#electronics #test-equipment #multimeter #siglent #owon # Overview Shopping for a benchtop DMM to sit alongside the [[Siglent Oscilloscopes|SDS804X HD]]. Four candidates under evaluation: OWON XDM1041 (~$121 AliExpress), UNI-T UT8804E ($269), Siglent SDM3045X ($399), Siglent SDM4055A ($739). Two adjacent Siglent models — SDM3055 ($574) and SDM3065X ($857) — are included because they land inside the same price band and change the calculus. Current front-runner: **OWON XDM1041** — but see the OWON Model Lineup section: the XDM2041 fixes three of its four weaknesses, and OWON shipped a genuinely new XDM1051 in June 2026. The decisive axes turned out to be 4-wire resistance, whether the remote interface is real SCPI or a vendor blob, and basic DCV accuracy — which does *not* track price across this set. # Resources ## Product pages - [OWON XDM1000 series (XDM1041 / XDM1241)](https://www.owon.com.hk/products_owon_4_1%7C2_digits_xdm1000_series_bench-type_digital_multimeter) - [UNI-T UT8804E](https://uni-trendus.com/products/ut8804e) - [TEquipment — Siglent SDM3045X](https://www.tequipment.net/Siglent/SDM3045X/Multimeters/) - [TEquipment — Siglent SDM4055A](https://www.tequipment.net/Siglent/SDM4055A/Benchtop-Multimeter/) ## Datasheets - [Siglent SDM4055A datasheet (EN01A)](https://www.siglenteu.com/wp-content/uploads/dlm_uploads/2025/04/SDM4055A_Datasheet_EN01A.pdf) - [Siglent SDM3045X datasheet (E04A)](https://res.cloudinary.com/iwh/image/upload/q_auto,g_center/assets/1/26/SDM3045X_DataSheet_E04A.pdf) - [UNI-T UT8804E "programming manual"](https://storage.googleapis.com/uni-tdocuments/Library/UT800E%20Series%20%26%20UT8804E%2CUT8803E%2CUT8802E%20Digital%20Multimeters%20-%20Programming%20Manual.pdf) — read this before buying, see below ## XDM1041 community + tooling - [TheHWcave/OWON-XDM1041](https://github.com/TheHWcave/OWON-XDM1041) — reverse-engineered SCPI command list plus a Python 3 logging utility (Linux + Windows). 64 stars. The SCPI list is more complete than OWON's own doc. - [XDM1041 teardown (Pallav Aggarwal, 2023)](https://pallavaggarwal.in/2023/12/30/owon-xdm1041-multimeter-teardown/) - [Eleshop customer reviews](https://eleshop.eu/review/owon-xdm1041-multimeter) - [essentialscrap — fixing HY3131 crystal noise](http://essentialscrap.com/owon_fix/) / [Hackaday writeup](https://hackaday.com/2021/07/14/fixing-noisy-measurements-on-an-owon-xdm2041-bench-multimeter/) ## Other OWON models - [OWON benchtop DMM index](https://www.owon.com.hk/list_benchtop_multimeters) — the authoritative lineup; ignore the AliExpress "wiki" comparison articles, which are AI-generated and get the counts wrong - [XDM1051 / XDM1251 (5½ digit, new 2026-06)](https://www.owon.com.hk/products_owon_5_1%7C2_digits_bench-type_digital_multimeter) · [spec PDF](https://files.owon.com.cn/specifications/XDM1051_Series_Specifications.pdf) - [XDM2041 (4-wire, trend chart)](https://www.owon.com.hk/products_4_1%7C2_digits__bench-type_xdm2041_digital_multimeter) · [Jameco listing](https://www.jameco.com/z/XDM2041-Owon-Technology-Lilliput-Electronics-4-5-Digit-Benchtop-Digital-Multimeter-55000-Count-True-RMS_2297631.html) - [Eleshop XDM1241](https://eleshop.eu/owon-xdm1241-multimeter.html) — clearest side-by-side of the 1041/1241 split # Comparison ## Headline specs | | OWON XDM1041 | UNI-T UT8804E | Siglent SDM3045X | Siglent SDM4055A | |---|---|---|---|---| | Price | ~$121 (AliExpress) | $269 | $399 | $739 | | Digits / counts | 4½ / 55,000 | 4⅚ / 59,999 | 4½ / 66,000 | 5½ / 220,000 | | DCV basic acc. (1 yr) | 0.05% + 5 | 0.025% (claimed) | 0.06% + 8 cts | 0.015% + 0.003% rng | | 4-wire resistance | No | **No** | Yes | Yes | | Reading rate | 60 rdgs/s | not published | 150 rdgs/s | 5–4,800 rdgs/s | | Remote | SCPI, USB + RS-232 | USB HID, read-only | SCPI, USB + LAN | SCPI, USB + LAN, VNC/web | | GPIB option | No | No | Yes | Yes | | Display | 3.5" 480×320 | 4.3" TFT | 4.3" TFT 480×272 | 5" 800×480 touch | | Temperature | K-type, PT100 | TC only, −40…1000 °C | RTD + TC | RTD + TC | | Capacitance | 50 nF – 50 mF | 6 nF – 60 mF | 2 nF – 10 mF | 2 nF – 10 mF | | Logging | 1,000 pts | 20k | 1 Gb NAND | 2M readings / 360M to file | | Scanner card | — | — | — | 16-ch optional (-SC) | | Safety | CAT I 1000 V / CAT II 600 V | — | — | — | | Size / weight | 235×88×65 mm, 0.45 kg | 2.27 kg | 3.76 kg, 294×260×107 mm | 3.8 kg, 379×260×103 mm | | Warranty | grey-market, none in US | 3+2 yr | Siglent | Siglent | ## The rest of the Siglent ladder | Model | Price | Digits / counts | DCV basic | Reading rate | |---|---|---|---|---| | SDM3045X | $399 | 4½ / 66,000 | 0.06% | 150/s | | SDM3055 | $574 | 5½ / 240,000 | 0.015% | 150/s | | SDM4055A | $739 | 5½ / 220,000 | 0.015% | 4,800/s | | SDM3065X | $857 | 6½ / 2,200,000 | 0.0035% | 10,000/s | Prices from TEquipment, checked 2026-09-14. ## Per-model notes **OWON XDM1041** — Half-width, 0.45 kg, AC-powered (the XDM1241 is the same meter with a battery). Real SCPI over a CH340 USB serial port at 115200 8N1. The best dollar-per-capability in the set by a wide margin, and the only sub-$150 option with a scriptable interface. No 4-wire, tiny log buffer, CAT I/II only. **UNI-T UT8804E** — The "programming manual" is not SCPI. It's a proprietary Windows DLL (`uci_ReadX`) over USB-HID exposing exactly two **read-only** commands, `data?;` and `disp?;`. Function and range cannot be set remotely. UNI-T's own FAQ says step up to the UT8805E/UT8806E for SCPI and 4-wire. Combined with no 4-wire resistance, this is a standalone display, not an automatable instrument. ==Disqualified on remote control.== **Siglent SDM3045X** — The hard one to defend. At $399 its 0.06% DCV spec is worse than the $269 UT8804E's claimed 0.025% and worse than the $121 OWON's 0.05%. What $399 buys is 4-wire, LAN + SCPI, RTD support, dual display, and Siglent support — not accuracy. A 4½-digit meter priced like a 5½-digit one. **Siglent SDM4055A** — The only genuine bench instrument in the group. 4× the counts and 4× the DCV accuracy of the SDM3045X, 4,800 rdgs/s, 2M-reading memory, auto 3 A/10 A current switching (30 A with the SCD30A external shunt), 16-channel scanner card option, and web/VNC remote — the same interaction model as the SDS804X HD. # XDM1041 Deep Dive ## JazzTech Electronics review — 2026-03-29 Most substantive recent video review, 24 min. Note: unit supplied by Banggood, and the description carries affiliate coupon codes — treat the framing accordingly, though the measurement results are run against a control meter and a PDVS2 mini reference source. All tests were run at the default speed setting [Video 0:00](https://youtu.be/zTtXa6ez7t0?t=0). **What measured well** - DC volts 1–10 V: "much better than spec," accuracy very good; polarity reversal read −0.9998 [Video 9:23](https://youtu.be/zTtXa6ez7t0?t=563) - mV DC clean all the way down to 0.1 mV [Video 10:32](https://youtu.be/zTtXa6ez7t0?t=632) - DC current 10 µA through 1 A, across both the 600 mA and 10 A sockets: no complaints [Video 11:43](https://youtu.be/zTtXa6ez7t0?t=703) - AC volts 10 V / 100 V / 200 V / 300 V all good [Video 12:55](https://youtu.be/zTtXa6ez7t0?t=775) - Resistance: 10 Ω and 10 kΩ dead on; 100 Ω, 1 kΩ slightly over; 100 kΩ slightly under [Video 14:37](https://youtu.be/zTtXa6ez7t0?t=877) - Frequency at 50 Hz very good, and it does read down to 1 Hz; duty cycle displayed simultaneously [Video 17:53](https://youtu.be/zTtXa6ez7t0?t=1073) - Diode mode gives forward voltage with an adjustable beep threshold — 3.29 V open-circuit, enough to light a blue LED (2.53 V) and a white one (2.58 V), which he notes is the first OWON to manage that [Video 8:02](https://youtu.be/zTtXa6ez7t0?t=482) - Overall: "reasonably accurate. Seems to be well within its spec" [Video 21:49](https://youtu.be/zTtXa6ez7t0?t=1309) **Where it fell down** - ==Capacitance: would not read the 160 pF residual of his decade box on auto range *or* manual 50 nF — "the first meter I think I've come across that won't"== [Video 15:57](https://youtu.be/zTtXa6ez7t0?t=957). Oddly, once 1 nF was switched in, it *did* show the residual on top of it [Video 16:31](https://youtu.be/zTtXa6ez7t0?t=991) - 1 MΩ read "more under than I would have expected" [Video 15:15](https://youtu.be/zTtXa6ez7t0?t=915) - AC current slightly over at 100 µA and 10 mA (in spec, but consistently high) [Video 12:55](https://youtu.be/zTtXa6ez7t0?t=775) - Frequency consistently slightly under at 1 MHz [Video 18:31](https://youtu.be/zTtXa6ez7t0?t=1111) - Temperature only "ballpark" [Video 19:06](https://youtu.be/zTtXa6ez7t0?t=1146) - Math function is max/min only, and takes two or three presses through menus to reach [Video 19:38](https://youtu.be/zTtXa6ez7t0?t=1178) - Data record works but menu navigation is "a little bit intensive" [Video 20:43](https://youtu.be/zTtXa6ez7t0?t=1243) - No trend graph on the display [Video 22:55](https://youtu.be/zTtXa6ez7t0?t=1375) - No dedicated Windows PC software he could find — SCPI is the only path [Video 21:18](https://youtu.be/zTtXa6ez7t0?t=1278) - Must hold the unit down to press the front buttons; not stackable [Video 2:17](https://youtu.be/zTtXa6ez7t0?t=137) **Useful details** - 30-minute warm-up, well under a full-size bench meter [Video 3:55](https://youtu.be/zTtXa6ez7t0?t=235) - CAT I 1000 V / CAT II 600 V only [Video 1:42](https://youtu.be/zTtXa6ez7t0?t=102) - Ships with probes (1000 V CAT III 10 A, "not particularly sharp"), croc clips, USB cable, manual, spare 250 V 1 A fuse [Video 3:22](https://youtu.be/zTtXa6ez7t0?t=202) - Utility menu has a **power-on setting: last-used or default** [Video 20:11](https://youtu.be/zTtXa6ez7t0?t=1211) — this matters, see below - ~£66 on Banggood at time of review [Video 21:49](https://youtu.be/zTtXa6ez7t0?t=1309) ## David Albert blog — 2025-12-01 [Owon XDM1041 DMM](https://www.dalbert.net/owon-xdm1041-dmm/). Focuses on the two things other reviews skip. **Frequency counter.** Tested against an SDG-2122X AWG slaved to a DOCXO reference. Accuracy dramatically exceeds spec — at 1 MHz the spec allows ±2 kHz and the meter read within 30 Hz. The real limits are signal level and resolution: | Input | Reading | Min Vpp for stable reading | |---|---|---| | 100 kHz | 99.997 kHz | 2.0 | | 1 MHz | 999.97 kHz | 2.0 | | 10 MHz | 9.9997 MHz | 2.5 | | 25–40 MHz | within 1–2 counts | 3.3 | | 60 MHz | 59.998 MHz | 3.6 | Practical consequence: at 4½ digits, a 16 MHz crystal resolves to 1 kHz per count = 62.5 ppm. Fine for confirming an oscillator runs; **useless for verifying a ±20 ppm or ±50 ppm crystal is in tolerance.** **SCPI.** Enumerates as a CH340 serial port, 115200 8/N/1. Firmware on his unit: **V3.3.0**. Confirmed working for automated DC current measurement: ``` *IDN? → OWON,XDM1041,2212364,V3.3.0,3 CONF:CURR:DC 500E-6 → -3=500mA, -4=50mA, -5=5mA, -6=500uA FUNC? → "CURR" RANGE? → "500 uA" AUTO? → 1 = auto, 0 = manual MEAS? → 1.557310E-05 ``` Also notes the USB port cannot power the meter. ## Firmware history — the settings-persistence complaint is stale Worth untangling, because it comes up in every older review: - A 2022-era GitHub note lists firmware **V2.1.2** (2021-05-25) as the newest available, with OWON saying no further updates were planned. That build had documented bugs: frequency misscaled when DUAL + FREQ are both selected, and RTD temperature readings freezing over the remote interface. - The most-cited Eleshop review (2025-03-22) complains the meter "returns to its default options" on every power-up and won't save measurement settings. - Dec 2025 unit ships **V3.3.0**, and the Mar 2026 review shows a power-on setting offering last-used or default [Video 20:11](https://youtu.be/zTtXa6ez7t0?t=1211). So OWON did ship at least one major firmware revision after declaring it done, and the headline usability complaint appears fixed. ==Verify the firmware version on arrival — AliExpress stock age is unpredictable and an old unit may still be on V2.x.== ## Recurring complaints across all reviews - Form factor cuts both ways: fits anywhere, but you must hold it to press buttons - UI has too many buttons for the functionality; Run/Stop and Dual are near-useless; can't change range while in math mode - K-type thermocouple drifts 4–5 °C within 30 minutes; PT1000 unsupported; PT100 hurt by lead resistance at low temps - Recorded measurements can't be pulled off to a computer through the UI (SCPI polling is the workaround) - Auto-ranging is slow, even in "fast" mode - Slight whine from the display above ~50% brightness - No sigrok support - Boot takes a few seconds ## The HY3131 crystal bug Petteri Aimonen ([essentialscrap](http://essentialscrap.com/owon_fix/), covered by [Hackaday 2021-07-14](https://hackaday.com/2021/07/14/fixing-noisy-measurements-on-an-owon-xdm2041-bench-multimeter/)) bought an XDM2041 for an automated test setup and found readings jumping around badly on the high-MΩ ranges. Teardown: - GigaDevices GD32F103CBT6 MCU plus a **HYCON HY3131** multimeter front-end chip - The HY3131 reference design calls for a **4.9152 MHz** crystal, chosen so sampling rejects mains hum. OWON fitted a **4 MHz** crystal instead, which aliases line noise straight into the measurement. - Swapping in a **3.072 MHz** crystal largely fixed it. He confirmed in comments that feeding an external clock through a 30 pF cap works for testing before committing to the swap. Two things follow that matter here: - He believes the **XDM1041 very likely shares the flaw** — similar chip, same 15 ms minimum sample interval, "very similar hardware but they realized the box was unnecessarily large." - ==He explicitly notes the stock timing "might work well enough with 60 Hz line frequency."== We're on 60 Hz, so this should be materially less severe here than for the European reviewers. It may still explain the "1 MΩ read more under than I would have expected" result in the JazzTech review. Caveat on the fix: a commenter points out changing the crystal throws off frequency measurement, since the counter derives from that clock. Separately, a 2024 commenter reports the AC millivolt display blanks to 00.0 below 5 mV, and OWON support confirmed that is intended behavior rather than a fault. # OWON Model Lineup Checked 2026-09-14 against OWON's own benchtop DMM index. **The XDM1241 is not a newer model** — it is the battery sibling of the XDM1041 in the same XDM1000 series, and OWON publishes a single spec PDF (`XDM1041_digital_multimeter.pdf`) covering both. ## XDM1041 vs XDM1241 — the only real differences | | XDM1041 | XDM1241 | |---|---|---| | Power | AC mains | Lithium battery, charges over USB-C or 5 V DC jack | | Connectivity | USB | RS232 + USB (per Eleshop) | | Dimensions | 235 × 88 × 65 mm | 200 × 86.5 × 64 mm | | In the box | Power cord, spare fuse | USB-to-DC cable; no spare fuse listed | | Eleshop price | €90 excl. VAT | €110 excl. VAT | | Reviews | 5 | 0 | Digits, counts, reading rate, display, accuracy tables, log depth, SCPI: identical. For a meter that lives on the bench the battery is a liability rather than a feature — one more cell to age out. Only worth the ~$25 premium if it needs to leave the bench. ## Where OWON's benchtop line actually stands | Model | Released | Digits / counts | DCV basic | Rate | 4-wire | Interface | Price | |---|---|---|---|---|---|---|---| | XDM1041 / XDM1241 | 2021 | 4½ / 55,000 | 0.05% + 5 | 60–65/s | No | USB, RS232 | ~$121 / ~$145 | | XDM1141 | 2023 | 4½ / 55,000 | 0.05% + 5 | 65/s | No | USB only | — | | **XDM2041** | 2019 | 4½ / 55,000 | **0.025% + 5** | 65/s | **Yes** | RS232 | $262.95 Jameco, €128 Eleshop | | **XDM1051 / XDM1251** | **2026-06** | **5½ / 120,000** | 0.05% + 5 | 85/s | No | USB, RS232 | ~$187 Banggood | | XDM3041 / XDM3051 | older | 4½ / 5½ | — | 150/s | — | USB/RS232/LAN/**WiFi** | €305 / €342 | | XDM6051 / XDM6061 | 2025-12 | 5½ / 6½, up to 2.2M | — | 1k / 10k/s | — | USB/LAN/RS232 | — | ## XDM1051 / XDM1251 — the genuinely new one Listed on OWON's site 2026-06, so roughly three months old. Over the XDM1041 it adds: - 5½ digits, **120,000 count** max reading (up from 55,000) - **85 rdgs/s**, with low/medium/high speed settings - **Dual display** — two measurement results at once - 3.5" **IPS** panel rather than TFT, same 480 × 320 - Temperature finally specified properly: K thermocouple −200…1300 °C and PT100 −200…800 °C, 0.1 °C resolution, ±(1% + 2 °C) - Math: min/max/average, dB/dBm, relative, AC+Hz - Same compact 200 × 86.5 × 64 mm body as the XDM1241 The catch: **basic DCV accuracy is still 0.05%**, identical to the XDM1041. This buys resolution, not accuracy. Still no 4-wire, still a 1,000-point log buffer. ==No independent reviews exist yet — too new. Nobody has verified the 120,000 counts are real counts rather than interpolation.== ## XDM2041 — the model that actually fixes the XDM1041's gaps Older (2019) and physically much larger (235 × 295 × 110 mm, a full-depth bench box), but it addresses three of the four complaints in this note: - **4-wire resistance** on 500 Ω / 5 kΩ / 50 kΩ at 0.1% + 10 / 0.1% + 5 - **DCV 0.025% + 5** — twice as accurate as the XDM1041 *and* the new XDM1051 - 2-wire resistance 0.1% vs 0.15% - **Trend analysis chart mode** — the exact feature JazzTech called out as missing [Video 22:55](https://youtu.be/zTtXa6ez7t0?t=1375) - Dual-line display, 3.7" 480 × 320, SCPI over RS232 The Eleshop reviewer who wanted 4-wire said plainly that size, not price, is why he skipped the 2041. ==US pricing is confusing: $262.95 at Jameco against €128 at Eleshop — check AliExpress before assuming the Jameco number.== # Decisions Pending. Current thinking: 1. **Rule out the UT8804E.** The read-only vendor-DLL interface and missing 4-wire make it a dead end for anything automated, regardless of the good headline accuracy number. 2. **Rule out the SDM3045X.** It's the one purchase in this set that's hard to justify at its price — worse DCV accuracy than meters costing a third as much, and $175 below the SDM3055 which fixes that. 3. **XDM1041 as front-runner is defensible.** At ~$121 it does the one thing the UT8804E can't (real SCPI), its DC voltage and current performance reviews well against reference sources, and the community SCPI documentation is better than the vendor's. Accept that it is not a bench reference: no 4-wire, 55k counts, CAT I/II, ballpark temperature, and capacitance that can't see 160 pF. 4. **The two-meter path.** Buy the XDM1041 now, keep the budget intact, and add a 5½ or 6½-digit Siglent later. A second meter is how you check the first one, and both end up used. 5. **If instead consolidating to one meter for the next decade:** SDM4055A at $739, or stretch $118 more to the SDM3065X for a full extra digit. 6. **The XDM1241 changes nothing** — same meter with a battery, ~$25 more. Skip unless portability matters. 7. **==The real open question is now XDM1041 vs XDM2041.==** If 4-wire turns out to matter (TODO item below), the XDM2041 gets it *plus* 2× the DCV accuracy *plus* the trend chart, for somewhere between +$30 (AliExpress) and +$140 (Jameco). The cost is bench footprint — it's a full-depth box, not a half-width one — and RS232 instead of USB serial. 8. **The XDM1051 is tempting but unproven.** More counts and a dual display, same 0.05% DCV, no 4-wire, and zero independent testing three months after launch. Not the one to buy sight-unseen. ## On hackability Don't expect a repeat of the scope trick. The SDS804X HD's bandwidth and options are firmware-gated on identical hardware, which is why the keygen works. The SDM tiers are real ADC and front-end differences — there is no license file that turns a $399 meter into a $739 one. There is an EEVblog thread titled "SDM3045X to SDM3055X-E Improvement Thread," but ==EEVblog is CAPTCHA-walled, so whether that thread documents a working modification or just an aspiration is unverified.== # TODO - ==Confirm which variant the AliExpress listing actually ships — product ID 1005006094754100 is a combined XDM1041 / XDM1241 listing; the $121.46 promo price points at the AC-powered XDM1041, but the page is CAPTCHA-walled and unconfirmed.== - ==Check firmware version immediately on arrival; target V3.3.0 or later.== - ==Decide whether the missing 4-wire resistance is actually a blocker for current bench work, or a hypothetical.== This now gates the XDM1041-vs-XDM2041 choice, not just the Siglent tier. - ==Price the XDM2041 on AliExpress — Jameco wants $262.95 but Eleshop is €128, so the real street price is probably $150–180.== - ==Measure bench depth before considering the XDM2041 — 295 mm deep vs the XDM1041's 65 mm.== - ==If an XDM1041 arrives noisy on high-MΩ ranges, check the crystal next to the HY3131 — a 4 MHz part is the known-bad one; 3.072 MHz is the community fix. Re-verify frequency-counter accuracy afterward if swapped.== - ==Confirm 10 A input fuse type and sourcing before the first high-current measurement — reviews mention a 250 V 1 A spare for the low-current path only.== - Pull the SCPI command list from TheHWcave's repo and test against the [[Siglent Oscilloscopes|SDS824X HD]] in the same pyvisa harness # References - [[Siglent Oscilloscopes]] — SDS804X HD unlock procedure and SCPI over telnet - [JazzTech Electronics review video](https://www.youtube.com/watch?v=zTtXa6ez7t0) (2026-03-29) - [David Albert — Owon XDM1041 DMM](https://www.dalbert.net/owon-xdm1041-dmm/) (2025-12-01) - [defpom — Owon XDM1041 55000 Count Multimeter Review](https://www.youtube.com/watch?v=xzsmyjWFtW8) — older but the canonical deep review