The DO-203 package is the JEDEC outline family for stud-mounted, hex-base rectifier diodes with solid terminals. Two variants carry almost all production volume: DO-203AA, still universally called DO-4, with a 10-32 UNF stud; and DO-203AB, called DO-5, with a 1/4-28 UNF stud. This page gives the dimensions, torque windows, thermal numbers and current ratings needed to specify, replace or correctly install one.

Key takeaways

• DO-203 means stud-hex base with solid terminals. DO-205 is the same idea with flexible terminals, per the JEDEC JEP95 DO index.
• DO-4 stud is 10-32 UNF-2A (M5 x 0.8 on metric parts). DO-5 stud is 1/4-28 UNF-2A (M6 x 1 on metric parts).
• Torque is not one number. Vishay specifies 1.36 N·m to 1.69 N·m on the nut for DO-4 with dry threads, and 1.07 N·m to 1.30 N·m if those threads are lubricated.
• DO-5 buys you roughly 3x the current and a much lower case-to-sink resistance: 0.25 °C/W versus 0.5 °C/W for DO-4.
• A plain part number means cathode-to-case. The “R” suffix means anode-to-case. Getting this wrong shorts your heat sink to the wrong rail.
• Distributor lead times on new DO-203 stock ran 40 to 45 weeks in mid-2026. Plan the buy, not just the design.

What the DO-203 outline actually specifies

JEDEC registers package outlines in publication JEP95. In the diode outline index, DO-203 is titled Stud-hex Base, Solid Terminals. Its sibling DO-205 is titled Stud-Hex Base, Flex Terminals. That single word — solid versus flex — is the family boundary, and it is the reason a DO-5 device has a stiff wire terminal while a DO-9 device has a braided flexible lead.

The index also records that the older DO-4 and DO-5 designations were archived and folded into DO-203 as variants AA and AB. Datasheets print both names. Vishay document 93492, for example, lists the case style as JEDEC DO-5 (DO-203AB) on the same page it prints outline drawing 95360 headed DO-203AB (DO-5). Treat the two as one part, not two.

An outline registration fixes the mechanical envelope only. It does not fix voltage, current, recovery time or junction temperature. Two devices in identical DO-203AB bodies can be a 50 V 35 A standard-recovery diode and a 1600 V 45 A device. The package tells you how it mounts, never how it performs.

DO-203AA (DO-4) vs DO-203AB (DO-5): the parametric comparison

Every figure below traces to the Vishay outline drawings and datasheets named in the source list. Dimensions are drawing maxima or ranges, not typical measurements.

Parameter	DO-203AA (DO-4)	DO-203AB (DO-5)
Stud thread, imperial	10-32 UNF-2A	1/4-28 UNF-2A
Stud thread, metric variants	M5 x 0.8	M6 x 1
Hex across flats	Not dimensioned on drawing 95311	17.15 mm to 17.38 mm (drawing 95360 rev. 27-Nov-2025)
Body diameter	Ø 6.8 mm feature called out	Ø 14.6 mm
Overall height, max	20.30 mm	25.40 mm
Terminal wire diameter	Ø 1.80 mm ±0.20 mm	Ø 3.80 mm to 4.10 mm
Approximate mass	7.0 g	17 g
Typical IF(AV)	12 A at TC = 150 °C	35 A / 40 A / 60 A at TC = 140 to 150 °C
IFSM, 60 Hz half sine	240 A	500 A / 800 A / 900 A
I²t for fusing, t = 8.3 ms	240 A²s	1040 A²s to 3400 A²s
RthJC, DC operation	2.0 °C/W	0.65 °C/W to 1.1 °C/W
RthCS, greased flat sink	0.5 °C/W	0.25 °C/W
VRRM range in family	50 V to 1000 V	50 V to 1000 V
TJ range	-65 °C to +200 °C	-65 °C to +200 °C
Max torque on nut, dry	1.69 N·m (15 lbf·in)	3.4 N·m (30 lbf·in)

The thermal column is the one that decides most designs. A DO-4 at 2.0 °C/W junction-to-case plus 0.5 °C/W case-to-sink puts 2.5 °C/W between the die and the heat sink before the sink itself is considered. Drop 8 W into that and the junction sits 20 °C above the sink face. The DO-5 equivalent, at 0.65 °C/W plus 0.25 °C/W, gives 0.9 °C/W — roughly a third of the internal thermal penalty for a package only 5 mm taller.

Worked example: can a DO-4 carry 10 A in a 60 °C enclosure?

Take a VS-1N1204A, 400 V, DO-203AA. Its maximum forward voltage drop is 1.35 V at 12 A average, so at 10 A average assume roughly 1.3 V and about 13 W of conduction loss. Junction-to-sink resistance is 2.5 °C/W, giving a 32.5 °C rise from sink face to junction. With TJ capped at 200 °C, the sink face may reach 167 °C — comfortable. The binding constraint is instead the sink itself: to hold the case near 100 °C in 60 °C ambient you need a sink of about 3 °C/W dissipating 13 W. The diode is not the limit here; the aluminium is.

Stud polarity and the “R” suffix

Vishay states the rule plainly in document 93492: the basic type number indicates cathode-to-case, and for anode-to-case you add “R” — 1N1188 becomes 1N1188R, 1N1186RA is the reverse-polarity 1N1186A. Distributor listings surface the same fact as “Stud Cathode” or “Stud Anode” in the diode-configuration attribute.

This matters because the stud is bolted to the heat sink, and the heat sink is usually chassis. In a positive-rail rectifier the cathode goes to the load, so a cathode-stud device electrically ties your sink to the output rail. Reverse-polarity parts exist so you can keep the sink at the other potential. If neither works, you insulate the stud — and then you pay for the insulator in thermal resistance and lose the low RthCS that made you pick a stud package.

Mounting torque, and why the datasheet gives four numbers

Stud packages fail at assembly more often than in service. Vishay application note 95556 is blunt about the mechanism: excessive torque, or driving the stud into a threaded hole, warps the hex base and cracks the die. Its only recommended fastening method is a nut and a washer.

The datasheets then split torque four ways, because friction differs by where you apply it and whether the threads are lubricated.

Package	Applied to	Thread condition	Torque
DO-203AA (DO-4)	Nut	Not lubricated	1.36 to 1.69 N·m (12 to 15 lbf·in)
DO-203AA (DO-4)	Nut	Lubricated	1.07 to 1.30 N·m (9.45 to 11.55 lbf·in)
DO-203AA (DO-4)	Device case	Lubricated	1.17 to 1.43 N·m (10.35 to 12.65 lbf·in)
DO-203AB (DO-5)	Nut	Not lubricated	3.4 N·m max (30 lbf·in)
DO-203AB (DO-5)	Nut	Lubricated	2.3 N·m max (20 lbf·in)
DO-203AB (DO-5)	Hexagon	Not lubricated	4.2 N·m max (37 lbf·in)
DO-203AB (DO-5)	Hexagon	Lubricated	3.2 N·m max (28 lbf·in)

The DO-5 limits carry a +0 % / -10 % tolerance, and the datasheet footnotes are the part most people miss: nut torque is the figure for pass-through holes, hexagon torque is the figure for threaded heat sink holes. Note also that lubricating the threads cuts the allowed torque by about a third, because a given torque now produces far more axial preload. If your line uses anti-seize on aluminium sinks, it is using the lubricated column whether the work instruction says so or not.

Heat sink preparation and the solder window

Application note 95556 sets numeric limits that most published guidance leaves qualitative. The sink contact surface should be flat within 0.03 mm with a levelling depth under 0.02 mm per DIN/ISO 1302, and the mounting face must be free of particles thicker than 0.05 mm. Mounting holes must be deburred. Vishay recommends a silicone- or petroleum-based joint compound such as Penetrox for the metal-to-metal joint, applied on the hex side.

For the terminal, the note gives a hand-soldering window of 260 °C to 275 °C for up to 10 minutes with lead-free solder when soldering the flexible lead to the diode eyelet, using a thermostatically controlled iron. That is a generous window compared with surface-mount parts, and it is one practical reason stud packages survive field repair.

When DO-203 is the wrong family

Vishay’s stud product list marks the ceiling clearly. DO-203AA tops out around 12 A to 22 A and DO-203AB around 60 A in standard-recovery silicon. Above that the flexible-terminal DO-205 family takes over: DO-205AA (DO-8) at 150 A, DO-205AC (DO-30) at 200 A, and DO-205AB (DO-9) from 300 A to 480 A across voltage classes up to 3200 V.

Going the other way, if your board is a PCB rather than a chassis and your current is under about 40 A, a TO-247AC part usually wins on cost, on assembly labour and on automated handling. Stud packages earn their keep where the thermal path is a machined chassis, where surge current is brutal, or where an existing enclosure was designed around a tapped hole twenty years ago.

Cost, lead time and second sourcing in 2026

This is where DO-203 sourcing gets uncomfortable. Checked in July 2026, DigiKey listed the Microchip S21160 (1600 V, 22 A, DO-203AA) at $28.98 each in hundreds against a 45-week manufacturer lead time. The 1N4530 (1200 V, 40 A, DO-203AB) sat at $53.84 in hundreds, also 45 weeks. Where stock exists it is not cheap: UES806, a 400 V 50 A DO-5 fast-recovery part, showed 100 pieces on the shelf at $71.18 each.

• Assume 40 to 45 weeks for anything not physically in a distributor bin. Order against the build, not the prototype.
• Cross-check the polarity suffix on every alternate. A 1N1186RA is not a drop-in for a 1N1186A.
• Distributor stock lists still carry RS and Vishay part numbers marked as discontinued. Confirm active status before you write the BOM line.
• For military and aerospace builds, JAN and JANTX screening exists — the 1N1199A family is available to MIL-S-19500/260, and JANTX1N4459 is qualified to MIL-PRF-19500/162. That screening path adds cost but preserves a defined second source.
• Buy the mounting hardware as a controlled item too. A wrong-grade nut on a 10-32 stud is how a torque spec quietly stops meaning anything.

Design mistakes that generate returns

• Tapping the stud into a threaded hole. Application note 95556 names this and excessive torque as the two mounting errors that crack the die. Use a clearance hole, a nut and a washer.
• Torquing to the dry number with lubricated threads. On a DO-5 that is 3.4 N·m applied where 2.3 N·m was allowed — nearly 50 % over.
• Ignoring the hex-versus-nut distinction. The higher hexagon figure only applies when the sink hole is threaded, which is exactly the configuration the mounting note warns against for other reasons.
• Assuming stud equals cathode. Both polarities ship in the same body with the same hex.
• Sizing the sink from RthJC alone. The case-to-sink term, 0.5 °C/W on DO-4, is roughly a quarter of the internal path and disappears entirely if you forget the compound.
• Reusing a stud diode after a hot removal. Once the hex base has been warped, the case-to-sink resistance never comes back to its greased-flat value.

Frequently asked questions

What is a DO-203 package?

DO-203 is a JEDEC-registered diode outline family for stud-mounted rectifiers with a hex base and solid terminals. The two production variants are DO-203AA, historically DO-4, and DO-203AB, historically DO-5. The registration defines mechanical dimensions and the stud thread only, not electrical ratings.

Is DO-203 the same as DO-4 or DO-5?

Yes. DO-4 and DO-5 are archived designations that JEDEC folded into DO-203 as variants AA and AB. Datasheets and distributor listings routinely print both, as in “DO-5 (DO-203AB)”. They describe the same physical part, so treat them as interchangeable names rather than distinct packages.

What thread size is a DO-203 stud?

DO-203AA (DO-4) uses a 10-32 UNF-2A stud, or M5 x 0.8 on metric versions. DO-203AB (DO-5) uses 1/4-28 UNF-2A, or M6 x 1 on metric versions. Both figures come from the Vishay outline drawings. Confirm against the specific datasheet, because metric and imperial variants share the same outline number.

How much torque should I use on a DO-5 stud diode?

Vishay specifies a maximum of 3.4 N·m (30 lbf·in) applied to the nut with dry threads, dropping to 2.3 N·m (20 lbf·in) if the threads are lubricated. Applied to the hexagon in a threaded heat sink hole, the limits are 4.2 N·m dry and 3.2 N·m lubricated, all with a +0 % / -10 % tolerance.

Is the DO-203 stud the anode or the cathode?

Either, depending on the part number. A base number such as 1N1188 is cathode-to-case. Adding “R” — 1N1188R — makes it anode-to-case. Distributors list this as “Stud Cathode” or “Stud Anode”. Never infer polarity from the body shape; check the marked symbol or the datasheet.

Can I replace a DO-203 stud diode with a TO-247 part?

Electrically, often yes below about 40 A. Mechanically it is a redesign: TO-247 mounts flat with a screw through a tab, needs a PCB or bracket, and typically has lower surge capability than a DO-5 rated at 800 A to 900 A IFSM. Verify I²t and surge margin before substituting, not just average current.

What to do next

If your current is under about 15 A and the chassis is already tapped, specify DO-203AA and torque the nut to 1.36 to 1.69 N·m with dry threads. If you are between 30 A and 60 A, or your surge duty exceeds a few hundred amperes, go straight to DO-203AB — the 0.25 °C/W case-to-sink figure and the 900 A surge rating pay for the extra 10 g and 5 mm. Above 60 A, stop looking at DO-203 and open the DO-205 datasheets. And in every case, order early: a 45-week lead time turns a good package choice into a schedule problem.

External references

• JEDEC JEP95 Diode Outlines index: https://www.jedec.org/sites/default/files/DOINDEX_0.pdf
• Vishay AN 95556, mounting instructions: https://www.vishay.com/docs/95556/mountingdo-203aa-do-203ab.pdf
• Vishay 93493, DO-4 12 A series datasheet: https://www.vishay.com/docs/93493/vs-1n1199a.pdf
• Vishay 93492, DO-5 35/40/60 A series datasheet: https://www.vishay.com/docs/93492/vs-1n1183series.pdf
• Vishay 95311, DO-203AA outline drawing: https://www.vishay.com/docs/95311/do203aa.pdf
• Vishay 95537, stud diode product list: https://www.vishay.com/docs/95537/productlsit.pdf

Image placeholders

• IMAGE 1: Side-by-side photograph of a DO-203AA and a DO-203AB rectifier with the hex bases and stud lengths visible against a scale rule. Alt: “DO-203 package comparison showing DO-203AA (DO-4) and DO-203AB (DO-5) stud rectifier outlines side by side”
• IMAGE 2: Annotated mechanical outline drawing of DO-203AB with stud thread, hex across flats and overall height called out. Alt: “DO-203AB (DO-5) stud rectifier outline dimensions including 1/4-28 UNF thread and 17.15 to 17.38 mm hex”
• IMAGE 3: Cross-section of a correctly mounted stud diode showing clearance hole, thermal compound layer, washer and nut. Alt: “Correct DO-203 stud diode mounting stack-up with clearance hole, thermal compound, washer and nut”
• IMAGE 4: Thermal resistance stack diagram comparing DO-4 and DO-5 junction-to-sink paths. Alt: “Thermal resistance comparison of DO-203AA and DO-203AB stud rectifier packages”

Internal link placements (fpga.io)

• INTERNAL LINK: “TO-247AC package” -> TO-247 outline, thermal path and creepage guide
• INTERNAL LINK: “junction-to-case thermal resistance” -> how to read RthJC and RthCS on a power datasheet
• INTERNAL LINK: “JEDEC outline numbering” -> JEP95 package outline naming explained
• INTERNAL LINK: “I²t for fusing” -> fuse coordination for rectifier diodes
• INTERNAL LINK: “heat sink sizing” -> natural-convection heat sink selection worksheet