Sansui-AUX1/AU-X11 A968B Transistors

SanthoshA

Active Member
Dear All,

I am currently working on restoration for my AU-X11 and have sourced couple of replacement transistors for the AU-X11.
The replacements are for Tosbhiba 2SA968, and i got them from a source telling me that these are genuine.
I am quite afraid and reluctant to go with these spares.
Here are the pictures below:

1. Toshiba A968(Y Rating)
2. Toshiba A968B(O Rating) - Replacement
2SA968B.jpg
2SA968B_Rear.jpg
Few graphs from Peak Atlas DCA75:
I have overlayed the graphs of one good condition original transistor rating Y over the other replacement transistor with an O rating.

BJT: Ic Vs Vce
upload_2018-1-21_13-50-5.png
BJT: hFE Vs Vce
upload_2018-1-21_13-50-49.png
BJT: hFE Vs Ic
upload_2018-1-21_13-51-33.png
BJT: Ic Vs Vbe
upload_2018-1-21_13-51-51.png
BJT: Ic Vs Ib
upload_2018-1-21_13-52-7.png

Most of the graph data matches the replacement transistors i have sourced.
My knowledge is still not good enough to make out the data from these, hence seeking help from fellow memebers.
Can any one please confirm if this data is sufficient to prove that the replacements are genuine?
Or do the replacements look fishy to your eyes, please let me know your inputs.
 
Last edited:
They look good to me but it is difficult to be sure, install them - use a DBT and see how it goes - you are using sacrificial OP transistors correct?
 
They look good to me but it is difficult to be sure, install them - use a DBT and see how it goes - you are using sacrificial OP transistors correct?
I have the originals in working condition should I use the On Semi ones instead, I have them in stock.
I wanted to do a compare between the other non touched board and the one I have upgraded the resistors to metal film .
Comparing the original outputs in the un touched board with the On Semi will not make a good comparison.

Or should I rather change the outputs to On Semi,
Use the amp for couple of hours and if everything is ok I will put the original output transistors in again, just to be on the safer side .

Please let me know your inputs.
 
Please let me know your inputs.

Let me put it this way, I installed the sacrificial OP transistors in mine as soon as I could - and they didn't get replaced with the originals until all the major works on the rest of the amplifier were completed and tested to my satisfaction.

And when I re-installed the originals I STILL used the DBT until I knew everything was good.

I am R-E-A-L-L-Y puzzled that you are not using the sacrificial OP transistors already! - this makes NO sense at all. :confused:

;)
 
Let me put it this way, I installed the sacrificial OP transistors in mine as soon as I could - and they didn't get replaced with the originals until all the major works on the rest of the amplifier were completed and tested to my satisfaction.

And when I re-installed the originals I STILL used the DBT until I knew everything was good.

I am R-E-A-L-L-Y puzzled that you are not using the sacrificial OP transistors already! - this makes NO sense at all. :confused:

;)

I was yet to decide on the sacrificial outputs, the only reason was to check the sound signature change if any introduced by metal films with the untouched board(left channel). I wanted to do a comparison with right and left channels by changing the fader .

I would also have to remove the driver board again to remount the originals, but yeah I will use the On Semi’s to be on the safer side .
I will just remove the original outputs on the untouched channel and use On Semis to make a good comparison .

To summarise I will do it this way:
1. Change all outputs on all boards the left and right channel to On Semi (one channel uses metal film and the other is untocuhed)
2. Do a test run for couple of hours and if all is ok swap all outputs to originals and compare both left and right channels .

I wanted to negate the effort involved in the replacing the output transistors twice, but yeah it’s safer not to brick them, and I truly understand their value .

I would probably use the Kapton insulators with the On Semi’s to avoid messy thermal grease .
And the micca insulators with thermal grease for the original outputs when the amp is ready .


Thank you John !
 
Dear All,

I am currently working on restoration for my AU-X11 and have sourced couple of replacement transistors for the AU-X11.
The replacements are for Tosbhiba 2SA968, and i got them from a source telling me that these are genuine.
I am quite afraid and reluctant to go with these spares.
Here are the pictures below:

1. Toshiba A968(Y Rating)
2. Toshiba A968B(O Rating) - Replacement
View attachment 1094316
View attachment 1094317
Few graphs from Peak Atlas DCA75:
I have overlayed the graphs of one good condition original transistor rating Y over the other replacement transistor with an O rating.

BJT: Ic Vs Vce
View attachment 1094311
BJT: hFE Vs Vce
View attachment 1094312
BJT: hFE Vs Ic
View attachment 1094313
BJT: Ic Vs Vbe
View attachment 1094314
BJT: Ic Vs Ib
View attachment 1094315

Most of the graph data matches the replacement transistors i have sourced.
My knowledge is still not good enough to make out the data from these, hence seeking help from fellow memebers.
Can any one please confirm if this data is sufficient to prove that the replacements are genuine?
Or do the replacements look fishy to your eyes, please let me know your inputs.

I have used those very same replacements in my X1. Nothing has blown up yet. They are good.

Both Y and O rank should be OK to use. Just make sure you pair O with O or Y with Y.
(Needless to say. Just a bit of extra info for anyone out here who might not know.)

Could you share a link to the Kaptons? (Too busy / lazy right now to look myself.)
I am seriously considering going that route too with my X11 restoration...when I finally get around to working on it that is.
 
Last edited:
I have used those very same replacements in my X1. Nothing has blown up yet. They are good.

Both Y and O rank should be OK to use. Just make sure you pair O with O or Y with Y.
(Needless to say. Just a bit of extra info for anyone out here who might not know.)

Could you share a link to the Kaptons? (Too busy / lazy right now to look myself.)
I am seriously considering going that route too with my X11 restoration...when I finally get around to working on it that is.
Thanks for the information, Ron. !
This clears things, and i am now more confident with the replacements.
For the TO-3 Kapton Insulators, the part numbers are the following:

Mouser No: 567-175-6-310P
Mfr. No:175-6-310P
Mfr.:Wakefield-Vette
 
To answer your original question explicitly, no, the data is not sufficient to prove the parts are genuine. However, the data looks fine and you may be in luck. Please please please, use the sacrificial TO3s though :)
 
To eliminate any doubt, these are available from BD Enterprises, are absolutely authentic, and are excellent replacements for the original drivers you are looking for, so long as you do not need the tab for collector connection:

2SA1837
2SC4793
 
To eliminate any doubt, these are available from BD Enterprises, are absolutely authentic, and are excellent replacements for the original drivers you are looking for, so long as you do not need the tab for collector connection:

2SA1837
2SC4793
I actually need the tab for the collector, as these transistors are used in the driver board section.
Any reason why it is a good replacement without the collector tab(i hope i did not misunderstand what you tried to convey). Or is it because they act as regular diodes alone ?

Basically with the collector tab cut off the transistors behave like regular PN/NP junction diodes, this is how it works in the phono boards. I do believe they also do not take part in current amplification too .
Another thought process, which i still don't understand is why were diodes not used and Sansui decided to use a transistor with collector tab cut off, in certain areas of the amplifier.
 
Last edited:
To answer your original question explicitly, no, the data is not sufficient to prove the parts are genuine. However, the data looks fine and you may be in luck. Please please please, use the sacrificial TO3s though :)
Thanks for your inputs, i have mounted the sacrificial transistors to the heat sinks already. !
 
The original part has the option to use the tab, the center leg, or both for collector connection. I assumed since all 3 legs are present on the ones in your pics, you would not need the metal tab for the collectors electrical connection. The part I recommend does not have the metal tab option.

I am just recommending a good and proven to me replacement that sounds correct and performs as good as the original part. With an amplifier like this, you want the right part there.
 
For a replacement transistor with no exposed 'metal tab' (which is used as the collector connection in AU-X1/AU-X11/AU-919 etc driver boards) you can form the collector lead and solder it to a ring terminal secured with the transistor mounting screw - I did this on my AU-X1.

I can't remember if I included a picture of the modification in my AU-X1 restoration thread, but it is very easy to do.
 
Last edited:
To eliminate any doubt, these are available from BD Enterprises, are absolutely authentic, and are excellent replacements for the original drivers you are looking for, so long as you do not need the tab for collector connection:

2SA1837
2SC4793

I am no expert, but i believe that I should bring this up to you guys:
I asked about these parts to member ilimzn who is also very knowledgeable about restoring audio equipment. He has used these as drivers, and he thinks that for the AU-X1 they may be working near their limit but he did recommend this pair as drivers for the AU-919. However he further recommends sanding the plastic insulation on the back of the transistors to expose the heat sink of the transistor and use thermal grease. This is for the thermal tracking of the bias current. Please read post 15 here:

http://www.diyaudio.com/forums/solid-state/286544-sansui-au-919-transistor-replacement-2.html

There is another transistor that is quite available still and it has an Ic maximum of 2 amps: 2SC5171/2SA1930. I could get a bunch of these in a local shop and they do not appear to be fakes.

More interesting to me, is that there may be another way to connect the collector in these insulated transistors using the original screw for connecting the collector to the circuit. This would require sanding off the plastic on the top side to expose the heatsink (which is connected to the collector) and use the screw in the usual way with a locking washer. Please see post #16 here to see the top side of the internal heat sink on a pair of 2sa1837/2sc4793:

http://audiokarma.org/forums/index.php?threads/sansui-au-d907-needs-some-recaping.799000/

I have not yet tried this but I am about to go this way on an AU-D907 that came with a bad pair of driver transistors.
I enjoy this thread. Please post pictures of your work if you can. We can all learn from your experience.

Jose

EDIT:
I am sorry. Now I saw your thread about the AU-X11. Your drivers are not thermally coupled to the output transistors so there is no issue about the thermal tracking of the bias current in your case and you may go directly with the transistors TO-220F without sanding the plastic at all.
 
Last edited:
The original part has the option to use the tab, the center leg, or both for collector connection. I assumed since all 3 legs are present on the ones in your pics, you would not need the metal tab for the collectors electrical connection. The part I recommend does not have the metal tab option.

I am just recommending a good and proven to me replacement that sounds correct and performs as good as the original part. With an amplifier like this, you want the right part there.

I truly get your point and concern, i have ordered the replacements 2SA1837 and 2SC4793, from BD Enterprises.
Should take about a week to reach my place. !
Even i was quite worried, about the replacements hence the sole reason to start this thread.
The A968B replacement which i used has an (O) rating whereas the C2238B's are rated (Y).

A mix and match of gain ranking between the PNP pair and the NPN pair could also be a problem right ?
And since these are used in pre-driver section, if i am not wrong, they are quite important to the overall performance ?
Will messing up with gain rankings of PNP and NPN pair create a problem ?

The Y's are supposed to have a better Hfe ranking than the O's, a picture from the data sheet.
upload_2018-1-24_14-47-51.png
I will keep you posted on the updates, and will also post the data here comparing the original A968B's with the replacements from BD Enterprises.
I will have to put my driver board installation on hold, but i don't mind waiting to get things done properly.
 
The Y's are supposed to have a better Hfe ranking than the O's, a picture from the data sheet.

The hfe ranks overlap and you may get an O and a Y to have nearly the same hfe gain. Probably this is not the best thing to do but PNP and NPN often (but not always) have different hfe values even if they belong to the same gain rank. I noticed that in several TO-220 transistors like the ones in question, the hfe for the PNP type is higher than that for the NPN type.
Other part that you may be able to find locally are the sanyo 2SA1011/2SC2344 which are exactly like the 2SA968/2SC2238 or the and sanken 2SA958/2SC2168 , both are plain TO-220 with collector tab, and also the panasonic 2SA2140/2SC593 (TO-220F, insulated tab). These 3 pairs are all rated at 1.5A collector current, they have 60MHZ or higher fT and were discontinued after the 2SA968/2SC2238 which are being faked left right and center. I saw your pictures of the restoration. Very nice work and well documented. If I ever encounter an AU-X1 or X11 this is definitely one of the places to learn about these.:thumbsup:
Jose
 
Back
Top Bottom