FAQ

The recommended replacement for the MD90F18 is the SXF20FF3. The recommended replacement for the MD90U25 is the SXF6523. If the diodes are operated at full rated voltage, the voltage stressed between leads will exceed the general design guideline of 10V/mil in air. It is recommended that the isolation voltage between leads be increased through the use of a conformal coating, the use of a dielectric liquid or vapor. 

A rule-of-thumb for determining whether additional encapsulation is required is the 10KV/inch isolation in air rule, which assumes relatively low humidity and clean air (no particulates on the diode). For example, using the 10kV/inch rule for the 1N6517 diode (5kV, 1A, 70ns), based on body size and distance between leads, it could operate up to 2900V in clean, dry air. Above 2900V, or in less clean conditions, the diode should be encapsulated or operated in oil or some other dielectric fluid.

Many factors can effect whether the 10KV/inch rule is effective in applications. Additional factors include humidity levels, thickness of the glass passivation, body length between leads, ambient temperatures, vacuum levels, and power levels.

Generally speaking, additional isolation should be provided for diodes operating a levels of 5KV or higher.

Occasionally in high frequency applications where voltage rise times are fast, matching Trr can be advantageous. This only works for single junction diodes (1KV or less). Individual junctions in high voltage multi-junction diodes (greater than 1KV) cannot be accessed.

Multi-junction diodes work very well in high frequency applications.

Soft recovery.

Yes and no. 100% of our diodes are tested under avalanche conditions at 100uA. Yes, all of our diodes have some avalanche capability.

The real question becomes, “How much?” That varies from diode to diode and depends on several factors including number of junctions, junction area, resistivity of the wafers, the energy or power dissipated, and more.

In avalanche mode, a diode is reverse biased. Leakage current increases exponentially as the reverse voltage increases. Dissipated power increases as leakage current increases, which heats up the diode junction. Heating up the diode junction increases the leakage current, which increases the dissipated power, and so on until the diode fails catastrophically. If the diode is kept cool, it will continue to operate indefinitely.

VMI diodes are tested in such a way as to ensure reverse leakage current is low enough at the rated PIV to help prevent thermal runaway commonly associated with avalanche mode operation.

If the customer knows what the avalanche requirement is, VMI can determine if the diode in mind will meet it.

Yes. Diodes can be connected in series and/or parallel in order to increase the reverse voltage and/or forward current capacities.

There are practical considerations to keep in mind such as thermal management, voltage isolation, and component count, to name a few.

When connecting diodes in parallel it is usually a good idea to match V_f. Differences in V_f between diodes will get worse as they heat up. Diodes with lower V_f will initially conduct more current, increasing their power dissipation and further lowering their V_f. This in turn increases their current share, continuing to heat up the junction, and so on until a catastrophic failure occurs. 

The I²T (Amps²*Seconds) rating is defined as the single cycle surge current, I_fsm, multiplied by a standard pulse width of 8.3mS (sine wave is assumed).

It is used as a rule-of-thumb to gauge surge capability at different pulse widths. It works because at higher currents, V_f (forward voltage drop), is dependent on the resistive element of the diode. V_f becomes proportional to current in the diode expressed by V_f = R_diode * I_f. The I²T calculation yields energy dissipated in the diode during the pulse duration. Power multiplied by pulse time gives the energy pulse. Energy dissipated during a surge current pulse is proportional to I²T and is usually the driving force behind a failure. The energy pulse causes localized heating which induces mechanical fractures or disruption of the silicon crystal structure. Calculating the maximum I²T can help determine if a diode will survive a current surge.

Example: A diode has an I_fsm rating 100A. Will it work at 150A surge for 1uS?

Solution:

1. Calculate I²T

I²T = (100A)² * 8.3mS = 83A²S

2. Determine if I²T under the new conditions is much less than the original calculation. Is the I²t calculation at 150A, 1uS much less than 83A²S?

Is 83A²S >> (150A)²*1uS?

= .023 A²S

Yes, .023A²S is << 83A²S

The diode should be able to handle a surge of 150A for 1uS.

Comment: When approaching fast pulse times (i.e. ns range), I²T using I_fsm is used as an upper limit. Operating I²t should not exceed half of the upper limit.

Plastic diodes have very high leakage current at extended high temperature, and are not recommended for the HT options.

Yes, and no. When trying to identify the cathode and anode end of a high voltage diode, the meter has to have enough voltage to overcome the forward voltage drop, Vf, of the device under test. We use a test setup that includes a 30V current limited power supply. When testing for polarity, the current should be limited to mA going through the diode when it is connected in the forward direction.

The most effective method of mounting the K25UF diode (and most all axial-leaded epoxy diodes) is to make a cut-out in the board just a tad larger than the dimensions of the diode body and drop the diode in so the leads lay flat across the solder pads. If possible, avoid lead-forming. If space does not permit it, is important to support the diode body while forming the leads. The leads are relatively short and thick, which enhances the current carrying capacity of the device. The epoxy body is easily damaged if the leads are not supported near the body during the lead forming process.

The highest level QPL diode VMI manufactures is JANTXV. We have on many occasions, “up-screened” JANTXV diodes to the space level by following testing and processing guidelines outlined by NASA. Please contact us if you are interested in space level diodes.

VMI does not have any in-house life test data, but the calculated MTBF (Mean Time Between Failure)at 55C and operating 60% in a stressful environment (Missile Launch conditions) is 43 years.

If you have questions about MTBF on specific diodes, please contact us.

Diode dimensions are typically nominal with a tolerance of +/- .005 inches (.127mm) unless otherwise specified. If you have questions about specific dimensions, please contact us.

Balancing resistors and capacitors are generally not required when using VMI diodes. A resistors connected in parallel with a rectifier is intended to balance the reverse bias voltage across the series rectifiers. As such, the resistor value needs to be selected such that the current through the resistor is fairly large compared to the reverse leakage of the diode, thereby making a stiff divider circuit. If there are relatively high leakage currents expected in the diode, this can result in a significant power loss in the balancing resistors.

Capacitors connected in parallel to resistors/diodes can be used to balance voltages across diodes during transient conditions, or ringing. Care must be taken to select the capacitors small enough so as not to impact the rectified waveform at the output of the diode.

VMI’s diodes are well-balanced for Ir and reverse recovery time (Trr).

Many of VMI’s diodes have up to twenty junctions in series. The series junctions operate under many kinds of applications, and all sorts of conditions without compensating resistors or capacitors with no problems.

VMI’s diodes can easily withstand 175C and 200C non-operating (storage) temperatures.

During operation, junction temperature become very important. As a general rule of thumb, junction temperature should not exceed 125C.

Junction temperature is relative to the operating parameters in the application. If reverse recovery losses are not a factor, care should be taken to ensure the diode does not go into thermal runaway due to reverse current (Ir) losses.

If reverse recovery losses are an issue, the junction temperature should be kept below the point at which reverse recovery losses initial thermal runaway.

Reverse recovery losses can be attributed to reverse recovery time (Trr) of the diode, junction temperature, operating frequency, forward switching current in the circuit, diode series impedance in the circuit, and available thermal paths to get the heat away from the junctions.

Due to the uniqueness of each application, reverse recovery loss analysis must be performed during circuit design and testing.

Answer: The SMF and SXF families start with the basic, glass-body, hermetically sealed diode, and over mold it in a rigid epoxy. Following the encapsulation process, the leads are trimmed and formed. The diode itself is hermetically sealed. The epoxy is extremely moisture resistant, but not water-proof.

The MVM series can be used to generate either a positive or negative output voltage.

• To get a negative output voltage from the MVM type series, connect your input signal to V2. The HV output will be V1, and the output voltage will be negative.

Keeping all things constant, the output voltage decreases as the temperature increases. There is a significant drop-off at temperatures higher than 150C. Other factors effecting output voltage include load current and operating frequency.

Other hyrid families exhibit the same phenomena. The actual output voltage varies depending partly on the number of stages and input voltage. Other factors include operating frequency, output current, and type of load.

• MVM302P08 High Temp Data
• MVM202P06 High Temp Data
• MVM402P10 High Temp Data

No. Most multipliers work best between 25KHz and 150KHz. There are exceptions to the rule, but generally the absolute minimum operating frequency is 10KHz.

Operating frequency directly impacts

• Ripple voltage (the higher the operating frequency, the lower the ripple voltage)
• Regulation voltage (the higher the operating frequency, the less drop in output voltage at load)

Higher operating frequencies have greater efficiency in the multiplier. For a slow 60hz signal, very large capacitances would be required to regulate the voltage
properly and have a consistent high voltage output.

We have manufactured many standard multipliers for space projects. Typically JANTXV level diodes are up-screened to space level, and additional testing and source inspections performed on the final assemblies. View a list of space projects VMI has contributed to.

The signal lines include interface filtering and suppression components designed to protect the signal integrity and internal components. If an arc occurs during operation, transients can be coupled into the signal lines. Using good grounding techniques as well as shielding the interface cable will normally be sufficient to reduce coupled transients to non-disruptive levels.

The high voltage cable is an option that is purchased separately. The optional cable is rated to 150kV maximum voltage. The connector is a standard Amphenol series 97 circular connector without the pin plug. The connector is used as a shield ground path to the supply frame for the high voltage cable.

There are three main protection systems –

• Over-voltage
• Over-current
• Arc/short-circuit

Over-voltage System: This circuit limits the voltage output of the high voltage transformer to approximately 110% of the rated output.

Over-current System: This circuit limits the output current of the high voltage transformer using a current sense transformer output compared to a preset reference.

Arc/short-circuit System: This protection encompasses a variety of circuits and components working together to protect the internal components from transient effects of an arc as well as to clamp the output current to the current regulation set value in the event of a short circuit condition.

The HVP series power supplies are 250W supplies. They are available as either a positive or negative output as follows:

• 50kV 5mA
• 80kV 3mA
• 125kV 2mA

There should be no damage.

Pin 7: 11k ohm
Pin 8: 100kohm during normal operation
Pin 9 and Pin 10: 10k ohm

All interface pins have protective diodes and series resistors.

Processing should be specified by the customer and reviewed by engineering. Most processing should be done on the final assembly. Diode processing is limited due to requirements for use in an onto coupler.

Custom electrical configurations are not available.

Custom mechanical changes to the lead configuration or package size will require a new mold (upwards of $10K) for the OC-100 style, and a new shell for the OC-250 style.

The packages can only increase in size.

Custom mechanical changes can take upwards of twenty weeks depending on vendors, yields, and schedules.

No. The package sizes have been optimized to accommodate the various LEDs and photo diodes.

No. The potting material has been specially selected to maximize the optical transmission of light from the LED to the photocopies.

No. The 10KV rating for the OC-100 and the 25KV for the OC-250 are maximum ratings.

No, but the output (Ir) of the photo diode is decreased by one half if only one LED is working.

No. The manufacturer of the LED determines the LED current range. The end user must not exceed the maximum current as specified by the manufacturer.

The Vf rating is for each LED shown on the schematic.

For the OC-100 family, each LED is rated at 100mA.

For the OC025 (2.5kV opto-coupler), the rated If of 50mA, is for a single LED. VMI recommends that all four LEDs are connected in series so each LED sees the same amount of current.

Gain is determined by output/input, and in this case, is also dependent on the level of applied reverse voltage. Defining “gain” as LED current/Optocoupler current, the following graph illustrates gain at two different levels of Vrwm.

The power gain is 2. This is the ratio of the output power of the photo diode divided by the input power to the LEDs.

PTR = (10KV x 100mA) = 1
(5V x 100mA) .5
1/.5 = 2

No. The current gain is fixed.

The current gain of the device is 1/1000. This is the ratio of the output current (diode leakage current) divided by the forward current through the LED(s).

CTR = 100mA = 10E-06 =. 001 = .1%.
100mA 10E-03

Gain changes over temperature. Gain decrease as temperature increases. It is linear over the temperature range of –25C to +150C.
VMI provides a graph of Gain vs. Temperature as part of the data sheet.

In order to build a 2W opto-coupler, the number of LED’s would have to double, and there would have to be a way to remove the heat from the high voltage diode. It is not possible to do all of that and keep the package size the same.

However, it is possible to connect two OC-100 style optocouplers in parallel to get the equivalent of a 2W device.

We don’t recommend it. By connecting two OC-250s in series, the volts-per-mil stress between the LED and the diode in the second OC-250 in series is greater than 350V/mil. 350V/mil exceeds design parameters for the molding material used in the OC-250.

Yes, two OC-250, or two OC100HG, can be connected in parallel.

The OC100HG is RoHS compliant.

Cross-talk can be an issue with these devices. Because they are high voltage, many end-users will encapsulate the optocouplers (and surrounding circuitry) with an epoxy or coating to further insulate the optocouplers. This epoxy is usually opaque, so it eliminates the cross-talk.

You can paint or otherwise coat and protect the optocouplers, but you will want to do some testing after the parts are coated to ensure the coating does not cause any arcing around the part at your applied voltage levels. Some customers paint the outside of optocouplers with a white paint to have more complete internal reflection in the optocoupler, increasing the gain of the device.

The two holes you mentioned are just depressions in the epoxy that are there to assist in centering the LEDs during molding. They are not significant to the operation of the optocoupler. They can be covered up if you like. 

Incident light will increase the baseline leakage current in the photo diode. To minimize the baseline leakage, shield the opto-coupler from stray-light sources. If possible, test your circuitry in darkness and ambient light to determine whether ambient light will increase the leakage current to unacceptable levels.

The thermal resistance of the module to air through the epoxy body is very poor. The best heat path is through the diode leads. Thermal resistance from the photo diode junctions to lead length is 6°C/W at 0.10 inch, and 12°C/W at 0.20inch. This assumes both leads are attached to an infinite heat sink.

The delay time between the LED current and reaction time of the photo diode is listed as 2µs in the data sheet. It is defined as T_on and T_off.

Of more concern is the capacitive load. The output current supplied by the photo-diode is dependent on the LED current, and the gain of the opto-coupler (refer to data sheet). The opto-coupler acts like a current source. The switching time will depend on how long it takes the current output from the opto-coupler to charge the load capacitance.

Over time the gain on the OC100HG will decrease as rapidly as 30% over a three-hour time frame when running both LEDs at the full rated current of 100mA.

Lower operating levels of LED current results in a slower decline in opto-coupler gain, but a gain drop has been observed in even very low LED current levels over time.

SPICE models for the optocouplers are not available individually, but they can be accurately modeled by using a current dependent current source in series with a high voltage diode. This will approximate the optocoupler function.

The OC150G and OC150HG leads are solder-dipped in Sn96. Peak recommended reflow tempeature for the LEDs, according to their datasheet, is 260°C for 10 seconds. VMI diodes are rated at 350°C for three seconds, but given that the LEDs are significantly lower in temperature recommendations, the LED recommendation is the gating item.