Marantz PM7200 Can't adjust Bias or DC Offset

[mossieurclo]

Thanks for reading me:

After i repaired the speaker protection board of that amplifier, all was ok.

So I then proceeded to adjust iddle current. Got it. Then DC offset: Ooh, boooy.

At that stage, on the left channel, after first seeing normal readings while moving the trimpot to get to the correct setting, suddenly readings were all over the place and did not respond to my turning the trimpot.

I played music to earphones, to find that there was no sound on left channel. Only distorted crackles on peaks.

I took out the board and took out the trim pot, measured it as being ok from end to end and from the moving wiper. Cleaned it anyway and retested as OK.

And, i resoldered 7265 that had broken solder joints. That component was probably the one acting up (relays clicking) when i pushed away the supply wires (touching 7265?) to have access to the trimpot.
(BTW why do these (TO126?) stand so high on their legs, wrapped in shrink tubing??)

I reinstalled the board. The audio is now heard on left channel. Distortion is hard to qualify on cheap earphones and with my old ears and in haste not to blow anything up.

But: i can't adjust either Iddle Current (.2mv) or DC Offset (160mv). Trim pots have no effect. With power off i rechecked that DC offset trimpot and it was still behaving correctly.

--I probed the supply lines as deep as i could, including around some transistors (maybe of interest: 7261 is E 53.2V, B -52.7V, E -53V). All seemed normal in my limited understanding of those circuits.
--I tested in circuit the first components branching out of + and -VB,
--and double diode tested 7265 and 7267.
--I did not probe under the copper shield (housing 7251, 7253, 7255, 7257, 7259) yet.

I hope there is an obvious "classic case" component causing that anomaly, which some knowledgable and experienced person here will point out to me.

I saw that i can unsolder (update: not easily!) that copper shield and probe those components. If someone points me to what i should measure there?

BTW all i have is a digital multimeter. It can test semiconductors. It also has capacitance capability but seems unable to measure >100?uf . To find good units in my NOS stash i rely on reading DC volts, discharge, remeasure DC volts, do long Ohms test, and remeasure DC volts to see if it went up. Served me well to date ;-)

Thank you all for your insights and help.

service manual:
https://www.hifiengine.com/hfe_downloads/index.php?marantz/marantz_pm7200_service.pdf

[ambrosia heart]

Components yellow circled are much cheaper than labour-cost, replacing them ALL.
Idling current must be adjusted properly; otherwise thermal protect or power transistors thermally broken down.

[mossieurclo]

Components yellow circled are much cheaper than labour-cost, replacing them ALL.
Idling current must be adjusted properly; otherwise thermal protect or power transistors thermally broken down.

Ah. these were my next targets for testing. You give me incentive

If 7267 tested ok on the double diode test, could it still be bad? 

I will replace it anyway, but i'd like to know 

[mossieurclo]

I got an e-mail saying Smoky had posted here...

Can't see that post. How come???

[mossieurclo]

I ordered a mega328 esr meter, and a solder sucker. 1 week?

I tested a lot of a lot of diodes and fusible resistors, those i thought could be tested in circuit. All tested ok.

A lot more tests to do, especially after i remove the copper shields (i at least succeeded in tinning the old butane irons, those could help)

So i plan to annotate a .pdf with my readings and post it, hoping someone more experienced than i will be there to read it and advise.

[ambrosia heart]

Just replace ALL the resistors and caps in the bias network.  If idling current still cannot be adjusted, then replace the photo-coupler and resistors connected to it.
Idling current must properly be adjusted ONLY IF components located in amplifying circuit is/are replaced except capacitors.
Bias network provides enough bias voltage for the output stage of 4 power transistors connected in parallel to work properly. 

[coromonadalix]

i think replacing all the resistors is going too far,  just remove them and check their values outside the board

the 7267 (2sc3419) transistor could be bad, this one should be on the heatsinks

I had a problem like this in the past in a sony amp,  it was the "sense" transistor on the heatsink who could not do its job


check page 34 (attached file)    it's the adjustment you may need before changing all parts ?? 

The best way  is to inject an 1khz 0db  signal and follow with a scope thru the 2 channels and do comparaisons between them  until you see signal distortion(s)

[mossieurclo]

Just replace ALL the resistors and caps in the bias network.  If idling current still cannot be adjusted, then replace the photo-coupler and resistors connected to it.

Isn't that photo coupler only for the Class A bias??

I won't use Class A, i need more power for Elipson1303...

[mossieurclo]

i think replacing all the resistors is going too far,  just remove them and check their values outside the board

I checked the fusibles (3317 to 3333) in circuit. They were ok. Can the multimeter "see through" the transistors??

the 7267 (2sc3419) transistor could be bad, this one should be on the heatsinks

I had a problem like this in the past in a sony amp,  it was the "sense" transistor on the heatsink who could not do its job

I diode tested 7267, out of circuit. Looked OK. I can imagine the "adjusting with temperature" part of it's job being broken, but the adjustment trimpots would still have effect, correct? As mentionned, they (Bias AND DC) have no effect.

The best way  is to inject an 1khz 0db  signal and follow with a scope thru the 2 channels and do comparaisons between them  until you see signal distortion(s)

I don't have a scope. I could try injecting that signal or playing music from a phone, and follow the audio on earphones rigged with the probes. Which is what i was doing when i found 7265 needing resoldering.

Can't wait to receive what i ordered...I even ran out of desoldering wick 

Thank you for your help (everyone).

[mossieurclo]

igottheshieldsoff igottheshieldsoff igottheshieldsoff igottheshieldsoff

! ;0 GOT  THE ;O SHIELDS OFF ;"!!!!!!! 

:banana::jump::banana:

[mossieurclo]

I ordered a mega328 esr meter, and a solder sucker. 1 week?

So i plan to annotate a .pdf with my readings and post it, hoping someone more experienced than i will be there to read it and advise.

I got the meter and solder pump today.

I calibrated the meter, rigged pins and alligators to extend the ZIF socket, tested a few cans in my NOS and 7267 of the Marantz that i had taken out (it's good :-().

That puny meter looks useful. 

For Google customers searching, a pretty good manual (not supplied with the lower priced item from Amazon.ca) for "LCR-T4 Mega328 Component Tester" is here:

https://kookye.com/download/manual/mega328_transistor_tester_user_guide-...

Now brace yourselves, friends. Data storm coming soon. Thanks for weathering it.

(i should not play with those words, sorry for those who live through those very real weather storms down south)

[Seekonk]

Surprised nobody asked the voltage across this diode array.  Not being able to adjust current would point to that.  This brings back memories when i put myself thru college repairing high end audio.  I knew Saul Marantz. He was a funny guy with great stories to tell.

[mossieurclo]

Surprised nobody asked the voltage across this diode array.  Not being able to adjust current would point to that.

I probed those voltages a while ago, after resoldering 7265 and finding myself not able to adjust Bias or DC (IIRC, should have done a journal):

B+: 56V
6275 Cathode: 55.8V
6271 Anode (6275 A): 1.66V
6269 C (6271 A): 1.32V
6267 C (6269 A): -0.33V
6265 A (6267 C): -1.33V
6273 C (6265 C): -1.663V
6273 A: -55.8V
B-: -56V

Are these what they should be?

This brings back memories when i put myself thru college repairing high end audio.  I knew Saul Marantz. He was a funny guy with great stories to tell.

Hmmm. I guess that makes you even more "venerable" than i? 

Thanks for those thoughts. Probing around diff pair coming up...   

[ambrosia heart]

I probed those voltages a while ago, after resoldering 7265 and finding myself not able to adjust Bias or DC (IIRC, should have done a journal):

B+: 56V
6275 Cathode: 55.8V
6271 Anode (6275 A): 1.66V
6269 C (6271 A): 1.32V
6267 C (6269 A): -0.33V
6265 A (6267 C): -1.33V
6273 C (6265 C): -1.663V
6273 A: -55.8V
B-: -56V

Are these what they should be?
Despite  of unable to know actual  values, 
these maybe what they should be.  Why?  They are symmetrical.   
Don‘t you see +/-1.66V, implying no DC protect and over-current protect as well as mid point voltage is zero.

[mossieurclo]

Despite  of unable to know actual  values, 

Actual values of what? Results of Forward and Reverse Bias, or voltage drop of the Diodes??

these maybe what they should be.  Why?  They are symmetrical.   

Yep. I saw the symmetry. I don't know enough to tell if they are the right voltages or not.

Don‘t you see +/-1.66V, implying no DC protect and over-current protect as well as mid point voltage is zero.

Do you mean that protection will not be triggered with those values (which is good?)??
Or that those values prevent protection from working??
And the mid point being 0V (-0.33 close enough?) is what is normal, or not? (in my limited understanding, that line should be at 0 for a push-pull stage?)

I am enclined to understand from what you say that all is normal in that area. Correct?? 

Thank you for your insight (and explanations for a noob?)

[xavier60]

By measuring the voltage drop across the 150Ω resistors, VAS current can be calculated. Ill edit this post if I can figure out what it should be.
Edit: I'd say 8mA. 

[xavier60]

The diodes circled in teal are for output current limiting. They can be temporarily disabled.

[mossieurclo]

Components yellow circled are much cheaper than labour-cost, replacing them ALL.

Hmm. I forgot to thank you for that nicely annotated schematic. Thank you 

But and why the popcorn? I am putting on a good entertaining thriller ?

[mossieurclo]

By measuring the voltage drop across the 150Ω resistors, VAS current can be calculated. Ill edit this post if I can figure out what it should be.
Edit: I'd say 8mA.

Last readings i took after resoldering 7265 were 0.8V "across" 3293, 1.4V "across" 3287. Hmm..suspect?

In circuit, 3287 reads 148.8Ohms, 3295: 151.5Ohms (3294: 148.6, 3288: 147.7). Is that critical?

I put quotes because these are readings not across but derived from voltages readings with black probe on ground.

Anyway i plan to redo all readings with a fully warmed up ramp.

[mossieurclo]

The diodes circled in teal are for output current limiting. They can be temporarily disabled.

Your're talking about 6265 to 6271?

The (maybe wrong) way i understand, removing (jumping over) those would put -1.6V AND +1.6V on the 0V "middle line". Would'nt that a be a short?? 

[xavier60]

The diodes circled in teal are for output current limiting. They can be temporarily disabled.

Your're talking about 6265 to 6271?

The (maybe wrong) way i understand, removing (jumping over) those would put -1.6V AND +1.6V on the 0V "middle line". Would'nt that a be a short?? 

Yes, 6271,6269,6267 and 6265. They can be left open or carefully check them for leakage. Their function is to clamp the maximum Base drive voltages with respect to the output line, limiting output current.

Important, 6267 in the idle current bias circuit is configured as a VBE multiplier. It clamps the voltage difference between the Bases of 7271 and 7173 at about 3V causing all of the transistors in the output stage to lightly turn on, idle current.
If this clamping action is lost for any reason, allowing a much larger difference voltage to be applied between the Bases of 7271 and 7173, destructively high current will flow from rail to rail, through the output transistors.

Ill wait for more certain readings to be taken across those resistors.
Edit: The other channel can be used as a reference.

[ambrosia heart]

Despite  of unable to know actual  values, 

Actual values of what? Results of Forward and Reverse Bias, or voltage drop of the Diodes??

these maybe what they should be.  Why?  They are symmetrical.   

Yep. I saw the symmetry. I don't know enough to tell if they are the right voltages or not.

Don‘t you see +/-1.66V, implying no DC protect and over-current protect as well as mid point voltage is zero.

Do you mean that protection will not be triggered with those values (which is good?)??
Or that those values prevent protection from working??
And the mid point being 0V (-0.33 close enough?) is what is normal, or not? (in my limited understanding, that line should be at 0 for a push-pull stage?)

I am enclined to understand from what you say that all is normal in that area. Correct?? 

Thank you for your insight (and explanations for a noob?)

After reading the schematic, words I written on this post must take it back! Your amps mid point voltage is slightly unbalanced. The voltage measured at mid point should be < 0.1V. This unbalanced voltage perhaps is the reason why idling current is unadjustable. 

If I were you, I would :
(1)measure all resistors on the bias network to see if they are bad. Measure them on board maybe a easy way. Only take them out to measure when the value is largely different from marked ones on service manual.
(2)Check all these components: photo-coupler, 2SC3419, 6265,6267, 6269 and 6271 with T4-LRC meter. I also bought it. Leakage cannot be tested with T4-LCR. 6265,6267, 6269 and 6271 will not get damaged even 2 channels (SR & R) power transistors CE shorted sparking off over-current protection ASO in Denon AVR 3312 . They only give bias to the drive stage of the power amp not adjustable . Idling current of output stage of power amp is adjustable.

Check photo-coupler  with  T4-LCR
https://karpsy.ru/en/printery/proverka-optopary-optopara-pc817-princip-raboty-i-ochen-prostaya-proverka.html

If they are OK, measure the voltages at these points , ab, cd...... with respect to ground. They must be highly symmetry!  Comparing the faulty channel with another one . Find out at which stage, the voltage becoming unbalanced and fix it. Every time I repaired a faulty amp, I will completely fix them before adjusting idling current. I don"t know  if idling current can adjust under voltage unbalanced condition  .

And the mid point being 0V (-0.33 close enough?) is what is normal, or not? (in my limited understanding, that line should be at 0 for a push-pull stage?)

Why did DC -0.33V not trigger protection? -0.33V is abnormal due to mid point voltage < 0.1V.

protection schematic analysis
(1)over-current protect (black arrows)
When a power transistor shorted, a large current flows from A to B and to 3337. At
the same time, a part of the current will now flow from B  to R1,R2 and D1.
D1 protect Qp from burning up.
Voltage sets up in R2 making Qp ON. When Qp is ON, making Vx potential largely drop,
changes Q2 from OFF to ON. Current flows from +B viaQ2, R2 to pin 1 of TA
7317P with C1 charged up. Your amp. will now be  under protection.

(2)DC protect (white arrows)
When +B and -B is unsymmetrical, mid point voltage becomes non-zero, a current will now be
flowing through R3344 to pin2 of TA7317p with C2 charged up. The threshold value of pin2
to make your amp. under protection is +/-0.9V.  Since your measured value is -0.33V, this value
will not spark off DC protection. You may download TA7317p datasheet to read. There may be
something wrong with your amp which may make idling current unadjustable.

[ambrosia heart]

see photos

[mzacharias]

Been working in audio over 40 years. Never seen 0.33 volts be enough to trigger any protection circuit.

It simply is not sufficient to endanger a speaker.

Less than ideal for sure, though.

A protection circuit will usually trigger around 1.2 to perhaps 2 volts DC.

Amplifiers equipped with conventional DC sensing are set this way so as to avoid false triggering with asymmetrical waveforms.

[mossieurclo]

After i repaired the speaker protection board of that amplifier, all was ok.

So I then proceeded to adjust iddle current, got it, then DC offset. Oh, bboy.
[...]

Summing up how that problem was resolved (thank you to all who contributed!!):

I replaced fusible resistor 3315 that had gone very high.

And lowered 3301 (and 3302) to 220R to bring DC back within adjustable range (7267 and 7268 having drifted?)

Now will be replacing lots of electrolytics, all other fusibles, and putting back 7289 that i blew up (remember: discharge those Caps!), to reenable Protection. Looking for that 2SA970 and other through hole legged critters i discovered that these are now really scarce, like in hard to find.

Are NTE replacements viable??