Solution for 2270 high bias - early serial number

It's becoming puzzling for me.
Meanwhile I bought Hitachi D669 that I am pretty sure are fake. You can't see it in the pictures but they don't look or feel like NOS at all. Like knocked off yesterday rather. I have never seen such shiny and thin (flat) legs on transistors either.

d669a.jpg d669b.jpg
 
Last edited:
So it is the matter of setting the idle current, right? I can set with no problem the 35V primary voltage, then DC Offset fairly close to zero.

Now the Idle Current, however I set it it goes up and up seems to never stop. Same in my 2270 with the round compensating/bias transistor 2SC984 and my other 2270 with the flat one 2SC496. Should it be 5mV for 2270 as stated in manual? I think I saw somewhere it should be 10mV rather.

And what if it's much too high? Will it overheat the amp? Cause more distortion?
 
The setting depends on where you measure. If you measure between J754 and J760 as recommended by many, the setting is 10mV. Are you giving it lots of time to stabilize? I find that it takes 20-30 minutes to stabilize. I keep an eye on it during that time and constantly adjust it to something under 10 mV. When it rises above about 15 mV, I turn it back down to 9. Eventually, it should stabilize. If not, there is trouble somewhere. One suspect would be the bias transistor H760. For better insight into this, read this thread where expert member Ecluser has an in depth analysis. He believes the design is bad and recommends moving H760 from the main heat sink to the smaller driver heat sink, and has good theory behind his recommendation. Good reading for anyone having this problem.

http://www.audiokarma.org/forums/index.php?threads/marantz-2270-bias-wont-go-below-15mv.322206/
 
So it is the matter of setting the idle current, right? I can set with no problem the 35V primary voltage, then DC Offset fairly close to zero.

Now the Idle Current, however I set it it goes up and up seems to never stop. Same in my 2270 with the round compensating/bias transistor 2SC984 and my other 2270 with the flat one 2SC496. Should it be 5mV for 2270 as stated in manual? I think I saw somewhere it should be 10mV rather.

And what if it's much too high? Will it overheat the amp? Cause more distortion?
There are two different methods depending on your amp board and I believe it's to do with units of serial number being below 3901 or above it but I'm not quite sure of the number. Take a look at the service manual error sticky thread, I believe the info is in there.

Overly high bias will generate excess heat due to increased current draw. I'm nmo amp guru but I don't think distortion rises but perhaps the current will rise far enough to just run away and cook the amp, taking out the outputs most likely.
 
The setting depends on where you measure. If you measure between J754 and J760 as recommended by many, the setting is 10mV. Are you giving it lots of time to stabilize? I find that it takes 20-30 minutes to stabilize. I keep an eye on it during that time and constantly adjust it to something under 10 mV. When it rises above about 15 mV, I turn it back down to 9. Eventually, it should stabilize. If not, there is trouble somewhere. One suspect would be the bias transistor H760. For better insight into this, read this thread where expert member Ecluser has an in depth analysis. He believes the design is bad and recommends moving H760 from the main heat sink to the smaller driver heat sink, and has good theory behind his recommendation. Good reading for anyone having this problem.

http://www.audiokarma.org/forums/index.php?threads/marantz-2270-bias-wont-go-below-15mv.322206/
Yeah, good, J760 is easily accessible. I may have some success. But when I turn on from cold I have like 1mV and then it goes in 20 minutes up to the desired 5/10mV
 
If you can't find D669, the following work but don't have as low Vbe, I don't think. KSC3503; JE243G; 2SC495; 2N4923
I wonder if KSC2690 would work theoretically, Vbe aside. I don't remember why I bought these some time ago, I like it as it is plastic encased all around without collector at the back like most.
 

Attachments

  • KSC2690A.pdf
    152.7 KB · Views: 10
^^^ That's quite normal.
My other 2270 with the round tube like bias/temp. compensating transistors 2SC984 goes over the target with trimming resistors turned to the limit.
But max of 13mV when should be 10mV is no tragedy, is it? Marantz never gets hot really, if you had other brands you know what I mean.
From my experience with other amps setting correctly the primary voltage (35V here) is most important for heat.
 
Last edited:
Yeah, I think I will stay with the original cylindrical transistors for now.
BTW, this mounting outside the original bracket seems so lame, but hard to figure out anything else or maybe there are better ideas around

2270biasmoddone.jpg2270biasmoddone.jpg
 
Randy is correct (as is usually is), the bias (idle) current is dependent upon serial numbers.

o s/n<3901 – J753 and J754 – set R763 for 5mV

o s/n >3901 - J754 and J760 – set R763 for 10mV
 
I add longer leads to the replacement bias transistor, cut the old transistor bracket and reattach it to the heat sink with the screw going through the hole of the new bias transistor. Don't forget the mica insulator and thermal compound.
Sorry for the poor pictures.

Tom
View attachment 933940 View attachment 933941 View attachment 933942
Yeah. This is good. Thx. You mean the new transistor goes between the bracket and the radiator with the screw through its hole? You don't really need to cut the bracket, do you? I reckon you cut off the part holding the cylindrical transistor, I don't really get what and why you did to the part holding the diode though :) that it looks so ... different :)

BTW. I see these white paint marks on the trimmers for marking their position. I have them too in one 2270. I thought it had been made by a user but it seems it was in production. And in mine the marks are where the trimmers are turned to the limit for smallest voltage possible so they had to realize there was a problem with the bias being too high and never too low.
 
Last edited:
Yeah. This is good. Thx. You mean the new transistor goes between the bracket and the radiator with the screw through its hole? You don't really need to cut the bracket, do you? I reckon you cut off the part holding the cylindrical transistor, I don't really get what and why you did to the part holding the diode though :) that it looks so ... different :)

BTW. I see these white paint marks on the trimmers for marking their position. I have them too in one 2270. I thought it had been made by a user but it seems it was in production. And in mine the marks are where the trimmers are turned to the limit for smallest voltage possible so they had to realize there was a problem with the bias being too high and never too low.

Yes, the transistor goes between the bracket and the heat sink (radiator). I cut the bracket off because it is not used with the new transistor. I bend the 'tab' so that it presses the diode against the heat sink as the bracket is now the thickness of the new transistor body away from the heat sink.

The white paint on the trimmers is to secure them from moving after they were adjusted.

YMMV
Tom
 
I made Vbe measurements a while ago using an Atlas Peak DCA55:
2SD669=0.68V
MJE243G=0.70V
2SD612=0.70V
2SC495=0.72V
KSC3503=0.76V

YMMV,
Tom

I used Peak as well. I the picture I put the transistors in ascending order of Vbe I got. I measured 2 times. I am not sure if all the transistors here can be used anyway. My 2SD669 is from Ebay, the other transistors from Mouser and Digi-key.
2SC2690 = 0.66V
2N4923 = 0.68V
MJE243 = 0.70V
2SD669 = 0.72V
KSC3503 = 0.74V

transistors.jpg peak1.jpg
 
Hi my 2270 has the idle current problem. I have had this 2270 since 1978 when I bought it for $65 dollars
From an acquaintance. Now that I have replaced the caps and some transistors including the output on 1side. With the idle current at 18.4 to 18.9 on both channels and barely knowing anything about electronics.
I find there is no serial number plate on this unit. HOW DID I MISS THAT FOR SO LONG?
I ask does anyone know how to tell if this is a low s/n unit or higher?
Jerry
 
You can download the service manual from Hifiengine or AK database. I think both of these will have several pages with the different amp board configurations. If the bias transistor is round and held to the heatsink by a rolled brass plate, it's likely fairly early. But it's possible that the bias transistor was replaced so that could be misleading. You might post a photo of one of the amp boards because there are guys on this forum who can tell by looking.

You might also mention where you are measuring the bias current because it makes a difference if your serial number is lower than 3900.
 
For s/n 1001 through 3900, the NPN output is connected Base to J759 and Emitter to J760. Starting with s/n 3901 the NPN output is connected Base to J759 and Emitter to J761.
 
Back
Top Bottom