, as they are 50 volt parts (rare).
(Fairchild KSK30 has worked for me)

Thus you need 4 transistor types:
a 50v jfet
a higher power pass transistor
a complementary pair.(counts for two)



Q1, Q2 = 2SC1735 BCE 100V, 0.5A 0.8W 130MHZ use KSC2682 (power pass) ECB (leads are REVERSE order)
Q3 = 2SA733 ECB 60V, 0.15A, 0.25W 190MHZ use KSA910 (npn of pair) ECB (leads are SAME order)
Q4, Q8 = 2SA904 BCE 120V 0.05A 0.2W 150MHZ use KSA910 (another npn of pair) ECB (leads are REVERSE order)
Q5, Q6 = 2SK34 SGD use KSK30 (jfet) SGD (leads are SAME order)
Q7 = 2SC1914 BCE 90V 0.05A 0.2W 150MHZ use KSC2310 (pnp of pair) ECB (leads are REVERSE order)

These KSA, KSC & KSK numbers can be looked up on the web at:
http://www.fairchildsemi.com/crossref/search.jsp

BEFORE YOU INSTALL THE TRANSISTORS, CHECK THE LEAD ARRANGEMENTS OF THE NEW AGAINST THE ORIGINALS!!!
 
This is the mantra of any good tech:

IF IT AIN'T BROKE, DON"T FIX IT...

Leave D5 and D10 alone. Your readings have PROVED they are ok.

The ONLY reason I have recommended replacing all 4 transistors in each regulator is a balance between
1) the possibliity of damaging (and having to get) good parts by using a piecemeal approach with MANY voltage readings and out of circuit tests that will take a while to talk you through
and
2) the probable low cost of the replacement transistors.

Plus the recommended replacements are MUCH more robust.
 
Help With SA-9800

Well I changed the transistors/JFets and turned on. No noise and correct voltages/polarity except for :

9. -55.2
12. +53
15. -70.9

Switched off and turned on again to double check voltages and very loud hum had returned. I measured again . As before but:

3. +70
4. +70

The +52 reg has obviously failed again.

Which of the transistors/jfets are the +52 reg? (to save me removing the wrong ones!!)
What would/could have caused the failure again?

Advice gratefully received-so close now!!
 
markthefixer said:
1 -52v regulated
2 -52v regulated
3 +52v regulated
4 +52v regulated
5 +32v regulated
6 +32v regulated
7 +18v regulated
8 0v ground
9 raw -48v UNregulated
10 -32v regulated
11 -32v regulated
12 raw +50v UNregulated
13 raw +68v UNregulated
14 0v ground
15 Raw -68v UNregulated
Click to expand...

pins 9,12 and 15 are raw dc and are fine.

Q1, Q3, Q5 and Q7 are the +52 volt regulator.

What was the source of the transistors you used, ESPECIALLY Q5, what was the part number of the Q5 jfet you actually installed? ( i think mouser was out of ksk30's when I last checked )

Is C8 the 47uF 63 volt capacitor getting warm? It could be extremely leaky, pulling too much current for the regulator to handle.

If I remember correctly, the +52 goes exclusively to the power amps, and they also have a 47uF 63 volt capacitor from +52 to ground, one on each channel. Any one of these could be in the process of failing, and drawing too much "leakage" current. While you could do ohmmeter tests on each of these caps, the leakage may not show up at the violtages that the ohmmeter uses to operate.

edit: inductor posted as I was typing, he is RIGHT, open up the wires connected to pins 3 & 4, it will eliminate the capacitors on the amplifier board., then if the +52v regulator survives, we know to look elsewhere,

IF you are poking around the caps on the amplifier boards, be VERY CAREFUL, the FRAGILE bias diode's connections are made here, and then the leads go to the bias diode itself, which is mounted on the heatsinks that have the power output transistors.

There is another sa-9800 thread here in the pioneer forum where we chased both power supply problems, a bad 47uf 63v cap AND a bad driver transistor, that, if i remember correctly, didn't sound too bad when run. there is valuable information in that thread that MAY apply to your situation. we will probably end up doing the same things...
 
Should have mentioned that the +52 reg had been replaced before!! and that C7 and C8 look as if they have been changed and are 100 uf and not 47 uf!!

Next I will strip out the 4 transistors to see which has failed.
 
hifileslie said:
Should have mentioned that the +52 reg had been replaced before!! and that C7 and C8 look as if they have been changed and are 100 uf and not 47 uf!!

Next I will strip out the 4 transistors to see which has failed.
Click to expand...

Yes, but PLEASE, what are the numbers of the transistors you used as replacements and which ones were replaced?

The 2sk34 Jfet is a 50 volt part, and these are geting rare, most are rated to about 35 volts. Also others have depended upon vendor's cross references that have turned out to be ill advised.

The C7, C8 caps really should be 80 or 100 volts, 47uF. Plus C8 being 63 volts has been over-voltage stressed and may have failed. It will need to be replaced, plus when powering up the +52 lines should be disconnected, and a current meter will be used to check the drain to each wire/amplifier channel.
 
The transistors I used were:

Q1 Q2 MJE 340
Q3 BC556
Q4 Q8 BC556
Q5 Q6 2N4338 ( Vishay 50v JFET)
Q7 BC546

I have tested them in circuit with a multimeter hoping to id which had failed but no obvious short circuits on any of them. Not sure about failure mode of a JFET-short circuit?
 
I think we need to be on the same "page" as far as transistor testing, this should help, PLUS if you follow the links, you should also find jfet testing procedures.....

markthefixer said:
<snip>

In the meantime, you will want to aquaint yourself with the DMM diode test setting and how it is used to evaluate transistors for damage.

This post should also help pierce any fog after reading EchoWars's landmark testing post. It establishes a methodical reporting format that eliminates confusion. Just copy it in toto and edit in your results for each transistor.
Click to expand...

I checked the specs on the transistors:
q1,q2 using MJE340 is an EXCELLENT choice. the original lead layout is BCE and the mje340's is ECB, reversed
Q3 using BC556 is marginally ok, as BC556 is a 65v Vceo transistor. the original lead layout is ECB and the BC556 is CBE, scrambled
Q4,Q8 using BC556 is NOT ok, as BC556 is a 65v Vceo transistor versus the original's 120v Vceo. the original lead layout is BCE and the BC556 is CBE, scrambled
Q5,Q6 using 2n4338 is an excellent substitute, the original lead layout is SGD and the 2n4338's is SDG, scrambled
Q7 using BC546 is NOT ok, as BC546 is a 65v Vceo transistor versus the original's 90v Vceo . the original lead layout is BCE and the BC546 is CBE, scrambled

Q4, Q8 and Q7 could benefit from a higher voltage rating, Pioneer did use 90v and 120v parts in those positions, and I would be uncomfortable using any part less than 80v (no load DC voltage across C2, also that's it's rating) even in those spots.

It could explain why it regulated correctly the first time (and failed on re-power up with something charged?). The Q1 pass transistor can handle the load, as long as Q3 can deliver the base current. You may want to check and see if you can find Zetex transistors in your area, the Zetex ZTX795A (PNP, 140V, 0.5A, 1W, 100MHz) would make an excellent substitute for Q3 and Q4 while the Zetex ZTX694B (NPN, 120V, 0.5A, 1W, 130MHz) would make an excellent substitute for Q7. Both have oodles of low current gain, while they can deliver a prodigous amount of current. It's just that they cost about $1 each.
 
You must log in or register to reply here.
Back
Top Bottom