cut_50_chineese_plasma

Cut 50 Parts Cross Reference NTE Digikey Alldata Sheet

Diodes	Mosfets metal-oxide semiconductor field-effect transistor	IGBTs insulated-gate bipolar transistor
		Substituting IGBTs for Mosfets

Diodes D92-02 200V 20Amp
If your temp light is on and will not go off this may be the culprit.
Link to stateside supplier.
Link to lowest price per order.

Upgrade to DPG60C300QB 300V 30Amp diode.

Link

Mosfets

Mosfets
wfw20n50 • 20n50
2SK2837 • K2837 500v 20A Stateside Supply
2SK2837 • K2837 Lower Price

Nte Cross Reference for 2SK2837
NTE2970 Amazon Ebay
500v 20 A
TK20J60U(F)Toshiba 600v 20A
Possible higher rated device. >> 2SK2611= NTE2973=900v 14 A

These are the specs on the wfw20n50

2sk2837

IGBT Upgrades? Design notes.
Ebay 600v IGBTs

insulated-gate bipolar transistor

I use the IRG4PC30W (600v 23A) IGBT's for rebuilding power supplies that use the TO-247 style devices that have 500 volt or less and 20 amp or less ratings. Generally the frequency of plasma cutter power supplies are in the 30 - 50 Khz range so they are a direct and improved drop in for what I have worked on. I have also found that the power supply heat sinks that the switching devices are mounted to also tend to run noticeably cooler with IGBT's as well.

Yes I would change out all of the switching devices if possible. If you blow out another set of the old ones you will have to take the machine apart again anyway. Plus you run the chance of the new IGBT's getting damaged if the other old ones went bad.

The spark gap on a Cut50D might be around .015' - 025" I think.
All though some will work better with wider or narrower gaps due to poorly matched parts in the high voltage circuits. Its sort of a tune it to work best set up. Cheap equipment tends to have that problem fairly often.

About substituting IGBTs for Mosfets:

Drop-in substitution
In some cases, you can simply substitute an IGBT for a MOSFET, and vice versa. The pinouts and drive requirements for the two devices are similar. IGBTs and MOSFETs are voltage-driven transistors, unlike bipolar transistors, which require dc base-current drive. However, "voltage-driven" doesn't mean that the driving source needs no current-sourcing and -sinking capability. To the driving circuit, the gate input of an IGBT or a MOSFET looks like a capacitor, with capacitance values reaching thousands of picofarads for large devices. The driving circuit must be able to rapidly charge and discharge this capacitance to ensure fast transistor switching.
International Rectifier (IR) gives an example of replacing a MOSFET with an IGBT in a paper that describes a high-voltage application that uses 500V switching devices (Reference 2 ). IGBTs shine in high-voltage circuits, because their silicon efficiency (power-handling capability versus silicon area) is much higher than the silicon efficiency of MOSFETs. For a given power output, the IGBT has lower conduction losses and a significantly smaller die area. The smaller area brings the added benefits of lower input capacitance and lower cost.
Gate-drive requirements for IGBTs and power MOSFETs are similar: They both require 12 to 15V for full turn-on and need no negative voltage for turn-off. IGBT and MOSFET losses are different. In a MOSFET, the power wasted in losses comes principally from conduction losses, and switching losses are negligible at frequencies as high as 50 kHz. In an IGBT, conduction losses are much lower than in a MOSFET, but switching losses become significant at high switching frequencies. The design example in Reference 2 illustrates the point.
The design example assumes a 7.5A switched dc current, a 310V bus voltage, a 125°C junction temperature, and a 50% duty cycle. First, consider the losses in a power MOSFET. Using data-sheet information, the on-resistance of an IRFP450 at 125°C is 0.816W. Therefore, the conduction loss at 125°C is 0.816W3(7.5A)2X50%=23W. Assuming 75-nsec switching times at a 50-kHz switching frequency, the switching loss is 6.5W. Thus, total power loss is 29.5W.
Now, substitute an IRGP430U IGBT for the MOSFET. You calculate conduction loss using the device's on-state collector-emitter voltage, just as you would for a bipolar transistor. From the data sheet, V_CE at 125°C is 2.03V. Conduction loss is 2.03VX7.5AX50%=7.62W. For similarly rated devices, the conduction loss of an IGBT is much lower than the conduction loss of a MOSFET. Moreover, this "similarly rated" IRGP430U uses only 40% of the silicon area of the IRFP450. The current density in the IGBT die is therefore much higher than the current density in the MOSFET, so the same power dissipation results in higher junction temperatures. To maintain junction-temperature parity, you must reduce the dissipation in the IGBT chip. The following formula yields the permissible IGBT power dissipation:

where R_SA is heat-sink-to-ambient thermal resistance;
R_CSM is MOSFET-case-to-sink thermal resistance;
R_JCM is MOSFET-junction-to-case thermal resistance;
R_SI is IGBT-case-to-sink thermal resistance;
and R_JCI is IGBT-junction-to-case thermal resistance.

The allowable power in the IGBT calculates to 23.2W. This power comprises both conduction and switching losses. Deducting the calculated 7.6W conduction loss, the permissible switching loss in the IGBT is 23.2–7.6=15.6W. The data sheet gives a switching-energy spec in millijoules. You divide the permissible switching loss by this figure to obtain the maximum switching frequency. In this example, the maximum frequency is 56.4 kHz, which is comfortably within the 50-kHz target figure. Figure 2 gives a graphical comparison of the relative conduction and switching losses in the MOSFET and the IGBT.
Figure 3 gives some further IGBT-MOSFET replacements. The ordinate of the graph is the maximum operating frequency of the IGBT as a function of the switched current. The conditions for the substitution are a 50% duty cycle and an equal 125°C junction temperature for both the IGBT and the MOSFET. IR maintains that, in general, you can replace a MOSFET with an IGBT of two die sizes smaller that has approximately 40% the die area of the most closely rated MOSFET.

DIY Plasma Cutter