• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Sony TA-N7B Transistors

Allvey

Allvey

New Member
Happy New Year!

I’m doing a complete overhaul to a Sony TA-N7B not only to increase its life span but also to improve its sound however apart from the bipolar transistors in the output stage (2SA473/2SC1173 to D45H8/D44H8) I’m not certain what other transistor should be replaced and by what (NOS or better grades?).

Could you please advise what transistors I should be looking to replace?

I’ve noticed a few mismatches between the schematics and the PCB’s, for example Q111: 2SC1962 on the schematics while 2SC1670 on the PCB’s, or Q151: 2SC1364 on the schematics while 2SC945 on the PCB’s, just to name a few.
Even though it’s likely the transistors on the PCB’s are valid equivalents shall I use as reference what’s on the schematics or on the PCB’s?

Big thanks.

Cheers,
Allvey
 
Register to hide this ad
Depends on how far you want to go.
The small signal stuff can be readily replaced by 2SA970/2SC2240, however there are parts I would not replace if it looks OK.
Others are a bit problematic due to case style and pin-out, but it can be done, The front end regulator pass transistors are very flexible, and they do not dissipate much power so almost anything decent in TO126 or even giant TO92 will work.
I'll get back to you on some of that via email, I'm having a very busy week extend into a busy week-end :(
 
ilimzn said:
Depends on how far you want to go.
The small signal stuff can be readily replaced by 2SA970/2SC2240, however there are parts I would not replace if it looks OK.
Others are a bit problematic due to case style and pin-out, but it can be done, The front end regulator pass transistors are very flexible, and they do not dissipate much power so almost anything decent in TO126 or even giant TO92 will work.
I'll get back to you on some of that via email, I'm having a very busy week extend into a busy week-end :(
Click to expand...

Thanks. I'll waiting be waiting for your email.
 
I restored two of those, the only transistors I replaced are the 2SA473/2SC1173 with d44h11/d45h11, as per ilimzn's advise. No other transistors are replaced and they both are dead quiet. First i used "nos" 2sa473/2sc1173 from littlediode. But I wasn't able to adjust the bias, now with the d44/d45 no problem at all. Both are running at 80-85mv.
 
The service manual states 30-70mv. Mirko advises to go a little bit higher. I have to look up his thread. It is not sounding creamy at all, but very tight and transparent, and I am running it on a ta-e88b
 
Harolda said:
The service manual states 30-70mv. Mirko advises to go a little bit higher. I have to look up his thread. It is not sounding creamy at all, but very tight and transparent, and I am running it on a ta-e88b
Click to expand...

...
 
Last edited:
ilimzn said:
Depends on how far you want to go.
The small signal stuff can be readily replaced by 2SA970/2SC2240, however there are parts I would not replace if it looks OK.
Others are a bit problematic due to case style and pin-out, but it can be done, The front end regulator pass transistors are very flexible, and they do not dissipate much power so almost anything decent in TO126 or even giant TO92 will work.
I'll get back to you on some of that via email, I'm having a very busy week extend into a busy week-end :(
Click to expand...

OK
:lurk:
 
Sorry for being absent, job demands.
There is a number of replacements one CAN make but not necesairly that one MUST make, if the original components are within spec.
Q101, 102 and 109 can be replaced with 2SA970
Q106 can be replaced by 2SC2240
Q107 and 108 make up a reference voltage generator which is dependant on transostor Vbe. This sets up the tail current of the input FETs but the variation will not be huge so nearly any decent NPN part will work. 100V Vcemax will be enough. Of course, 2SC2240 can be used.
Q163/164 also 2SC2240 but could just as well be more general purpose transistors.
Q167/168 also 2SA970 but again, could be more general purpose
Q161/162 are used as current sources to set up the zener current for D161/2. This is not overly critical but at least needs to be in the ballpark. The thing is, the only way to know for sure is to measure the voltage drop on R167/162 and calculate the current.
Q103/104 could be almost any JFET with Idss higher than the input stage tail current. This is actually surprisingly low at only 1mA, but this was chosen for minimum noise from the input JFET. All the usual suspects could be used here but things like 2SK170 et al should be avoided - high capacitance. 2SK107 I think is what I used. Contrary to intuition, these do not need to be well matched. Any difference will just translate to a small difference in drain voltage of the input JFET pair, which does not amount to anything significant.
Q144-147 could literally be any old general purpose small signal transistors, taking into account the pinout. These are the overcurrent protection circuit. 2SA1015/2SC1815 will do just fine. They only see a couple of volts across C-E under normal circumstances.
Q163/168 are more difficult to replace. Even though almost anything with Icmax=0.5A or more and Ucemax=80V or more in a TO126 or larger case will work, the problem is the pinout, which is EBC rather than the traditional BCE or ECB. There are painfully few that fit and have the same pinout, in the low numbered MPS series, some older Toshiba 2SA/SC and BF7xx series. Sony has their own, 2SA835/2SC1663. However, all of these are very difficult to get, which is why even in the original EU/US versions Sony used TO220 transistors with legs bent around in a rather ugly fashion. MJE340/350 or BD139/140 would work just as well. The regulators only work with a few mA, so there is no need for additional heatsinking - but the pinout problem remains.
Q110/111 can have the same problem in reverse. In some cases the originals are EBC bent into a BCE shape. 2SB649/2SD669 will work here, ECB pinout is compatible with the board.
However, Q130/131 are more difficult to replace. Again the same problem, EBC pinout and TO202 case with center heatsinking fin, all of which are counted on by the PCB layout. Again, 2SB649/2SD669 would work just fine but for the pinouts. If these need replacing (failed due to output stage failure) it is best to use Q110/111 in theor place if the EBC variant was fitted there, and replace Q110/111 by the recommended replacements above.
Regarding the output stage, D44H11/D45H11 are excellent. There are some Japanese replacements that could work but are difficult to get and therefore much more expensive. D44/45/H11 have the bonus of extremely flat current gain, which is as such the best one can get for a 10A transistor (and probably better than many smaller ones!) and this is something the output stage can really benefit from. That being said, the same pair is not so good for non-switching applications as a bare BJT, and as many has great speed and beta but atrocious SOA ratings. Fortunately, the hybrid BJT/VFET combo caters for exactly that kind of problem and in this application it is perfectly fine - MUCH more so than the originals.
Finally, there is the input dual JFET and this is quite difficult to replace with anything but itself, sadly long obsolete. More modern ones may not be suited because they are high gm units designed to operate at much higher currents. So far I have been lucky and no degraded parts so no replacement was needed.
 
I know it is an old thread but I hope you v-fet guys are still around. I got my ta-n7b restored by my tehnician and he used MJE15030G/15031G pairs on output stage. I would like to hear your opinion on that. Amplifier works nicely and bias is set on 50mV.
 
One of my N7 has a strange crackle in the right channel, so I checked the DC Bias and it was way of: Left Channel 21mV and Right Channel 0,0mv and no way to adjust the Bias correctly. The R-CH stays at 0,0mV no matter what, L-CH stops at max 21 mV and can only be adjusted lower but not to the recommended 30-70mV.
so could the 2SA473/2SC1173 be the problem?
 
Last edited:
kentlabs said:
One of my N7 has a strange crackle in the right channel, so I checked the DC Bias and it was way of: Left Channel 21mV and Right Channel 0,0mv and no way to adjust the Bias correctly. The R-CH stays at 0,0mV no matter what, L-CH stops at max 21 mV and can only be adjusted lower but not to the recommended 30-70mV.
so could the 2SA473/2SC1173 be the problem?
Click to expand...

Too little bias is most likely the traditional fading 2SA473/2SC1173.
0V bias, however, is more worrisome, which sounds the bias stage is out of service all together.

Besides those cascode transistors, some of fusible resistors are also "fading" out, in the sense that they increase in value.
Notably the ones on the VFET socket boards are prone to that effect, and should be replaced while the cascodes trannies are replaced.
Dale CMF55 will do fine (yes they are fusible type )

Besides fading fusible resistors, you could in your case also have a blown fusible resistors.
The driver trannies have already 3 of the them around...
With one down, the bias circuit won't be able to get anything up there. Then again, with one down yo
Hence, replacing all the fusible resistors (with correct 0.25W or 0.5W rating) is another thing that always should be done sooner or later in the N7B.
 
damn, the N7 has been in service for a full restoration not long ago. They must have really worked hard for the money....
Thanks Oilmaster, I will check all the fusible resistors next and will see what else they have "overlooked".
 
The sony fusable resistors may check cold okay but can go open as read many many on sony threads. Might consider better bias VR pots if the cheapie plastic to bournes. many rebuilders swap out and better registration.
hope it helps.
bink
 
thanks Binkman, yeah already read about that when searching AK for advice. I'm going to replace them with Bourns 50K Pots as well as all the fusible resistors, the BJT's to
D44H11/D45H11 and the the main filter caps to 27,000uf - 80v Nippon Chemicon ones. anything else I should pay attention to? I am trying to put a list of parts together for a complete restoration. I came across what is called Mirko's Kit for the N7 and was wondering if someone would share some info on that since Mirko doesn't seem to be around at AK anymore.
 
Last edited:
I completely recapped the unit and replaced the double vd2121 diodes next to the things you mention
 
You must log in or register to reply here.
Back
Top Bottom