Marantz 18 receiver needs help

[The Dude]

I had a receiver returned a couple of times years ago. The customer lived nearby, so I went over to his house. His pet rabbit had chewed through the speaker wires. Sorry, I'm not going to warranty that!

[Leo_Gnardo]

Comments?

Time to "think outside the box." In this case, the box that holds the Marantz 18. 52 Bill, olddawg and The Dude just gave you some important suggestions. I hope you'll follow through on them and hopefully head off another transistor meltdown.

[The Dude]

Not to hype on it, but...... Another time, a customer volunteered that he had built his own stereo speakers. Each one had 4) 6x9 4 ohm car speakers wired in parallel giving him a one ohm load per channel. Unfortunately, you never know what the customer is doing with the amp after you fix it. You can either find out and make him bring in the speakers and wires, or keep fixing the amp for free.

[bigdrums925]

I don't have the idle current set to 80mV anymore in an attempt to keep the current draw down during turn on.
Somewhere in this thread was a comment that 80mV seems high even though the service manual recommends it.
At one point I attempted to reduce it a little as a final courtesy but the access hole shorted to the metal trim pot slot and bam, had to replace the power transistors again.
I know I'm supposed to do that with an insulated tool but the access is quite small and the pot hard to turn. My expensive and embarrassing mistake.

The bias/current trim is set as low as I could make it, referencing my amp meter on the variac because I can't get to full power up yet.
However, with the 8 ohm dummy load and an 800 hz signal, volume at 50% & 50% variac, the voltage drop is 6mV sometimes less.
So I increased the variac to 54% where it showed a .3 amp draw on my amp meter. This raised the voltage drop to 105mV.
Nothing seems to be getting hot yet. But what do I do next and still insure I don't risk blowing any more expensive parts?
In other words how far can I push this test? Until I see current draw of .8 - 1.0 amp perhaps?
I can surmise the voltage drop across R35 would be as high as 300-500mV at that variac level.
Or is temperature the major concern here? I have a digital infrared thermometer if that is the case.

BTW I can't be absolutely certain but believe the underlying issue with the returns is something I missed or a combination of small things that begin to fail after time, not an issue of user error. Normally I am very suspect of user errors but not in this case. That with not knowing exactly what parts to replace with what currently available parts & which numbers has me on edge. Thanks Marantz! I thought them better than that.
For what its worth it tests fine here, but I should probably be running it harder and longer before returning it to the customer to ferret out any possible hidden issues.
It's rather old and difficult to work on particularly considering how well loved these old Marantz's are. Seems whatever I do to it has a high risk of another failure.

[bigdrums925]

advice well taken dude. I think we could be bros from different mothers!
Everyone else as well. I will double check for any issues on his end

Leo: I got it. and yes, this group of extremely talented and experienced techs is the best! Keeps me in business.
Hopefully I can be of similar service down the line.
Thank you all.....

[g1]

When I remove the output transistors (on the right channel) the current draw is very low all the way up to 100% variac.

I'm not sure if I'm understanding correctly, do you mean current draw for the whole amp? So left side draw is low but with right side it is much higher?
If that is the case, I would not release the amp again until you can get the currents roughly equal.

[nickb]

I don't have the idle current set to 80mV anymore in an attempt to keep the current draw down during turn on.
Somewhere in this thread was a comment that 80mV seems high even though the service manual recommends it.
At one point I attempted to reduce it a little as a final courtesy but the access hole shorted to the metal trim pot slot and bam, had to replace the power transistors again.
I know I'm supposed to do that with an insulated tool but the access is quite small and the pot hard to turn. My expensive and embarrassing mistake.

The bias/current trim is set as low as I could make it, referencing my amp meter on the variac because I can't get to full power up yet.
However, with the 8 ohm dummy load and an 800 hz signal, volume at 50% & 50% variac, the voltage drop is 6mV sometimes less.
So I increased the variac to 54% where it showed a .3 amp draw on my amp meter. This raised the voltage drop to 105mV.
Nothing seems to be getting hot yet. But what do I do next and still insure I don't risk blowing any more expensive parts?
In other words how far can I push this test? Until I see current draw of .8 - 1.0 amp perhaps?
I can surmise the voltage drop across R35 would be as high as 300-500mV at that variac level.
Or is temperature the major concern here? I have a digital infrared thermometer if that is the case.

BTW I can't be absolutely certain but believe the underlying issue with the returns is something I missed or a combination of small things that begin to fail after time, not an issue of user error. Normally I am very suspect of user errors but not in this case. That with not knowing exactly what parts to replace with what currently available parts & which numbers has me on edge. Thanks Marantz! I thought them better than that.
For what its worth it tests fine here, but I should probably be running it harder and longer before returning it to the customer to ferret out any possible hidden issues.
It's rather old and difficult to work on particularly considering how well loved these old Marantz's are. Seems whatever I do to it has a high risk of another failure.

So you are saying that the idle current on one side is high while the other is high, even too high at 54%? Then you need to root cause that.

It's not a good sign that you cannot get to zero idle current as R27 can be set to zero. That will turn Q2 off leaving just the bias regulator diode in circuit implying that it is bad ( this was the RH channel - right?). Measure the voltage across C29 with R27 set to lowest idle current and report the drop across R36. Don't let the drop across R36 go over 100mV (say). Also check R37 are R42 are OK.

Leave the load off until you get the idle current sorted. Yes, the idle current in the service manual is very high. Personally I'd set it much lower (like a tenth of that) by using a scope or THD analyzer. What does the good channel take? Maybe use that as a guide.


Yes, temperature is a concern as the transistors will fail if the junctions get hotter than about 120C.

[bigdrums925]

I think I've found where the problem is located. R36 looks like a potentiometer with only 1 terminal while using the metal case as the other contact.
I would think there is a wiper inside like a regular pot and that reversing the leads while measuring the resistance of the pot directly would give the same resistance.
On the R channel amp, the problem child, it shows the same value which depends on the adjusting position, yet the same in either direction
But on L channel amp I get a diode reading, infinity in one direction, about 900 ohms in the other.
So I will probably have to unsolder a bunch of parts to get R36 out to see how it works. Seems odd to me because I'm measuring directly from the case to the only terminal.
I realize of course I must be measuring through the bias control diode, but I can't see well enough to see why.
I'll report back when I know more

It's not a good sign that you cannot get to zero idle current as R27 can be set to zero. That will turn Q2 off leaving just the bias regulator diode in circuit implying that it is bad ( this was the RH channel - right?). Measure the voltage across C29 with R27 set to lowest idle current and report the drop across R36. Don't let the drop across R36 go over 100mV (say). Also check R37 are R42 are OK.


I think C29 is mislabeled on the new drawing and should be labeled C25 (220nf) because both the original and new drawing show a C29 being located near Q10 and at 470pf.
Both the L & R channel bias regulator diodes measure good in circuit, 862 ohms/infinity. Strange looking things and Marantz p/n on them.

[The Dude]

I'm looking at the schematic in your first post. Is that the one being used here? If so, R36 is shown as a .47 ohm 5W resistor- not variable, so it doesn't jive with your post. Can you clarify? Without knowing for sure what part we're discussing, I'd be more apt to believe that there is an open junction (either diode or transistor) on right/problem child channel.

[g1]

R26 not R36? (and R27 for other channel) It is drawn as a rheostat so the actual part may only have 2 contacts. I'd expect it to measure the same with probe both ways when using the resistance (not diode) range.

[The Dude]

Agree. It should read 2.5k at one end of its travel and near shorted at the other end, if that's the VR we're talking about. The two channels won't read exactly the same unless the rheostat is adjusted for the same resistance.

[g1]

And if you do mean the rheostat (R26/R27), and you were using the diode range, do diode checks on B-E junction of Q1 & Q2.

[bigdrums925]

For some reason I am getting identical readings both directions on the R channel rheostat
But not on the working L channel
Since it seemed all other resistance & voltage measurements similar except for bias, it makes me think this must be source of the problem

I recall finding many errors in the original Marantz drawings so I'm not sure I'd trust them much. Besides whom else in 1969 would use outsourced parts branded
with only their own proprietary numbers, then let all that critical information get lost forever. I actually had a lot of respect for Marantz until working on this unit.
Some areas have modular circuit boards, while critical higher voltage areas, power amp section looks like a first attempt by a junior HS kid doing point to point wiring
with multiple components stacked on top of each other and almost no way to identify (or see) which parts correspond to which on the drawings. The bias control diodes
for example are buried under the related transistor and capacitor and look like a box capacitor not a diode. I didn't even know it was down there because I couldn't see it.
At least 6 other critical junction areas and can't be probed safely without risk of shorting out. Worse, some measurements can't be made in an energized state. And it's all tucked not so neatly under the chassis lips to further prevent access.
Even difficult unsoldering is common to check individual parts, but in this case many leads come together haphazardly with no support and very close to other terminals.
So removing 3 or more parts just to see the suspect is the norm. Forgive my rant.

So I don't know what to do now cause the working channel is the one that differs from what the drawing(s) imply.
All I know at this point is I have the bias trim as low as I can go on the bad channel and can only variac to 50% before the voltage drop goes through the roof. 49% is stable
but go 51% the current draw climbs rapidly but predictably, but not the voltage drop which rises very rapidly so I can't test very well.

[bigdrums925]

I unsoldered the trim pot and Q2 is soldered to that. Q2 tests bad. So how do I know what to replace it with since its Marantz's proprietary number 34-1014?

[bigdrums925]

I'm looking at the schematic in your first post. Is that the one being used here? If so, R36 is shown as a .47 ohm 5W resistor- not variable, so it doesn't jive with your post. Can you clarify? Without knowing for sure what part we're discussing, I'd be more apt to believe that there is an open junction (either diode or transistor) on right/problem child channel.

Frankly Dude, I can never tell myself which drawing I'm referencing at any given moment because I seem to have to jump around to verify which part I'm working on.
But ya, its the rheostat R27. Sorry bout that.
Again turns out you were right because Q2 is bad and I need to find a replacement.
Can anyone tell me what common transistor I can use as a replacement since it's label Marantz 34-1014