The sound from the right channel of my Sherwood SEL-400 amp dropped out suddenly and had a "hissy" fit, no smoke but I did smell something cooking when I opened up the amp and noticed that a very faint smell seemed to come from the Right channel output transistors.
The schematic indicates that output the transistors are SJ1936 & SJ1937, these are present in the Left Channel and they each have a "blue dot".
The Right Channel has Motorola SJ1902 & Texas Instrument SJ1903 which are obviously replacements so the amp must have had this problem before.
I pulled the outputs to test them and this is what the PEAK Atlas showed. (Three of them read what is indicated with an asterik*, the fourth does not). My guess is that the Right Channel transistor Texas Instrument SJ1903 is faulty since it is the odd man out. Can someone please confirm this?
Right Channel:
Motorola SJ1902
NPN Darlington Transistor
*Diode Protection between C-E
*Resistor Shunt between B-E
*hfe not accurate due to B-E res
Current gain hfe=27
Test Current Ic=2.50mA
Base-Emitter V Vbe=0.860V
Test Current Ib=4.096mA
Leakage Current Ic=0.000mA
Texas Instrument SJ1903
PNP Darlington Transistor
Current gain hfe=34
Test Current Ic=2.50mA
Base-Emitter V Vbe=1.172V
Test Current Ib=3.785mA
Leakage Current Ic=0.000mA
Left Channel:
Motorola SJ1936
NPN Darlington Transistor
*Diode Protection between C-E
*Resistor Shunt between B-E
*hfe not accurate due to B-E res
Current gain hfe=22
Test Current Ic=2.50mA
Base-Emitter V Vbe=0.807V
Test Current Ib=4.148mA
Leakage Current Ic=0.000mA
Motorola SJ1937
PNP Darlington Transistor
*Diode Protection between C-E
*Resistor Shunt between B-E
*hfe not accurate due to B-E res
Current gain hfe=23
Test Current Ic=2.50mA
Base-Emitter V Vbe=0.841V
Test Current Ib=4.113mA
Leakage Current Ic=0.000mA
The schematic indicates that output the transistors are SJ1936 & SJ1937, these are present in the Left Channel and they each have a "blue dot".
The Right Channel has Motorola SJ1902 & Texas Instrument SJ1903 which are obviously replacements so the amp must have had this problem before.
I pulled the outputs to test them and this is what the PEAK Atlas showed. (Three of them read what is indicated with an asterik*, the fourth does not). My guess is that the Right Channel transistor Texas Instrument SJ1903 is faulty since it is the odd man out. Can someone please confirm this?
Right Channel:
Motorola SJ1902
NPN Darlington Transistor
*Diode Protection between C-E
*Resistor Shunt between B-E
*hfe not accurate due to B-E res
Current gain hfe=27
Test Current Ic=2.50mA
Base-Emitter V Vbe=0.860V
Test Current Ib=4.096mA
Leakage Current Ic=0.000mA
Texas Instrument SJ1903
PNP Darlington Transistor
Current gain hfe=34
Test Current Ic=2.50mA
Base-Emitter V Vbe=1.172V
Test Current Ib=3.785mA
Leakage Current Ic=0.000mA
Left Channel:
Motorola SJ1936
NPN Darlington Transistor
*Diode Protection between C-E
*Resistor Shunt between B-E
*hfe not accurate due to B-E res
Current gain hfe=22
Test Current Ic=2.50mA
Base-Emitter V Vbe=0.807V
Test Current Ib=4.148mA
Leakage Current Ic=0.000mA
Motorola SJ1937
PNP Darlington Transistor
*Diode Protection between C-E
*Resistor Shunt between B-E
*hfe not accurate due to B-E res
Current gain hfe=23
Test Current Ic=2.50mA
Base-Emitter V Vbe=0.841V
Test Current Ib=4.113mA
Leakage Current Ic=0.000mA
Last edited:
Googling SJ1936 & SJ1937, I found this website which suggests that the NTE249 / NTE250 are appropriate substitutes.
Next question, can this information be trusted?
https://vetco.net/products/npn-si-darlington-transistor-100v-16a-comp-nte250-nte249
Next question, can this information be trusted?
https://vetco.net/products/npn-si-darlington-transistor-100v-16a-comp-nte250-nte249
Data sheets are available for the NTE249 & NTE250 but not for the SJ1936 & SJ1937.
Revising my question, can I trust www.vetco.net to be a reliable source?
Revising my question, can I trust www.vetco.net to be a reliable source?
Vetco might be reliable but NTE is not what you want in your amp, unless there is absolutely no other solution.
Here's the service manual
Hugo
Here's the service manual
Hugo
The current gain values will probably be far off if measured with the PEAK.
Edit: the PEAK is good for small signal transistors due to it's voltage and current limitations.
Here is a thread proposing replacements, which seem OK to me:
https://audiokarma.org/forums/index...00-output-transistor-subs-equivilants.282181/
Mouser has the 2N6059 and the 2N6052 but they are utterly expensive.
BDW83C and BDW84C can be used too, but will be hard to find.
Furthermore, I recommend to only buy from reputable sources.
Hugo
Edit: the PEAK is good for small signal transistors due to it's voltage and current limitations.
Here is a thread proposing replacements, which seem OK to me:
https://audiokarma.org/forums/index...00-output-transistor-subs-equivilants.282181/
Mouser has the 2N6059 and the 2N6052 but they are utterly expensive.
BDW83C and BDW84C can be used too, but will be hard to find.
Furthermore, I recommend to only buy from reputable sources.
Hugo
I saw that too but keep in mind that Newark / Farnell is a business to business company.
You need a VAT number.
Hugo
You need a VAT number.
Hugo
If that s 100V/16A transistors then that should be good, otherwise there s the BDX66B/67B that are 100V/16A
that are TO3 as well, at worse and depending of the amp power you can use their TOP3 equivalent BDV64C/65C.
that are TO3 as well, at worse and depending of the amp power you can use their TOP3 equivalent BDV64C/65C.
Not in US. No VAT number needed, they sell to anybody with a pulse.
Bought stuff from them many times.
In the early days (before the internet) all the big distributors required a tax ID. Mouser and Digikey didn’t but they were small potatoes with limited stock in those days. The first one that carried real audio transistors who sold to anyone with a pulse was MCM (spin off of Newark, now Re-absorbed).
What are the specifications that I need to pay attention to when trying to find substitute transistors?
Here's a data sheet for the NTE249 / NTE250 transistors
https://4donline.ihs.com/images/Vip...9-1.pdf?hkey=6D3A4C79FDBF58556ACFDE234799DDF0
Here's a data sheet for the NTE249 / NTE250 transistors
https://4donline.ihs.com/images/Vip...9-1.pdf?hkey=6D3A4C79FDBF58556ACFDE234799DDF0
Aside from just the basic ratings, which any TO-3 will do you need to pay attention to the secondary breakdown breakpoint. You need a transistor that is happy on +/-50 V supplies. IOW, it must handle a useful amount of current at 50V. MJ11016 is on the edge of useability, 2N6284 is not suitable. Either is fine on +/-40. MJ11022 handles full power to 50V, which is far safer. They are $9 and the PNP is out of stock at Mouser. All of these are available in TO-247 (MJH package), and I suspect the TO-3 packaging is on its way OUT (keep eyes peeled for EOL notices). They are getting stock from ON now, so you may be able to get the 11021 in your lifetime.
Do NOT mix families or brands between an NPN and PNP pair. Mixing voltage grades is fine. The B-E resistor values need to be the SAME between a pair unless you plan on running the bias hot enough to overcome differences and force high current into both outputs. Even then the distortion will be higher than if they had the same resistors.
Do NOT mix families or brands between an NPN and PNP pair. Mixing voltage grades is fine. The B-E resistor values need to be the SAME between a pair unless you plan on running the bias hot enough to overcome differences and force high current into both outputs. Even then the distortion will be higher than if they had the same resistors.
Those are what I would be using. Power handling in all the other suggested types falls off with voltage too soon.