2325 - Inop Right Channel

Osibisa

AK Subscriber
Subscriber
I am coming to find that Marantz is like a Lay's potato chip. Bought this 2325 knowing that "it lights up, but no sound", figured it would be another interesting project.

Well 1/2 of dial & indicator lamps actually work and removing the cover revealed a missing dial cord shield, a few missing screws, and the relay cover off rattling around the interior. Inspection of P800 showed signs of the wires to J810 & J811 being removed and soldered back, so some attempt was made to service the board.

The original LY2 relay was still in place and I noticed that the contacts which would have fed the right channel were well worn. I suspect the previous owner removed the cover and tried to reposition the contact points to compensate for wear instead of replacing. This board has since been rebuilt (once the old relay was removed I saw signs of possible arcing)

Upon power-up the relay closed after normal delay then opened about 15 seconds later. Figured I'd pull the right amp module and power up again, relay closed and stayed that way. Swapped module to left side and relay would not close, so bad module. Preamp is ok.

So down to my question, modules were never removed prior to me taking them out, visual inspection shows no burnt components although cap C706's jacket has shrunk. In this scenario what would be the usual suspects to focus on other than the four outputs?
 
Typically two scenarios,fault in protection circuit (component, solder joints, wires,,,) or abnormal voltage detected and the protection circuit is doing it's job.
Since the amp was ok for 15 seconds suggests a voltage drift that triggers protection, especially with C706 mentioned.
Zener H716 is a possible candidate.

Suggest measure Vdc at H716, H717 on both channels
 
Well that doesn't appear to be the problem, removed the four outputs and did some testing.
Got expected results so far and interesting enough the relay closed when I started then opened again after 15 - 20 seconds.
Would be a lot easier to test if the module wasn't in the chassis, saw that a member built a harness but of course I can't put my hands on it now.
 
I assume that you checked the zener voltages after the relay opened, 15-20 seconds?

I'll have a look at the sm to get some test points, need to confirm that it's excessive Vdc or maybe a fault in the protection cct.
 
After the relay opens, measure the Vdc at pin J808 on the relay/P800 board, then measure at pin J809.
Ideally less than 10mV, needs to be about 650mV (b-e drop) to trigger protection.
 
Sorry it took so long,

Yes I did check the zener voltages and they were in spec.

Here are the measurements from P800:

J808 = 7.4 mV
J809 = 57.5 V

Really appreciate your help here.
 
Last edited:
J809 57.5V is not good, almost rails voltage on the speaker line.

The two common scenarios are, 1) short on the positive half of the amp, dumping excessive
positive voltage where it shouldn't be, 2) open circuit on the negative half of the amp so
a negative voltage is missing that would normally cancel out a pos voltage.

There is no easy way to track this down, only painstaking measurement. However
H706,7,8,9 are seen as strong candidates (outputs pulled?). Please measure Vdc at e/c/b of
each of the transistors, measure at connecting component, eg H708e at R743, one slip of
the probe... Black probe connected to chassis as per normal
 
I had already pulled the outputs before taking measurements, will check the connecting components as suggested. Did notice R719 getting warm faster than the same resistor in left channel if that is relevant.
I'll post findings after testing.
 
Figured I knock the measurements out tonight:

J706 = 61.8v
J709 = 60.5v
J708 = 59.1v (should be 7.5mV)

J707 = 62.0v

J702 = 59.2v (should be -7.5mV )
J703 = -62.0v
J705 = -62.0v

J704 = -62.1v
 
Last edited:
H708b is 59v at relay trip. Cold it was 0.5v then started to rise after 15 - 20 seconds.
 
In a situation like this, I generally start by testing every semiconductor. If you have the proper test device to check them in circuit, it saves a lot of time.
If you only have a DMM or other out of circuit tester, then removal may be your only option.

Tom
 
Tom, I had mentioned in my last post that when "cold", H708(B) measured within the expected voltage.
Since I only have a DVM, isn't it possible that a component might test as good when it failed in-circuit with power applied?
 
It's possible that a component could test good and not work in-circuit, but I have rarely seen that issue.
Voltage testing power amps with negative feedback is a challenge, they are kinda like a snake with it's tail in it's mouth.

I use an ancient Huntron Tracker, something like a Model HTR1005. It will test semiconductors in circuit at various voltage levels.

Tom
 
In a situation like this, I generally start by testing every semiconductor. If you have the proper test device to check them in circuit, it saves a lot of time.
Yes, certainly agree with this as a general start however...

when "cold", H708(B) measured within the expected voltage.
Since I only have a DVM, isn't it possible that a component might test as good when it failed in-circuit with power applied?
I would expect parts to test OK when cold

H708b is 59v at relay trip. Cold it was 0.5v then started to rise after 15 - 20 seconds.
OK, something is failing with time, the obvious candidate is H706, however this can be something of a sucker bet as the cause may be somewhat subtle.

J708 = 59.1v (should be 7.5mV)
H708b, so b-e is "only" 0.1-0.2V?
Suggest diode test H708b-e for possible short, in circuit ok

Suggest replacing H706 2SA818, possibly KSA1381 however different pinout, maybe others can chime in with better sub.
 
Apology, I have not done the necessary checks.

Incircuit testing can be quite tricky due to alternate paths.
I've just checked for alt paths and don't see anything that would explain your results.
I would expect the 0.539 is the b-e drop, so it's not a near short which was my question.
With +59Vdc on H708b I'd be looking at H706, but like it's been said, these issues can
be quite tricky.

Nice graphic from markthefixer on diode testing

DiodeTest.JPG

Maybe consider replacing H706 with the working channel?
 
I'll take a leap of faith and replace H706, H707, and H714 for good measure. I'd rather not interfere with the working channel.

My understanding based on a few previous threads with a similar issue is that these subs are suggested:

H706 (2SA818) ---> KSA1381
H707 (2SC1628) ---> KSC3503
H714 (2SC496) ---> MJE243G

Do you concur?
 
Others can do a better job on transistor substitution.

My take is,
2SA818 ebc 180V, 5V, 50mA 1W 70-240hfe(5V/10mA) 120MHz 5pf
KSA1381 ecb 300V, 5V, 100mA, 7W, 40-320hfe(10V/10mA) 150MHz 2.3pf
graphs are mostly nice and linear, looks ok to me

2SC1628 2SA818 compliment
KCS3503 KSA1381 compliment
ecb 300V, 5V, 100mA, 7W 40-320hfe (10V/10mA) 150mHz 2.6pf
looks ok to me

2SC496 ecb 40V 5V 1A 1W 40-240hfe (2V/50mA) 100MHz 10pf
MJE243G ecb 100v 7v 4A 1.5W 40-180hfe(1V/200mA) 40mHz 50pf
MJE243 does not appear a great match, graphs are not real linear either.
I think you can do better. I don't like the MJE243 non-linear graphs.

Suggest wait for others to chime in.
 
TO-202 parts (2SA818 and 2SC1628) don't have the collector as the center pin and are hard to source.
I don't like bending the legs around to make a TO-126 or TO-220 part fit in place of a TO-202.

I don't particularly like the MJE243 for the reason you state, 2SC496 is readily available.

Tom
 
Back
Top Bottom