ZXY-ED8-10 1:1 Audio Isolation Transformer, 600 Ohm to 400 Ohm Line Matching, ED8X10-2.0H Laminated Core, Gull-Wing Terminals in a 14 mm Package
A 1:1 audio isolation transformer in a 14 mm laminated package, meant for the point where a 600 ohm line meets a 400 ohm device input. Both windings are specified identically, 105 ohm of DC resistance and 2.0 H of inductance, which is what keeps a 1:1 part balanced instead of merely isolated. The band is quoted where it is used rather than as a headline: plus or minus 0.12 dB from 300 Hz to 4 kHz, 2.3 dB of insertion loss at most, and a return loss that stays above 28 dB down at 300 Hz.
- Turns ratio 1:1 within plus or minus 2 percent, winding 1-2 to winding 3-4.
- DCR 105 ohm on each winding within plus or minus 15 percent; inductance 2.0 H minimum.
- 600 ohm on 1-2 with a 400 ohm load on 3-4, measured at 1 kHz and 0 dBm.
- Plus or minus 0.12 dB from 300 Hz to 4 kHz; insertion loss 2.3 dB maximum.
- Return loss 28 dB minimum at 300 Hz, 35 dB minimum at 1 kHz and at 4 kHz.
- Hi-pot 1250 V AC at 1 mA for 2 s; laminated E-I core in a 14.1 by 7.95 by 5.0 mm outline.
1. Product Name and Marking
ZXY-ED8-10 is the product name and ED8X10-2.0H is the type: ED8X10 for the package, which measures 0.555 in by 0.313 in by 0.197 in, and 2.0H for the 2.0 henry minimum inductance. The part carries no printed label. What little marking there is sits on the bobbin between the core and the clamp, where it is too small to read in the photographs and too dark to photograph against the varnish; the only place the type appears in full is on the documents.
That makes the paperwork the identity of the part, and the paperwork is not one voice. The Chinese sheet that travels with this family is a two-row table for B2161 and B2132, at 1.0 H and 1.2 H, with 125 ohm, 135 ohm and 155 ohm of DC resistance and a hi-pot of 1 mA for 60 seconds. The English description and the specification sheet describe ED8X10-2.0H alone, at 2.0 H, 105 ohm on both windings and a hi-pot of 1 mA for 2 seconds. Both are plausible sheets for a family of 1:1 line transformers; only one of them is this part.
2. What It Does
It is a passive 1:1 transformer, two windings on one laminated core, no supply and no active parts. The signal goes in on one winding and comes out on the other with no copper path between them, so a difference in earth potential between two pieces of equipment stops at the core. With a 1:1 ratio the voltage is not meant to change, which is the point: the part is there to break a ground loop and to present 600 ohm to a line and 400 ohm to a device input, not to add or remove level.
A 1:1 part is harder to get right than a step-down, because both windings have to be the same. The sheet shows that it is: 105 ohm on 1-2 and 105 ohm on 3-4 to the same tolerance, and 2.0 H minimum on both when they are measured in parallel. Matching the two halves is what keeps the balance stable across the band, and balance is what a balanced line relies on to reject what it picks up on the way.
3. The Winding Pair and the Two Impedances
Two windings, four terminals, and a naming convention that is worth reading twice. Pins 1 and 2 are one winding and pins 3 and 4 are the other; the schematic draws 1 and 2 on the line side and 4 and 3 on the device side, which is why the ratio is written 1-2 to 3-4 rather than 1-2 to 4-3. The DC resistance is 105 ohm on each winding within plus or minus 15 percent at 25 degrees C, and the inductance is 2.0 H minimum with the two windings tied in parallel, measured at 250 Hz and 0.5 V.
| Parameter | Value | Condition |
|---|---|---|
| Turns ratio, 1-2 to 3-4 | 1:1, plus or minus 2 percent | Ratio test |
| DCR, winding 1-2 | 105 ohm, plus or minus 15 percent | 25 degrees C |
| DCR, winding 3-4 | 105 ohm, plus or minus 15 percent | 25 degrees C |
| Inductance, parallel | 2.0 H minimum | 250 Hz, 0.5 V, windings tied |
| Impedance | 600 ohm on 1-2 | 1 kHz, 0 dBm, 400 ohm on 3-4 |
| Dielectric strength | 1250 V AC, 1 mA, 2 s | Primary to secondary |
| Compliance | RoHS | Materials and process |
The two impedances are not a ratio, they are an interface: 600 ohm is what a line looks like and 400 ohm is what the device input looks like. The sheet states it as 600 ohm on 1-2 measured at 1 kHz and 0 dBm with 400 ohm on 3-4, which is the only way the number means anything; measured open circuit it would read higher and tell a designer nothing about the circuit it will sit in.
4. What the Signal Sees
Three numbers describe the part from the signal's point of view and this sheet gives all three, each with a condition. Frequency response is plus or minus 0.12 dB from 300 Hz to 4 kHz against a 1 kHz reference, which is flat enough that the transformer is not the thing colouring the channel. Insertion loss is 2.3 dB maximum at 1 kHz and 0 dBm into a 400 ohm load, which is the price of the copper and the core and is normal for a part this size. Return loss is 28 dB minimum at 300 Hz and 35 dB minimum at 1 kHz and 4 kHz, which says the input actually looks like 600 ohm across the band rather than only at the design frequency.
| Parameter | Value | Condition |
|---|---|---|
| Frequency response | plus or minus 0.12 dB | 300 Hz to 4 kHz, referred to 1 kHz |
| Insertion loss | 2.3 dB maximum | 1 kHz, 0 dBm, 400 ohm load |
| Return loss at 300 Hz | 28 dB minimum | Same load |
| Return loss at 1 kHz | 35 dB minimum | Same load |
| Return loss at 4 kHz | 35 dB minimum | Same load |
The band itself is 300 Hz to 4 kHz, which is the voice band and not the full audio band. That is not a limitation hidden in the small print, it is the application: this is a line and intercom part, and quoting it to 20 kHz would be a different design with a different core. If a programme path needs 20 Hz to 20 kHz, this is the wrong family, and the sheet is honest enough to say so by stating the band it was measured over.
5. Construction, Glue and Terminals
The build is a laminated E-I core clamped over a glue-dispensed bobbin, with the winding covered by polyimide tape and the terminals formed as gull-wing feet, two at each end. The glue is the detail that gets mis-described most often: the English description calls the part potted, which suggests a filled can, while the photographs show an open clamp with adhesive at the core joint and around the bobbin where it meets the core. The second reading is the one the pictures support, and it is also the one that matters for the buyer, because a clamped core can be re-clamped and a potted one cannot.
| Pin | Winding | Note |
|---|---|---|
| 1 | Winding 1-2, one end | 600 ohm line side; 105 ohm to pin 2 |
| 2 | Winding 1-2, other end | The pair is the line winding on the schematic |
| 4 | Winding 3-4, one end | 400 ohm device side; 105 ohm to pin 3 |
| 3 | Winding 3-4, other end | Drawn below pin 4 on the schematic |
6. Mechanical Dimensions
The outline is three figures in inches: 0.555 in long within plus or minus 0.020 in, 0.313 in wide and 0.197 in high, each within plus or minus 0.010 in. Converted, that is 14.1 mm by 7.95 mm by 5.0 mm, and the conversion is given here because most drawings that reach a mechanical engineer are in millimetres; the tolerances are quoted in inches because that is how they were written, and they do not convert to round metric values.
| Dimension | Value | Note |
|---|---|---|
| Overall length | 0.555 in, plus or minus 0.020 in | 14.1 mm converted |
| Overall width | 0.313 in, plus or minus 0.010 in | 7.95 mm converted |
| Overall height | 0.197 in, plus or minus 0.010 in | 5.0 mm converted |
| Terminals | Four, gull-wing, two at each end | Lead form not dimensioned |
| Mounting | Solder terminals, surface or wire | No land pattern in the set |
| Weight | Not stated | Not published |
7. Application
The band, the impedance pair and the 1:1 ratio put this part in three jobs: breaking a ground loop between two pieces of audio equipment, coupling a 600 ohm line to a 400 ohm device input, and sitting in a voice-band channel where 300 Hz to 4 kHz is the whole signal. It is a line part rather than a microphone part, since 2.3 dB of insertion loss and 2.0 H of inductance are both chosen for a signal that is already at line level.
For a ground loop the connection that matters is the one between the two windings: keep the line side on pins 1 and 2 and the equipment side on 3 and 4, and do not tie either winding to the chassis at both ends, or the loop comes back through the shield of the cable instead of through the core. The 1250 V barrier is what allows the two sides to sit at different earth potentials in the first place.
8. Ordering and Marking
Order by type, ED8X10-2.0H, and add the product name ZXY-ED8-10, because the family sheet that circulates with the Chinese documents carries B-numbers at other inductances and a different hi-pot duration. Say which ratio and which band you are buying, 1:1 over 300 Hz to 4 kHz, so that a substitute from the same family is not accepted on the grounds that it is also a 1:1 line transformer.
| Item | Status | Why it matters here |
|---|---|---|
| Dimension picture | Three callouts wrong | A 14 mm length labelled Overall Height, and 4.7 mm and 8.5 mm against 7.95 mm and 5.0 mm on the drawing |
| Width and height | Appear swapped in that picture | Its width figure matches the drawing's height and its height figure matches the width |
| Part numbers | Two families in one set | The Chinese sheet lists B2161 and B2132 at 1.0 H and 1.2 H, not ED8X10-2.0H at 2.0 H |
| Hi-pot duration | 2 s against 60 s | The English documents say 2 s, the Chinese sheet says 60 s at the same 1250 V and 1 mA |
| Inductance test | 250 Hz against 1 kHz | 2.0 H parallel at 250 Hz and 0.5 V against 1.0 H per winding at 1 kHz and 0.3 V |
| Construction | Potted against clamped | The description says glue-sealed potted; the photographs show an open clamp and glue at the core joint |
| Footprint | Not dimensioned | No pad drawing and no lead form, so the land pattern has to come from a sample |
| Marks | Four logos, no numbers | CCC, CE, RoHS and UL appear as artwork; no certificate or file number is printed |
| Temperature, class | Not stated | No operating temperature range and no insulation class anywhere in the set |