Varistor voltage tolerance decides whether two Walsin voltage steps are really different parts. The export prints plus or minus 10 percent on all 379 rows. We computed the band of every step: 30 of the 34 adjacent steps overlap, so a bench reading can belong to two catalogue steps at once.
Venatech Limited, an authorised Walsin distributor in Hong Kong, holds the metal oxide varistor parametric export from Walsin Technology (Walsin): 379 rows, 35 distinct varistor voltages. This post is written for buyers and engineers in Hungary who cross a varistor by its voltage step. It uses that export only. We hold no varistor datasheet, so every figure is the column as published. Every part number below is real.
Do the plus or minus 10 percent bands of neighbouring Walsin varistor voltages overlap?
Yes, on 30 of 34 neighbouring pairs. We parsed all 379 rows, sorted the 35 distinct varistor voltages, and computed 0.9 times and 1.1 times each one. A step’s band overlaps the next step’s band when 1.1 times the lower voltage exceeds 0.9 times the upper. That holds on 30 pairs. Two pairs touch exactly and two leave a gap. The arithmetic uses integers, so a touching pair is not a rounding artefact.
| Outcome for the pair | Pairs | Example pair (Vdc) | Real parts | Match / mismatch against “each step is a distinct band” |
|---|---|---|---|---|
| Bands overlap | 30 | 300 and 330 | SR301K05D, SR331K05D |
Mismatch — 300 reaches 330 V, where 330 starts at 297 V |
| Bands touch at one point | 2 | 18 and 22 | SR180K05D, SR220K05D |
Mismatch — both bands contain 19.8 V |
| Gap between bands | 2 | 120 and 150 | SR121K05D, SR151K05D |
Match — the bands are 3 V apart |
We have not seen this counted anywhere. The 35 steps have 34 neighbours, and 30 of them, 88.2 percent, cannot be told apart by a varistor voltage reading alone. 367 of the 379 parts sit on a step whose band overlaps a neighbour. The 12 that do not are at 18, 22 and 27 Vdc.
Which four neighbouring steps do not overlap?
Two of them touch and two leave a gap, all at the bottom of the ladder or at 120 to 150. We listed the four pairs whose ratio clears 1.1 divided by 0.9, which is 1.2222. No pair above 180 Vdc reaches that ratio: the widest step there is 1.179, between 240 and 283 Vdc, and the narrowest anywhere is 1.040, between 750 and 780 Vdc.
| Pair (Vdc) | Ratio | Top of the lower band (V) | Bottom of the upper band (V) | Real parts | Match / mismatch |
|---|---|---|---|---|---|
| 18 and 22 | 1.2222 | 19.8 | 19.8 | SR180K05D, SR220K05D |
Mismatch — touching, gap of 0.0 V |
| 22 and 27 | 1.2273 | 24.2 | 24.3 | SR220K05D, SR270K05D |
Match — 0.1 V gap |
| 27 and 33 | 1.2222 | 29.7 | 29.7 | SR270K05D, SR330K05D |
Mismatch — touching, gap of 0.0 V |
| 120 and 150 | 1.2500 | 132.0 | 135.0 | SR121K05D, SR151K05D |
Match — 3.0 V gap, the only real gap |
A 0.1 V gap is smaller than any instrument tolerance a bench would carry, so only the 120 to 150 pair is clear in practice. The low-voltage corner and the 120 to 150 pair are the only places the printed tolerance leaves neighbouring steps apart. Everywhere else the ladder is spaced more tightly than its own printed tolerance.
Where is the overlap deepest?
At the top of the ladder, by more than 100 V. We subtracted the bottom of each upper band from the top of each lower band and ranked the 30 overlapping pairs. The six deepest all sit between 750 and 1181 Vdc, and the deepest is 910 against 950 Vdc, where the two bands share 146 V.
| Pair (Vdc) | Overlap (V) | Real parts | Match / mismatch |
|---|---|---|---|
| 910 and 950 | 146.0 | SR911K10D, SR951K34R |
Mismatch — 910 reaches 1001 V, 950 starts at 855 V |
| 950 and 1000 | 145.0 | SR951K34R, SR102K10D |
Mismatch — 950 reaches 1045 V, 1000 starts at 900 V |
| 1000 and 1068 | 138.8 | SR102K10D, SR112K10D |
Mismatch — 1000 reaches 1100 V, 1068 starts at 961.2 V |
| 750 and 780 | 123.0 | SR751K10D, SR781K10D |
Mismatch — 1.040 ratio, the narrowest step in the file |
| 1068 and 1181 | 111.9 | SR112K10D, SR122K53D |
Mismatch — the 1181 Vdc step has one part, SR122K53D |
The steps at the top are not thin on parts either. The 950 Vdc step has 3 parts, the 1181 Vdc step has 1, and parts that share 146 V of band cannot be told apart by a voltage reading but can differ widely in peak current: SR911K10D is rated 2,500 A and SR951K34R 40,000 A. See the varistor voltage ladder for which footprints reach each step.
Does the overlap reach beyond the next step?
On 17 of 33 two-step pairs, yes. We repeated the test between each step and the one two places above it. The first such pair is 205 and 240 Vdc, and the 17 run up to 1000 and 1181 Vdc. For those steps one band covers three catalogue values, so a part measured at 330 V is inside the 300, 330 and 360 Vdc bands at once.
That is arithmetic on a printed number, not a statement about Walsin’s production distribution. It tells you what the printed tolerance allows and nothing about where real lots sit. We hold no distribution data.
What does this mean for selecting a varistor in Hungary?
Select on the working voltage and on peak current, not on a varistor-voltage reading. The export pairs each varistor voltage with exactly one maximum working voltage, 35 for 35, so the Vac column is a clean selector where the Vdc column is a blurred one. At 275 Vac the paired varistor voltage is 430 Vdc; its band runs 387 to 473 V, and the 470 Vdc step starts at 423 V.
| Field | Walsin export, SR391K10D |
Your incumbent part | Match / mismatch |
|---|---|---|---|
| Maximum working voltage | 250 Vac | to be filled in | Match only if equal or higher in your circuit |
| Varistor voltage | 390 Vdc, printed plus or minus 10 percent | to be filled in | Mismatch risk — band 351 to 429 V overlaps both neighbours |
| Peak current | 2,500 A | to be filled in | Compare the waveform, which the export does not state |
| Disc size | 10 mm | to be filled in | Compare the footprint |
Check the working voltage first, then peak current, then disc size. Treat the varistor voltage as a guide to the step, never as a pass or fail test on an incoming lot. For the peak current inversions that also trap a crossing, see disc size against peak current.
We hold 28 Walsin datasheets, 451 pages, none of them covering varistors, so the tolerance, its test current and its reference temperature are not in the documents we hold. Ask Walsin or us for the varistor documentation on your part number and confirm against both datasheets before you qualify in your own circuit.
What is in Hong Kong is on ready stock, 1,125 part numbers as at 23 September 2026, a dated snapshot and not a live feed. Cross a varistor with the free cross-reference tool, and see the range on the varistor product page.
The short version
Walsin prints plus or minus 10 percent on all 379 varistors and the 35 voltage steps are spaced closer than that tolerance. 30 of 34 neighbouring pairs overlap, 2 touch exactly at 18 and 22 Vdc and at 27 and 33 Vdc, and 2 leave a gap. Only 120 to 150 Vdc is separated by a real 3 V. The deepest overlap is 146 V, between 910 and 950 Vdc, and 17 of 33 two-step pairs overlap as well. 367 of 379 parts sit on a step with a neighbour inside its band. We will not tell you two steps are interchangeable. We will show you the bands. Qualification stays with you.
Send us the varistor voltage and working voltage off your drawing and we will show you which steps your tolerance reaches. A question back: when your incoming inspection measured 330 V, which step did it decide you had bought?
Frequently asked questions
What tolerance does Walsin print on its varistor voltage?
Plus or minus 10 percent, on all 379 rows of the export. It is the only value in the Tolerance column, so the column does not distinguish one part from another. The export does not say what the tolerance applies to or at what test current, and we hold no varistor datasheet to confirm it.
Do Walsin varistor voltage steps overlap within the printed tolerance?
Yes. Of 34 neighbouring pairs among the 35 distinct varistor voltages, 30 overlap at plus or minus 10 percent, 2 touch exactly and 2 leave a gap. The deepest overlap is 146 volts, between 910 and 950 volts direct current. The steps are spaced closer than the tolerance allows.
Which Walsin varistor steps are cleanly separated?
Only the 120 and 150 volt direct current pair has a real gap, 3 volts between 132 and 135. The 22 and 27 pair is 0.1 volt apart, and 18 and 22 and 27 and 33 touch exactly. Every pair from 150 volts direct current upward overlaps.
Should I select a varistor by its varistor voltage in Hungary?
Use it to find the step, not to decide. The maximum working voltage maps to the varistor voltage one for one across 35 steps, so the working voltage is the cleaner selector. Then compare peak current and disc size. Qualification against the datasheet stays with you.
Is a measured 330 volts inside more than one step?
On the printed tolerance, yes. The 300 volt step reaches 330 volts at its upper edge, the 330 step is centred there, and the 360 step starts at 324 volts. A single reading cannot tell the three apart without a datasheet limit.
Does Venatech hold a Walsin varistor datasheet?
No. We hold 28 Walsin datasheets, 451 pages, covering 26 MLCC and 2 chip resistor series. Varistors, disc capacitors, inductors, thin film, radio frequency and antenna parts have none, so every varistor figure here is from the export and should be confirmed with Walsin.