Mrantz Model 1090's Power tansistor Q723 always burned

Dear coachers, After I have studied all the skills and information from Mr. ecluser’s thread carefully: http://www.audiokarma.org/forums/archive/index.php/t-272813.html. I know how to adjustment the BIAS between Power Transistor finally, Thanks Mr. ecluser very much!!! I know more on my favour. I like 1090 warm and power voice very much! Everything of Right channel always in good. But I have still some troubles on My Marantz Model 1090 repair:
1. Major problem is Q723(NEC B618, Power Transistor, left channel) in main amp always burn after I changed it by 2nd hand B618.
Now the third has broken.
Before burned:
1) Test the B618 I would be used, It’s good at 216 Hfe.
2) Solder it to board and install heatsink correctly.
3) Inspect all 4 power transistor on main board, They are good .
4) Inspect the speaker terminals, Good, no shorted.
4) Power on firstly in fear and trembling, No fuse burn and other abnormity.
5) I adjust R771 let the voltage drop to 6mv on 0.47 Ohms (R763). I heard music for approximately 10 minutes then Power off—— I am ready to finetune after a while as Mr. Eclusters experience, I keep the test clip on the resister.
6) But power on again, the Voltage always in 1 mv?! I feel so strange, and inspecting the power transter, Oho! Q723 C-E shorted like before!?
Who could help me?! I am so confused the root cause: B618’s quality? Effect by Other around part ? I inspct them by digital Multi Meter, they are all good.
:tears:

slodering point photo on Tone Mode

Updated: could anybody give me a clear photo of all soldering points on 1090 Tone Mode selector?
Thank very much!:scratch2:

Do you have the proper mica insulators under the output transistors? If so are they in good condition?

Thanks for your remind. but I installed the mica insulator correctlly, actually it always be adhibited on the original post of huge cooper since 1977.
BTW, I changed the Q723 (Power Transistor, Left chanel) 3 times, and the longest normal working time is 1 month, the shortest is 10minutes this time.
I think before twice that I didn't not adjustment the BIAS after changing, so Q723 burned.
But in 3rd time, I power on after I inspected all the output terminal and transistor of left channel are in normal. Then I adjust R771 let the voltage drop of R763 is 5.6mv. and the speaker output is 7mv, at the same time right chanel is good: I connet a speaker on it and listen the music for 10 minutes.
During this period, the BIAS(5.6 mv) and 7mv(Idling output) are steady. Play CD, 5.6mv is lightly change around 5,6mv, It seems perfect.
I turn it off. After about 30 minutes, I power on again with happy, but the BIAS always is 1mv(no speaker contectted, no any speaker swithch on)?!
all the fusees are good. But I think if I connnected the speaker to LC, the two F001 would pop like the fault before.
So I think the Zener Diode Q741(W06B) is not in good condition, but inspect it (take it off from board) by DMM, it seems no problem?
antthing else? Thanks!

Zener diodes generally work or they don't, if you are getting the regulated voltage at the zener then it's ok.

Thanks, Mr. BKtheking! I just afraid it's unsteady condition or not.
Maybe as Mr. ChrisMarantz said, I didn't adjust the BIAS and burned Q723 twice after it was be replaced, Maybe the mico isolatoer become to bad condition. Now I disassembly the heatsink and check the mico isolating performance carefully, or change it by new one.
My poor Q723, My poor Shining 1090!

My poor board be soldered again and again. as before, Q723 shorted between C and E. Cast an eye on the mico, there are some cracks. I will buy some new mico flake and B618.
By the Way, if the voice style would be changed if I replace current Nec B618/D588 with Sanken 2SA1491/2SC3855?

Dear coachcers, normal mica insulator has installing hole, it lager than my crack on eye. Does it could not effect the isolating performance worse?

About the hole of mica

Sanken Transistor heatsink suggestion

Here is a picture of the mica insulators from a 1090 that I rebuilt. There are no holes in the insulator.

I'm guessing from what chris posted they are solid mica's with no holes, if your's have holes you might be shorting the case to ground.

About 1090 tone mode selector and Q723's mico isolator.

Thanks very much!
Mr. Bktheking, I have sent the order of mica isolator for MT-200, It's larger than the original (25*30 mm). and Its minmum package number is 1K. My God!
Mr. ChrisMarntz, I catched your word clearly before. Do you have the picture of 1090 tone mode selector soldering points from bottom side under the speaker switcher? Like my picture.
My dog was missing, So I have to do this agently in this week!

tone mode selector

Buring wating the Q723 (NEC B618), I Explain on the soldered drop junctions clear picture I wanted of 1090 tone mode selector.
There is a fitful fault in pre-amplifier: Both of 2 channel sound volume like buzz ( but without distortion or noise) when I switch the post of the selector somtimes, I afraid there are some ostensible soldered drop junction under the board. Thanks!:scratch2:
And keep looking for my puppy!:tears:

I'm not clear on the problem you are having with the tone selector switch.

I looked through my pictures and I don;t have one showing the bottom of the board.

I may have a spare board that I can take a picture of for you.

What do you want to see in the picture?

Thanks, ChrisMarantz.
I want to get the correct defination on soldered drop junctions like my picture. I doubt that my some junctions were disconnected. My service manual's diagram is not clear.

Sounds like the switches need a good cleaning.

Thanks! Cleaning always been down each half year.

Maybe My 1090 is repaired. Thanks everybody!

Mr. ChrisMarantz, I have changed mica insulator and B618 by new part. and then I adjust R771(LC) until 10mA(9.5mV) BIAS current is reached between emitter of Q721 & Q723. 12mA Idling current Service Manual asked, but I consider the tolerance on measuring & working status.
Likewise, Adjust R772 of Right Channel (Good channel before) until 9.5mV. By the way, I measuring value is 0mV before adjustment. Why? The root cause is Long time or primary value by marantz?
Beginning on cold, the BIAS current is approximately 1mV, and increase to 9.5mV I set up before within 1 minute.
Everything is OK. then I connect 8 ohms speaker on system 1 and System 2 at the same time(4 Boxes), Play CD in 11 o’clock volume(It’s always normal working status for me).
If I adjust the BIAS current to 10mV, it will up to 13mV after I play music 1 hour, And up to 15mV after 6 hours. So I decrease the BIAS voltage to 9.5mV. I would monitor the current in one week.
Here is my Max value: 256mV when I play dancing music.:thmbsp:

It sounds like you have the problem solved :thmbsp:

As for the bias drifting upwards with time/use, do you have the rubber pads in place behind the heat sink that push bias compensation transistors (Q711 & Q712) against the heat sink?

These transistors need to be in contact with the heat sink to work correctly.

In the 3rd picture you can see how the soft rubber pads fit between the clamping bar and the back of the transistor, pushing it against the heat sink for a good thermal contact.

Puzzle on the relationship of Power Transistor's between hfe and current valuve?


Thanks very much for your clear pictuer, Mr. ChrisMarantz.
Yes, the orginal good soft rubber pads always lie there behide Q711/Q712(LC/RC). I pay more attention on cooling.
and now, after 6 hours continous working, the premeditatedly 9.5mV are steady with silent background.
BTW, I add the copper wing heatsink on Q803 in pre-amplifier before.

Here is my new question:
What is the relationship of power transistor between hfe and current value?
I replaced the bad transistor of LC’s Q723 by new one. And I purchase 5 piece of B618. hfe value: 60/106/114/131/168. Which one is better? Why?
I installed 114-hfe this time, and I find the LC’s current value always less than RC’s in working.
Is it normal except the different recording level in each channel?