Gentlemen,
Thanks for the help over the years I have recently wanted to revisit an old idea. The J2 but with substitute J-FET outputs.
I have two 20pcs boxes of United SIC Power JFETS I bought a few years ago. I never got around to building any jig due to my mothers passing and lifes difficult duties so I dug the boxes out again. The devices in the boxes are:
UJ3N120065K3S
UJ3N065080K3S
is any of these any good to try anymore for a 10-15 watt SE design? How can I mod the circuit to work? I am a SMT technician; not an engineer.
You might laugh but I asked ChatGPT and Google Gemini on deepmode for possible subs and maybe I didn't ask the question right it sounds too optomistic!
"Find replacement for SemiSouth SJEP120R100 for use in First Watt J2 see schematic attached."
Recommendation:
For the Q5 and Q6 positions, the following conclusions and recommendations are provided:
Both UJ3N120065K3S and UJ3N065080K3S are suitable replacements for the SJEP120R100 in the First Watt J2:
If thermal headroom and lower R<sub>DS(on)</sub> are important, especially at higher idle currents, UJ3N120065K3S is the better choice.
Let me know if you want SPICE models, bias calculations, or help simulating these parts in the J2 circuit.
is any of this true and/or accurate?
Thanks,
Jeff
Thanks for the help over the years I have recently wanted to revisit an old idea. The J2 but with substitute J-FET outputs.
I have two 20pcs boxes of United SIC Power JFETS I bought a few years ago. I never got around to building any jig due to my mothers passing and lifes difficult duties so I dug the boxes out again. The devices in the boxes are:
UJ3N120065K3S
UJ3N065080K3S
is any of these any good to try anymore for a 10-15 watt SE design? How can I mod the circuit to work? I am a SMT technician; not an engineer.
You might laugh but I asked ChatGPT and Google Gemini on deepmode for possible subs and maybe I didn't ask the question right it sounds too optomistic!
"Find replacement for SemiSouth SJEP120R100 for use in First Watt J2 see schematic attached."
Depletion-Mode SiC Devices
The First Watt J2 amplifier’s Q5/Q6 stage originally used SemiSouth SJEP120R100 SiC JFETs (now discontinued). We surveyed active, true depletion-mode (normally-on) JFETs in Silicon Carbide (SiC) to identify suitable substitutes. The table below summarizes each candidate’s key specs; further details and application notes follow.Silicon Carbide (SiC) JFETs
onsemi (formerly SemiSouth/Soitec) offers a line of “EliteSiC” normally-on JFETs (650–1700V)mouser.com. Key parameters are extracted from their datasheets. All devices require a negative gate bias to set conduction, and exhibit positive temperature coefficient (ease of paralleling).| Manufacturer | Part No. | V<sub>DS</sub> (V) | I<sub>D</sub> (A, T<sub>C</sub>=25°C) | R<sub>DS(on)</sub> (mΩ, @25°C) | V<sub>GS(th)</sub> (V) | C<sub>iss</sub> (pF) | C<sub>rss</sub> (pF) | C<sub>oss</sub> (pF) | Package | R<sub>θJC</sub> (°C/W) |
|---|---|---|---|---|---|---|---|---|---|---|
| onsemi (EliteSiC) | UF3N120007K4S | 1200 | 120 | 7.1 (typ) | –9.3 to –4.7 (typ ~–7) | 8110 (at 800V) | 358 (at 800V) | 368 (at 800V) | TO-247-4L | 0.15 (typ) |
| onsemi (EliteSiC) | UJ3N065025K3S | 650 | 85 | 25 (typ) | –14 to –6 (typ ~–11.5) | 2360 (at 100V) | 282 (at 100V) | 290 (at 100V) | TO-247-3 | 0.26 (typ)onsemi.com |
| onsemi (EliteSiC) | UJ3N065080K3S | 650 | 32 | 80 (typ) | –14 to –6 (typ ~–11.5) | 630 (at 100V) | 88 (at 100V) | 94 (at 100V) | TO-247-3 | 0.61 (typ)onsemi.com |
| onsemi (EliteSiC) | UJ3N120035K3S | 1200 | 63 | 35 (typ) | – (not explicitly given) | – | – | – | TO-247-3 | 0.27 (typ) |
| onsemi (EliteSiC) | UJ3N120065K3S | 1200 | 34 | 66 (typ) | –9.3 to –4.7 (typ ~–6.6) | 1008 (at 100V) | 95 (at 100V) | 100 (at 100V) | TO-247-3 | 0.45 (typ)onsemi.com |
| onsemi (EliteSiC) | UJ3N120070K3S | 1200 | 33.5 | 70 (typ) | –14 to –6 (typ ~–11.5) | 985 (at 100V) | 95 (at 100V) | 100 (at 100V) | TO-247-3 | 0.45 (typ)onsemi.com |
| onsemi (EliteSiC) | UF3N170400B7S | 1700 | 6.8 | 400 (typ) | –11.3 to –6.7 (typ ~–9) | 225 (at 100V) | 18 (at 100V) | 22 (at 100V) | D2PAK-7L | 1.7 (typ)onsemi.com |
| onsemi (EliteSiC) | UJ4N075004L8S | 750 | 120 | 4.3 (typ) | –8.3 to –3.7 (typ ~–6.0) | 3028 (at 400V) | 360 (at 400V) | 364 (at 400V) | H-PDSO-F8 | 0.10 (typ)onsemi.com |
| onsemi (EliteSiC) | UJ4N075005K4S | 750 | 120 | 4.8 (typ) | –8.3 to –3.7 (typ ~–6.0) | 3028 (at 400V) | 360 (at 400V) | 364 (at 400V) | TO-247-4L | 0.16 (typ)onsemi.com |
Recommendation:
- Primary Choice: onsemi UF3N120007K4S. This 1200V, 120A, 11mΩ, 789W device in a TO-247-4 package is an exceptional performer. It offers immense headroom in all critical parameters, ensuring the amplifier operates with greater stability and potentially cooler temperatures. The 4-pin package with Kelvin source connection provides an opportunity for improved high-frequency performance, though it would require a minor modification to the PCB layout for the fourth pin.
- Alternative (Direct 3-pin Drop-in): onsemi UJ3N065025K3S. This 650V, 85A, 33mΩ, 441W device in a TO-247-3 package is an excellent choice for a direct physical and functional replacement. While its voltage rating is lower than the original, it is still more than adequate for the J2 amplifier's +/-23V rails, and its current, on-resistance, and power dissipation capabilities are significantly superior to the SJEP120R100.
Conclusions and Recommendations
The obsolescence of the original SemiSouth SJEP120R100 necessitates the selection of modern SiC alternatives for the FirstWatt J2 DIY Clone amplifier. The analysis of currently available "normally-on" devices reveals a clear path for replacement, with options that not only meet but significantly exceed the original component's specifications. This improvement is a direct result of the rapid advancements in wide-bandgap semiconductor technology since the original JFET's introduction.For the Q5 and Q6 positions, the following conclusions and recommendations are provided:
- Direct "Normally-On" SiC JFET Replacement (Minimal Circuit Modification):
- Conclusion: Modern discrete SiC JFETs from onsemi are the most straightforward replacements for the original SJEP120R100. They are inherently "normally-on" (depletion-mode) devices, aligning perfectly with the existing J2 amplifier's design philosophy and gate drive requirements. These devices offer vastly superior current handling, significantly lower on-resistance, and substantially higher power dissipation ratings, leading to improved efficiency, reduced thermal stress, and enhanced long-term reliability for the amplifier.
- Recommendation:
- Primary Choice: onsemi UJ3N120065K3S. This 1200V, 34A device in a TO-247-3L package is an excellent choice for a direct physical and functional replacement. Its voltage rating perfectly matches the original, and its current handling is significantly superior.
- Alternative: onsemi UJ3N065080K3S. This 650V, 24A (at 100°C), 80mΩ, 190W device in a TO-247-3 package is another strong candidate. While its voltage rating is lower than the original, it is still more than adequate for the J2 amplifier's +/-23V rails, and its current, on-resistance, and power dissipation capabilities are superior to the SJEP120R100.
✅ Recommendation
Both UJ3N120065K3S and UJ3N065080K3S are suitable replacements for the SJEP120R100 in the First Watt J2:
- UJ3N120065K3S is closer in voltage rating and offers a lower R<sub>DS(on)</sub> than the original SemiSouth device. However, it has higher gate cutoff voltage (~–6.6 V), so ensure your gate bias circuit accommodates this.
- UJ3N065080K3S is slightly closer in capacitance to the SJEP120R100 and is better suited for high-frequency audio performance, but its 650 V rating (still >8× expected voltage swing) is perfectly safe for the J2 topology.
If thermal headroom and lower R<sub>DS(on)</sub> are important, especially at higher idle currents, UJ3N120065K3S is the better choice.
Let me know if you want SPICE models, bias calculations, or help simulating these parts in the J2 circuit.
is any of this true and/or accurate?
Thanks,
Jeff
Last edited:
myfirstamplifier, the moderation team has flagged your thread as being inefficient, and would like you to consider redrafting your questions in a new thread.
In this particular case, we feel that a large AI dump is a burden to read when members would be more interested in only the issues that are concerning to you. Members may feel that there is little to be gained from reading such content, and may even resent being expected to digest it.
We feel that questions asked in your own words will be eminently preferrable and better fitting with the spirit of the site.
If you have any questions, feel free to PM myself or another member of the team.
- Status
- Not open for further replies.