• 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

Pioneer SA-7500 II power stage - transistors replacement

M

Moncho

Member
Hi guys,

I'm (since yesterday) the proud owner of a "Pioneer SA-7500 II". The face plate and general condition is really good, but it doesn't sound fine. The left Channel is dead and the right has a lot of "hiss" (like white noise), so I opened.
Inside I found that the right channel was "repaired", one power resistor is different to the originals and also the power transistor were changed (Q14 and Q16).

I'll measure all important voltages and at the same time try to identify the most relevant "issues". Anyway, as the power section has been modified I want to change all e-caps and transistors in it, so they "match". I have no experience replacing transistors, I've found some options (in stock) by my self and have also taken some recommendations from other threads here in AK.
So, could you please give me some feedback related to the replacements? Are they ok? Do I have to pay extra attention to certain parameters?

PS: I won't bother to find the original packaging, I just need the proper replacement and then I'll found a way to mount them.

Thanks !!!

Q13 / Q14 (PNP)Q15 / Q16 (NPN)
OriginalOption1 (Digikey)Option2 (mouser)OriginalOption1 (Digikey)Option2 (mouser)
2SB618AMJL1302AGMJE15031G2SD588AMJL3281AGMJE15030G
Pc [W]8020050Pc [W]8020050
Vcb [V]150260150Vcb [V]150260150
Vce [V]120260150Vce [V]120260150
Veb max [V]555Veb max [V]555
Ic continuous [A]No info158Ic continuous [A]No info158
Ic max [A]72516Ic max [A]72516
TJ [°C]150150150TJ [°C]150150150
ft [MHz]73030ft [MHz]83030
hFE min607540hFE min407540


Q9 / Q10 (PNP)Q11 / Q12 (NPN)
OriginalOption (Digikey)OriginalOption (Digikey)
2SB527KSC2690AYS2SD357KSA1220AYS
Pc: Ta/Tc [W]0.75 / 101.2 / 20Pc [W]0.75 / 101.2 / 20
Vcb [V]110160Vcb [V]110160
Vce [V]80160Vce [V]100160
Veb max [V]65Veb max [V]55
Ic continuous [A]No info1.2Ic continuous [A]No info1.2
Ic max [A]22.5Ic max [A]0.82.5
TJ [°C]130150TJ [°C]130150
ft [MHz]35155ft [MHz]35175
hFE min5060hFE min5560


Q5 / Q6 (NPN)Q7 /Q8 (PNP)
OriginalOption (Digikey)OriginalOption (Mouser)
2SC1438KSA992FATA2SA857KSC1845FTA
Pc [W]0.50.5Pc [W]0.50.5
Vcb [V]150120Vcb [V]150120
Vce [V]120120Vce [V]150120
Veb max [V]55Veb max [V]55
Ic continuous [A]No infoNo infoIc continuous [A]No infoNo info
Ic max [A]0.050.05Ic max [A]0.050.05
TJ [°C]150150TJ [°C]125150
ft [MHz]65100ft [MHz]100100
hFE min150150hFE min150150
 
Last edited:
Register to hide this ad
To replace those "bat-wings" transistors, I usually go for the mje15032/15033 (or 15031/15030 in your case), they are easier to fit, even if the package is only rated for 50W instead of the 80W of the original parts. We need to do some math based on the supply voltages before confirming that the transistors won't be stressed over 50W in that application.

The drivers and smaller transistors look OK to me. Keep in mind that the differential pairs (those transistors face to face at the power amp input) need to be matched pairs.
 
elnaldo said:
To replace those "bat-wings" transistors, I usually go for the mje15032/15033 (or 15031/15030 in your case), they are easier to fit, even if the package is only rated for 50W instead of the 80W of the original parts. We need to do some math based on the supply voltages before confirming that the transistors won't be stressed over 50W in that application.

The drivers and smaller transistors look OK to me. Keep in mind that the differential pairs (those transistors face to face at the power amp input) need to be matched pairs.
Click to expand...
Hi Enaldo,
Thanks a lot for your feedback.
The transistors that you usually use rated to 50W ... which math do you do to avoid stress over 50W in that application? The Amp is rated to 45W per channel at 8Ohms and 50W per channel at 4 Ohms, I doubt that I'll crank it up to the top using 4 Ohm speakers. In any case, if there is nothing wrong with the other transistor that I have found, then I could just use those and forget about any other restriction, thus they are rated to 200W they also handle more voltage and current keeping "Veb max of 5V". The power supply will remain the same, so the amp will not extract 200W out of them, but I still have to avoid any type of overheat". Am I right?

In relation to the drivers, I know that the have to "match", in fact there is thread here in AK, where @EngineerNate posted a pdf file, from someone called "Ian Fritz", with a pretty nice explanation and method to match transistors. These are fairly "cheap", so I'll just use brute force, buy 10 of each and match them.

Once again, thanks a lot for your reply and Hope to be able to repair this beautiful looking amp :)
 
If the amp is rated to max. 50W/channel at 4 ohm, I'd say that the 50W device should be safe to use. As you say, probably you will never push the amp to the limits.

The other transistor, the only thing I see wrong is the package. they are TO-3P or TO-247, and they will be harder to fit, the footprint is different, thicker leads too, but if you find a way to do a polite job installing them, of course they are an extra insurance. There are 75 or 80W transistors in to-220 packages, but usually those are not designed or suited for audio application and they are not as linear as the mje15030.

Another thing to keep in mind, the replacements have a much higher bandwidth, up to 30MHz vs 7MHz. That could introduce an oscillation in some amps, specially in old designs. It never happened to me, but it could happen. It's easy to solve adding a small ceramic capacitor, but probably the circuit is already prepared to kill any oscillation. I need to check the schematic again.

Regarding the matched pairs, I see in the schematic that your amp uses the 2SA798 double transistors. Those must be replaced with matched pairs of ksa992. There is a long thread at AK about replacing the 2sa798, since they always fail. The rest don't need to be matched parts.
 
Last edited:
Edit: sounds like, from later posts here, the following recommendation is not the best for the original poster's situation. Read careful before following any of my advice here for this unit.

A company (one person, really... a fellow vintage stereo repair person) named Hoppe's Brain makes adapters including one that allows you to choose a modern, available TO-3P or TO-247 device and mount it where a MT-200 device was originally used. Quite clever, seems to work for many well, and allows better selection of replacement devices.

hoppesbrain.com

MT-200 to TO-3P / TO-247 Transistor Heat-Spreaders / Adapters

MT-200 to TO-3P / TO-247 Transistor Adapters / Heat-Spreaders
hoppesbrain.com hoppesbrain.com

Hoppes Brain MT-200 adapter 20240223_163720.png

No affiliation with the vendor other than I think their products are cool... and made by "us" for "us".
 
Last edited:
I would only change what’s blown but change with complementary pairs.
I do not believe the Hitachi batwing is the same package as the sanken mt-200?
Imo for the outputs I recommend the Toshiba 2sa1943n/2sc5200n.
For drivers Toshiba have tta004b-q/ttc004b-q
Also this thread might get more responses in the pioneer section
 
dlucy said:
A company (one person, really... a fellow vintage stereo repair person) named Hoppe's Brain makes adapters including one that allows you to choose a modern, available TO-3P or TO-247 device and mount it where a MT-200 device was originally used. Quite clever, seems to work for many well, and allows better selection of replacement devices.

hoppesbrain.com

MT-200 to TO-3P / TO-247 Transistor Heat-Spreaders / Adapters

MT-200 to TO-3P / TO-247 Transistor Adapters / Heat-Spreaders
hoppesbrain.com hoppesbrain.com

View attachment 3301271

No affiliation with the vendor other than I think their products are cool... and made by "us" for "us".
Click to expand...
I think the original transistors are not that big. They are those kind of To-220 with lateral metal wings.
 
elnaldo said:
If the amp is rated to max. 50W/channel at 4 ohm, I'd say that the 50W device should be safe to use. As you say, probably you will never push the amp to the limits.

The other transistor, the only thing I see wrong is the package. they are TO-3P or TO-247, and they will be harder to fit, the footprint is different, thicker leads too, but if you find a way to do a polite job installing them, of course they are an extra insurance. There are 75 or 80W transistors in to-220 packages, but usually those are not designed or suited for audio application and they are not as linear as the mje15030.

Another thing to keep in mind, the replacements have a much higher bandwidth, up to 30MHz vs 7MHz. That could introduce an oscillation in some amps, specially in old designs. It never happened to me, but it could happen. It's easy to solve adding a small ceramic capacitor, but probably the circuit is already prepared to kill any oscillation. I need to check the schematic again.

Regarding the matched pairs, I see in the schematic that your amp uses the 2SA798 double transistors. Those must be replaced with matched pairs of ksa992. There is a long thread at AK about replacing the 2sa798, since they always fail. The rest don't need to be matched parts.
Click to expand...
Hi Enaldo,

thanks again. In fact you just nailed it. I just measured Q3 and Q4 -> 2SA798 and the voltages aren't ok, I read; almost 0V in each base but 1V positive in the common emitter, the collectors are ok at -47V and it should be Base 0V, Emitter -0.6V and collectors -47V (as in the picture).
Sorry if I'm bit slow, but you are saying that each 2SA798 will be replaced by a pair of ksa992. All of the new ksa992 must be a close to each other as possible AKA match? So there will be no discrepancy inside each "new 2SA798" and neither between channels? and There is no need to match any other transistor?

Thanks again :)
 
rcs16 said:
I would only change what’s blown but change with complementary pairs.
I do not believe the Hitachi batwing is the same package as the sanken mt-200?
Imo for the outputs I recommend the Toshiba 2sa1943n/2sc5200n.
For drivers Toshiba have tta004b-q/ttc004b-q
Also this thread might get more responses in the pioneer section
Click to expand...
Hi rcs16,

thanks a lot for your transistors recommendation, they are in stock !!!!
I'm sorry if I posted in the wrong place, I have been an active member since a week or so and I'm still confuse with all the available options and thought ... hmmm ok, my question is related to transistors, soo here it goes :)
 

Double Transistor 2SA798: Is There a Replacement?

I have been working out an issue on a Pioneer SA 9500II and on the amp board there is a double PNP transistor, a 2SA798. I have removed it and replaced it on the boards a total of four times now and it is getting a little beat looking, to the point that I fear to have to test it or remove it...
audiokarma.org

You need a matched pair at each location. you don't need to match between different pairs. They will be close enough if you buy from the same batch.

In my experience, the 2sa798 will become noisy sooner or later. That's why I replace them in my repairs. But if it's your unit, you could leave them alone until they fail, what could happen in months, years, or never in your life.
 
A TO-3P transistor should also work well for the application, if you don't mind drilling and tapping new holes in the heatsink. The TO-3P has a similar copper heat-spreader surface-area to the BatWing, so no problem there.

I wouldn't be so much worried about a higher Ft, as long as the amp is compensated well. However, the NJW1302/3281 have large silicon dies, and pretty high Cob and Cib, which also affects the response and compensation required by the amp. I would consider a lower capacitance TO-3P like the ST 2SD1047.

Square wave testing into a difficult load like a 2-ohm resistor in series with a 1uF cap should reveal any instability. (Keep it under 10V P-P) You can experiment with compensations caps and see what happens.
 
Phloodpants said:
A TO-3P transistor should also work well for the application, if you don't mind drilling and tapping new holes in the heatsink. The TO-3P has a similar copper heat-spreader surface-area to the BatWing, so no problem there.

I wouldn't be so much worried about a higher Ft, as long as the amp is compensated well. However, the NJW1302/3281 have large silicon dies, and pretty high Cob and Cib, which also affects the response and compensation required by the amp. I would consider a lower capacitance TO-3P like the ST 2SD1047.

Square wave testing into a difficult load like a 2-ohm resistor in series with a 1uF cap should reveal any instability. (Keep it under 10V P-P) You can experiment with compensations caps and see what happens.
Click to expand...
I think I'll just drill the plate to make a new TO-3P fit.
Looking the datasheet for the ST 2SD1047, I don't find his "complementary" PNP? Does it exist? I mean, does ST produce a par NPN/PNP complementary transistors where the NPN is the one that you recommend?
Another thing, if I test it and any stability appears, where should I put a cap to compensate? which type of cap (ceramic)?
 
Yeah, the 2SB817 is the compliment. These are not produced anymore. I only mention it as an example of a low-Cob transistor, I don't have any actual experience with this Kenwood. The compensation caps already exist in the design, and their values can be adjusted. Usually they put in ceramic caps for compensation, which works fine, but if you want to experiment with better caps, silver-mica or high-spec ceramics like NPO/GOG types are also excellent.

JohnAudioTech has a good video on this!

Also
 
Yeah, the 2SB817 is the compliment. These are not produced anymore. I only mention it as an example of a low-Cob transistor, I don't have any actual experience with this Kenwood. The compensation caps already exist in the design, and their values can be adjusted. Usually they put in ceramic caps for compensation, which works fine, but if you want to experiment with better caps, silver-mica or high-spec ceramics like NPO/GOG types are also excellent.
Phloodpants said:
JohnAudioTech has a good video on this!

Also
Click to expand...
Thank, I'll watch the videos.
I'm thinking about to use the original holes in the heatsink to fit a "clamping" mechanism. A single clamp should be enough to hold the TO-3P package in position and I don't need to drill or tap anything.
 
Hi guys,

just to close this thread I want to say, that after a lot of changes, finally I've only replaced the differentials Q3 and Q4, each 2SA798 for 2xKSA992 and the output transistors Q13 to Q16 using 2SA1943N/2SC5200N. Also a bunch of diodes and resistors but nothing complicated and only as a preventive action.

The design of this amp is not flexible at all and any little change makes also a huge change in the bias current, which is also not adjustable.
In general I'm happy how it sounds and it works according to the manufacturer specifications. Anyway I want to add the possibility to adjust the DC Offset and a bit of bias current adjustment, to do so I've opened this other thread and I already have some kind of solution but I haven't had any time to put all together and test it.

I just want to thank you guys and wish you a happy listening :)
 
I arrived at this post hoping to confirm my substitution choice for 2SC1438.
The original poster, Moncho, was asking if his subs were acceptable.
I saw cross references substituting NPN for PNP and vice-versa.
There seems to be no mention of this in the responses.
Am I missing something here?
(I realize I am late to this posting)
Larry

all PNP - OK
Q13/14
Original PNP 2SB618A
Option 1 PNP MJL1302AG
Option 2 PNP MJE15031G

all NPN - OK
Q15/16
Original NPN 2SD588A
Option 1 NPN MJL3281AG
Option 2 NPN MJE15032G

INCORRECT
Q9/10
Original PNP 2SB527
Option NPN KSC2690AYS

INCORRECT
Q11/12
Original NPN 2SD537
Option PNP KSA1220AYS

INCORRECT
Q5/6
Original NPN 2SC1438
Option PNP KSA992FATA

INCORRECT
Q7/8
Original PNP 2SA857
Option NPN KSC1845FTA
 
lebrun53 said:
I arrived at this post hoping to confirm my substitution choice for 2SC1438.
The original poster, Moncho, was asking if his subs were acceptable.
I saw cross references substituting NPN for PNP and vice-versa.
There seems to be no mention of this in the responses.
Am I missing something here?
(I realize I am late to this posting)
Larry

all PNP - OK
Q13/14
Original PNP 2SB618A
Option 1 PNP MJL1302AG
Option 2 PNP MJE15031G

all NPN - OK
Q15/16
Original NPN 2SD588A
Option 1 NPN MJL3281AG
Option 2 NPN MJE15032G

INCORRECT
Q9/10
Original PNP 2SB527
Option NPN KSC2690AYS

INCORRECT
Q11/12
Original NPN 2SD537
Option PNP KSA1220AYS

INCORRECT
Q5/6
Original NPN 2SC1438
Option PNP KSA992FATA

INCORRECT
Q7/8
Original PNP 2SA857
Option NPN KSC1845FTA
Click to expand...
Hi,

I'm so sorry but I can't help you. As I've wrote, I did change and try some transistors, but I couldn't make it work.
I wish you luck and if you succeed, please let me know.

Have a nice weekend :)
 
You must log in or register to reply here.
Back
Top Bottom