Product
Support FAQs
Layout
Are SynQor converters footprint and pin compatible with
Power-One and Tyco (Lucent) modules?
Yes, SynQor offers the same footprint and pin out configuration of the
Power-One and Tyco/Lucent half-brick, quarter-brick and eighth-brick
converters.
What type of thermal relief should I use?
In general, to provide improved solderability, customers will want to add
thermal reliefs to the output pins of the converter. The design of these
thermal reliefs should be carefully considered to prevent heating due to
resistive power losses. SynQor has written a
white paper on this subject entitled "Thermal and Reliability Study."
In general the aspect ratio of the conductive paths of the thermal relief
spokes should be wider than they are long, to reduce their resistance. A
significant fraction of the resistance between the converter and the load is
the first few centimeters of copper from the output pins and the power / ground
plane. Reductions in output resistance can be made by placing a copper pad on
every layer of the PCB around each of the output pins. Place 8 to 12 vias
surrounding the output pins, tying all the additional copper pads to the power
and ground planes. These suggestions will limit the resistance (and power
dissipation) in the board and increase the reliability of the plated through
holes.
^ back to top ^
Conducted EMI
What is the best strategy to minimize EMI?
There is no one perfect EMI strategy for all applications, but some basic
thought before-hand can make the task much easier. The first step is to make
sure that the location of components minimizes noise. For example, de-coupling
capacitors should be located as close as possible to the converter, especially
X-caps and Y-caps. Use ground planes to minimize radiated coupling, minimize
the cross sectional area of sensitive nodes, minimize the cross sectional area
of high current nodes that may radiate such as those from common mode
capacitors. The location of the EMI components is critical; avoid placing the
converter in close proximity to your filter to avoid noise coupling back into
the filter. Keep in mind that you are not just filtering the power supply, but
all the circuitry that the converter is powering as well. Most of today's
communication cabinets employ as much local filtering as possible at the card
level, and then another filter at the power entry module, where the power feeds
will enter your cabinet. SynQor has an EMI
application note entitled EMI Characteristics that further details
these principles and includes recommendations for 1 and 2 stage filters that
are designed around SynQor's converters. Please contact
SynQor directly for design assistance, or email
.
What type of conducted line filter should I use?
While a pre-designed EMI filter may be adequate for a particular converter,
there is no guarantee that it will suppress other sources of conducted EMI that
are present on your board, such as the noise from high-speed processors and
other digital logic devices. Better value and performance will be obtained with
a discrete filter design. The key to a smooth EMI compliance process is to
design as much filtering as possible onto your circuit pack. This will allow
the flexibility to tweak and modify components when the initial testing is
performed. It is best to over-design up front than be caught off guard;
components can easily be deleted once the initial testing is completed. Adding
components to an existing PCB design is much more difficult and can yield
unpredictable results. SynQor has suggested one and two stage filters that are
simple and reliable to implement, and for much less cost than an off-the-shelf
filter. These suggested filter circuits are in SynQor's EMI
Application note. In general most modern communications equipment uses
a local EMI filter on each of the circuit cards, and a final shielded filter
located in or near the power entry module.
How do I choose Y-Caps?
Y-caps are the EMI capacitors that are connected from the input power feeds to
chassis ground. Sometimes they are connected from each converter's power output
terminal to chassis ground as well. SynQor filter designs use 2700pF Y-Caps.
The voltage rating depends on the insulation and isolation safety rating of the
-48VDC supply. If you are unsure of these attributes, use capacitors rated to
2000V. If the -48V is a reinforced insulation scheme, then 100V rated
capacitors will suffice.
What does the input filter of your converter look like?
All SynQor half bricks use a C-L-C input filter. All SynQor quarter bricks and
eighth bricks have an L-C filter. The value of these components can be found in
SynQor's EMI application note,
or in the appropriate product
datasheets.
What if I don't have a chassis ground?
While it is preferable to have and to use a chassis ground, it is not
mandatory. Under some circumstances a chassis ground will not be available for
the input conducted EMI filter. If this is the case, you will want to use a
different filter topology. Contact
for this circuit.
What type of filter should I use for multiple converters?
For multiple converters, you can still use SynQor's recommended filter design,
just be sure to size the individual components to handle the maximum current at
the lowest input voltage.
I currently have my input and output filtering arranged
for a different manufacturer's converter. Will your module work with this
filtering or must it be changed?
In general, a filter that has been previously designed around a 200 to 500KHz
fixed frequency converter will be adequate for an equivalent SynQor converter.
SynQor's converters have lower common mode noise due to their lack of a
baseplate. Baseplated units have higher common mode noise as the switching
noise of the power semiconductor's couples through the parasitic capacitance of
the baseplate.
^ back to top ^
Radiated EMI
Are there any special concerns with an open frame
converter?
SynQor open frame converters are no worse, and often better, than baseplated
converters. In general most baseplated units have a plastic cover, and in 3
Meter and 10 Meter far field measurements offer no advantage compared to the
SynQor open frame design. SynQor's unique topology and reduced common mode
noise have a significant advantage over other manufacturers' converters in
regards to reducing radiated emissions.
^ back to top ^
Near-field EMI
How will an open frame converter perform versus a
baseplated converter?
With few exceptions, most customers have found the near-field characteristics
of SynQor's open frame converters to be no worse, and often better than the
typical baseplated converter. SynQor has several suggestions for reducing
near-field radiation- the first of which is to place a ground shield beneath
the converter. SynQor suggests placing a primary referenced ground shield
beneath the converter's primary circuitry, and a secondary referenced ground
shield beneath the converter's secondary side circuitry. The isolation barrier
on all SynQor converters is easily found by looking for the 1.4mm clearance gap
located on both the top and bottom of the PCB, or by locating the
optoisolators, which also identify the barrier. (Note: All of SynQor's magnetic
assemblies are considered primary referenced.) With regard to the near-field
EMI above the typical baseplated converter, a baseplate is not a perfect
"shield" as commonly thought. Unfortunately the baseplate is tightly
coupled to the converter's high frequency switching nodes (in particular, the
drains of the primary-side MOSFETs). The shield is therefore
"bouncing" and attempts to stop it from doing so by grounding it to
the output are challenging. Typically the circuit path to ground is hindered by
the parasitic inductance of the PCB connection. As such, the baseplate is not
well-grounded at high frequencies, and radiates significant noise. There are
also still the sides of the module to contend with, as these are not generally
shielded. If there is sensitive circuitry, it is best not to place it directly
over a converter or close to the edges of the converter unless it has a ground
shield directly over or under the conduction path, as this will reduce any
coupling. It is also helpful to note that the amount of noise that can be
coupled is proportional to the cross sectional area of the conduction path; the
smaller the loop, the less noise that will couple.
What about other Parasitic Baseplate Effects?
If you could hold the baseplate quiet by grounding through a low inductance
path, the result would be a great deal of common-mode current that would flow
from the primary side switching nodes to the baseplate (through parasitic
capacitors) and to the output ground. This then creates other problems that
need to be addressed, mainly conducted common mode noise. Common mode noise
tends to be a challenging problem to control. SynQor's converter design
eliminates this problem as it has very little parasitic capacitance to the
output ground. What noise it does have is effectively controlled with a common
mode capacitor that we place on our converter. Compared to the industry
standard Class B conducted noise filter, you will find that the SynQor
converter needs half as many filter stages due to the reduced common mode
noise. Also see SynQor's EMI Application
Note.
What about magnetic fields?
Baseplated converters will offer improved protection from near B-field
radiation. At most frequencies, a baseplated converter will be approximately
10dB/uM more quiet than the open frame design. If B-field noise is a critical
design consideration, any SynQor converter can be ordered with an optional
baseplate, which will offer the same reduction in B-field noise.
^ back to top ^
Enable Circuits/Inrush Control
Why do I need an inrush controller?
All converters act as negative resistors, and all power distribution paths have
some inductance. These two elements require that some electrolytic capacitance
be placed across a converter's input terminals to compensate for the negative
resistance, preventing a possible oscillation. If a circuit card containing a
converter is hot swapped into a back plane, the charging of the electrolytic
capacitor can cause a low voltage transient or glitch on the 48V supply, which
could upset other circuits sharing the common 48V feed. To prevent this glitch
on the input rail, an inrush control circuit should be placed in series with
the 48V feed. An inrush control circuit limits the charging of the 48V rail. A
good example of a controller for an inrush circuit is Linear Technology's
LT1640 and LT1641, Summit Micro's SMH4804, or Supertex's HV300 series.
How do I select an inrush controller?
You can drive the enable of a DC/DC converter with the inrush controller's
power good or enable signal provided there is no EMI filter between the
controller's ground reference and the converter's ground reference. If there is
an EMI filter located between the inrush controller and the enable circuit, the
noise which the EMI filter rejects will appear on the enable lines, and could
cause the modules to turn on and off randomly depending upon the size of the
injected noise. The inrush controller and the converter must have the same
reference. This can be avoided by leaving the modules permanently enabled, not
using the Power Good signal, moving the common mode filter to come before the
hot swap circuit, or adding some isolation between the Power Good output and
the input to the modules (such as an optocoupler).
Why is there a 200ms delay at turn on for your new family
of converters?
Most all of SynQor's converters have a 200ms startup delay at initial turn on,
and after a fault condition such as OVP or Over Temperature Shutdown. After
recovering from any fault condition, the converter will not turn on for 200ms.
When the bus voltage is below the under voltage lockout, this is considered a
fault condition. This behavior is detailed in our data sheets, and should be
studied before implementing an inrush or sequencing strategy. The delay ensures
that the start up behavior is always consistent and well controlled. If the
input voltage were to glitch momentarily to zero volts, then return to the full
bus voltage, the delay gives the control circuitry time to return to proper
startup conditions. Since the control circuitry has no voltage when the input
voltage is brought to zero, it is difficult to distinguish between power up
after a glitch and initial power up. Therefore, the module treats all
recoveries from undervoltage lockout identically. The delay on power up also
guarantees that the input voltage is well within the operating range before
powering up the output load. Note that after the delay time has passed, the
module will turn on quickly when the ON/OFF pin is set to the enabled state.
How do I sequence multiple converters?
Sequencing requirements need to be considered in the preliminary stages of the
power supply design. Usually these requirements are driven by ASICs or
processors, which have separate Core and I/O voltages. Often these voltage
rails must turn on in a specific order (sequence) or are required to have no
more than some maximum voltage difference between these rails. If this maximum
difference is violated, the chip can be damaged or even destroyed. In general
there are three (3) ways to sequence the turn on characteristics of multiple
converters.
The first method is to turn the converters on in a specific sequence with
either a control chip such as Summit Micro's SMH4804, or with discrete
circuitry. A simple solution is to have the output of one converter drive an
optoisolator that enables the second converter, and so on. In general most
sequencing requirements will want the lowest voltage to turn on first, and off
last. It is important to use an optoisolator to enable the other converter as
the enable is a primary referenced signal, while the output of a converter is a
low voltage isolated SELV signal.
Another method often used is to tie diodes between the different voltage rails
in a manner that while powering up the diodes will conduct, but when the
converter outputs are fully on, the diodes are reversed biased. For example a
diode between the 5V rail and the 3.3V rail, with the cathode connected to the
5V rail, will force the 5V rail to follow the 3.3V rail while turning on, but
once the 5V rail is at 5V, the diode will be reversed biased. This forces the
difference in voltage between the two rails to be no more than one diode drop
apart. Conversely, 3 diodes with a 0.7V drop in series from the 5V rail towards
the 3.3V rail will insure that the 3.3V rail is charged should it come up after
the 5V rail.
The last and most complex solution is to place FETs in series with each
converter's output, and enabling the FETs once the converters are fully turned
on. By carefully controlling the turn on of the FET gates, the voltage rails
can be brought up in strict adherence to any sequencing specification. Such a
solution can be built with discrete components or by using a specific
controller such as Summit Micro's SMT4004. One note of caution when
implementing these solutions: if the sense lines are connected on the output
side of the MOSFETs, the converter will not be able to sense its output voltage
at turn on until the MOSFETs are on. This will cause the converter to raise its
output until it reaches over voltage protection. You must either connect the
sense lines directly to the converters output and trim up to compensate for the
FETs on resistance, or add additional FETs to connect the sense lines after the
main FETs are enabled.
Any suggestions on using Summit Micro's hot swap
controllers?
It is important when defining a system's sequencing requirement that SynQor's
200ms initialization period be considered. As an example, if you are using a
programmable delay to enable multiple converters in a specific sequence, you
must make sure the delay is approximately 200ms. When using the Summit Micro
devices, make sure to use their 160ms delay setting, and use the second delay
tap to enable the first converter. As an example for a three converter solution
you use the SMH4804 where PG#2 would enable the first converter, PG#3 for the
second and so on.
^ back to top ^
Component Selection
How do I choose an input electrolytic capacitor (input
E-cap)?
SynQor has written an application note
on how to select an input electrolytic capacitor and why this capacitor is
needed. The paper can be downloaded from the Application Notes page of our
website. In general you will find that each converter will require a capacitor
that is approximately 47uF to 100uF with a minimum voltage rating of 100VDC,
and an ESR in the range of 0.5 to 1 Ohm.
How do I choose a CM Inductor?
Select an inductor that is rated for the desired input current. Leakage
inductance in the common mode inductor will provide reduction of differential
mode input current ripple. SynQor's recommended conducted EMI filters specify a
primary inductance of 360uH (0.36mH) and a leakage inductance of 3.5uH. A
popular choice for these inductors are Pulse Engineering's Self-Leaded SMT
Common Mode Chokes, such as P/N P0353.
^ back to top ^
General Application
Can I create negative voltages with SynQor converters?
All of SynQor's converters are fully isolated which allows them to generate
negative output voltages. Simply connect the Vout+ terminal to the system
output ground, and a negative voltage will be generated on the Vout- terminal.
I have a negative 48V bus. Can I use SynQor converters?
Because the converter is isolated, the Vin+ terminal can be tied to the input
ground of a -48V system. The Vin- terminal should then be connected to the -48V
rail. Keep in mind that the primary side signal pins are referenced to the Vin-
terminal in all cases.
How do I read the codes on the product label?
Each product has a label which reveals the product part number, revision letter
code and serial number for that module. This information can be found on the
label in the specific locations detailed in
this photo.
^ back to top ^
Safety Issues
Where do I find the safety certifications?
All of SynQor's safety certificates can be found
here.
Where are the conditions of acceptability?
All safety certifications, including the conditions of acceptability can be
downloaded in PDF form here.
How do I choose a fuse?
The Conditions Of Acceptability (COA) section of the safety file details the
type of fuse required for each converter. SynQor's product datasheets specify
the recommended fast blow external fuse to be used.
What is creepage and clearance?
Safety certification requires that the primary 48V circuitry (Primary) and the
low voltage output (Secondary) circuitry be separated from each other. SynQor's
converters use a Basic level of insulation, with 2000V of isolation. As such,
all SynQor converters have a minimum distance of 1.4mm between primary and
secondary circuitry.
What is "insulation rating" and "isolation
rating"?
Insulation refers to the design parameters used to isolate the 48V primary side
signals from the low voltage secondary side signals. This covers the
specifications for creepage and clearance between traces and components, as
well as the type of insulation used in the transformers. The transformers
transfer energy between the primary (Input) and secondary (Output) stages of
the converter. All SynQor converters use Basic Insulation which meets the most
stringent requirements for 48V based DC systems. Isolation refers to the
breakdown rating of the isolation stage, either in the transformer or between
isolated components such as optoisolators. SynQor tests and guarantees all
converters capable of withstanding a 2000V breakdown.
Are any dangerous voltages exposed on the surface of the
converter?
In the United States, UL defines an "Unsafe" voltage as 60V, this is
called the SELV (Safety Extra Low Voltage) limit. At input voltages of 0 to
60V, the highest voltage will be across the midbuss capacitors, this is an AC
signal, operating at the switching frequency, with a maximum peak to peak
amplitude of about 58V. Above a DC input of 60V the maximum voltage is the
input voltage across the primary switching FETS, which is approximately Vin
plus a few volts.
Are safety covers recommended for the converters or is a
"danger high voltage" or similar label required on the surface of the
converters?
If the input voltage remains below 60VDC, then all voltages are SELV, and no
special care is needed. If the input voltage rises above 60VDC, then there will
be non SELV voltages on the top side of the converter. The exact need for
warning labels or covers, depends on the safety agency approvals needed. For
UL, If non-technicians are to be allowed to change system configurations, i.e
add more memory cards, the machine should be powered off to prevent access to
the non-SELV voltages.
^ back to top ^
Performance Issues
How do I reduce output noise and ripple?
In general SynQor's converters have significantly lower output noise and ripple
when compared to traditional isolated converters. This is due to SynQor's
patented power topology, and excellent design practices. All of SynQor's ripple
and noise measurements defined in our data
sheets are made with 10uF of Tantalum and 1uF of Ceramic capacitance
across the converter's output. This is not required, but is used so that our
converter's measurements conform to standard industry practices. Further
reduction of output ripple and noise is accomplished through the addition of
both Tantalum and Ceramic capacitance. Ceramic capacitance provides attenuation
of output ripple, while the inherent ESR of tantalum capacitors provides
damping. It is possible to provide additional filtering through the use of LC
filters, but care should be taken when placing these filters within the sense
lines, as this can effect the converter's stability. Consult the factory for
further support on this topic.
What is SynQor's crossover frequency and phase margin?
All SynQor converters are designed to have adequate stability margin (at least
20dB of gain margin, and 50 Degrees phase margin) over variations of input and
output conditions, including input voltage range, output voltage range, output
current range, and load impedance. Performance is verified by using Bode plots
and examining gain and phase margin.
How much output capacitance can I drive with your
converter?
All SynQor data sheets define
the maximum allowable startup capacitance with a fully loaded output. If the
output load current at startup is reduced, then a larger amount of capacitance
than defined in the data sheet
is possible. An interesting benefit of SynQor's topology is that a large amount
of output capacitance has little effect on the converter's stability!
How do you trim a converter?
All SynQor converters have a trim pin, which allows trimming the nominal output
voltage higher and lower. In general, to trim a converter low, a resistor is
connected from the trim pin to the -Sense pin. To trim up, connect a resistor
from the trim pin to the +Sense pin. The value for the resistor is calculated
with the formulas provided in the converter's data sheets. You can also use the
Trim Resistor Calculator found in the Application
Notes section. The trim formulas match accepted industry standards for
half, quarter and eighth bricks.
Can I trim a converter actively?
Trimming or margining converters with an external active circuit is possible
and SynQor has several sample circuits available depending on your
requirements. Please consult your local SynQor
representative, or email
for details.
How does Over Voltage Protection (OVP) work; does it
track the trim?
SynQor's OVP protection is fixed and does not track the output voltage when
using the trim function. Customers should use care when trimming the converter
to insure the OVP is not activated during transient conditions, or when series
diodes or FETs are used on the converter's output. On SynQor's first generation
half brick converters containing HNA in the part number, the OVP protection is
measured across the sense lines. Care should be taken to insure that the output
pins do not become disconnected from the sense lines. If the output pins become
disconnected, and there is some capacitance across the sense lines, damage
could occur to the converter's output stage. This could happen if the output
pins are not properly soldered, or if MOSFET's are placed in series with the
output. On SynQor's new Kilo, Mega, Giga, and Tera half bricks, as well as the
QNA/QGA quarter bricks, the OVP sense point is located across the output pins
which avoids this condition.
Can I put converters in series?
SynQor's converters can be placed in series, however should one converter turn
off, a protection mechanism should be implemented to insure that the other
converter is disabled.
What are sense lines? Do I need to connect them?
Sense lines are used to compensate for resistive drops along the power
distribution path. The sense lines should be connected at the point of load, or
at a minimum connected to the output pins at the converter. If the sense lines
are not connected, the regulation and set point specifications contained in the
data sheet will not be met. No damage will occur if the sense lines are
left unconnected.
^ back to top ^
Full Feature Converters
What is the OR-FET signal used for?
The primary function of the full feature converters OR-FET signal is to supply
a voltage source higher than the converter's output voltage which can provide
current to turn on the gate of an N-Channel "ORing" MOSFET. The
signal has no implied "intelligence". ORing FETs are used to isolate
the output of a converter in the event that the converter's output experiences
a short fault. This is preferable to "ORing" diodes that have
significant power losses at high output currents. This signal is very low
power, and is not capable of supplying more than 50mW of power. Any control
circuitry drawing beyond 50mW of power should be driven off the converter's
main output.
The OR-FET pin can also be used as a Power Good signal: when the converter is
operating properly the OR-FET voltage will be much higher than the converters
output, providing a positive indicator of a converter's health even in a system
where converters are directly connected in parallel.
More details are available in our Full Feature
Application Note.
Can I use the Cshare pin as an output current monitor?
The current share pin will give a voltage that is proportional to the output
current, but only for a single module. If two converters are sharing a common
load by using the current share connection the voltage will represent the
average current of the two converters. This signal is referenced to the primary
side of the converter, so to interface with a controller the signal will have
to be brought across the isolation barrier through an optoisolator (unless of
course the controller is on the primary side, then a direct interface is
possible.) Any load impedance added to the CSHARE signal should be above
100kOhms. Loading on this signal will affect the current sharing performance.
More details are available in our Full Feature
Application Note.
How do I synchronize modules?
SynQor Full Feature converters have a pin to provide synchronization with an
external clock. The signal should be a 5V TTL level rectangular wave with a
duty cycle between 25% and 75%. The CSYNC signal is referenced to the Vin- pin
of the converter. When synchronizing different output voltage converters, you
should select the highest frequency specified for any converter as the common
frequency. Converters will not synchronize properly at frequencies below that
specified in their datasheet. More details are available in our Full Feature
Application Note.
Should I synchronize modules?
While Synchronizing converters may make EMI characterization and filter design
simpler, it can also cause converter harmonics to stack on top of each other,
creating a more difficult EMI problem to solve. Generally EMI specifications
require measurements be quasi-peak, and it is more beneficial to leave the
converters not synchronized. Certain systems require synchronization so that
the output ripple is at a single frequency. Applications such as wireless
communications equipment, systems with extremely fast clocks, or sensitive
optical circuits may find that the benefits of synchronizing the converters
output ripple outweigh the EMI benefits of having un-synchronized converters.
More details are available in our Full Feature
Application Note.
^ back to top ^
Parallel Converters
Can I parallel converters for more power?
SynQor's family of Kilo, Mega, Giga, Tera and some Peta half bricks can be
ordered with the full feature set which allows converters to be paralleled for
more power, as well as to take advantage of additional control features.
How do I connect SynQor's full feature converters in
parallel?
To connect full feature modules in parallel, you simply connect the current
share pins and the start sync pins of the sharing converters together. In
addition, make sure that the Vin+ and the Vin- pins are tied together, as the
Vin- pins provide a common reference for the current share signal. Outputs
should be connected together at a common point with the sense lines. SynQor has
an application schematic detailing these connections. Please refer to the Full
Feature Application Note for
this schematic.
How do I trim converters in parallel?
A trim circuit should be supplied for each individual converter, therefore each
converter should have a trim resistor. Make sure that all trim resistors are
the same value.
Where do I connect the sense lines for parallel
converters?
The sense lines of converters in parallel should be connected together at the
exact same point for balanced transient responses and the best output voltage
regulation.
^ back to top ^
Handling
How should I store and handle open frame converters?
Open frame converters can be damaged from poor handling, excessive mechanical
shock, or from a static electric discharge. The units should be:
-
Carefully handled and not subjected to mechanical stresses
-
Treated as an ESD sensitive component
-
Stored in a static protective container which physically protects the converter
-
The converters should not be stored in plastic bags, or stacked on top of one
another in any way
Are SynQor's converters water wash compatible?
SynQor converters are compatible with a water wash process, provided that the
converters are dry before powering them on. SynQor uses a no clean flux, and as
such the flux residue on our converter may react with other chemicals from the
manufacturing or wash process. Generally this manifests itself as a white
powder residue on the converter. This is usually benign, but any such residue
should be analyzed to confirm its reactivity. Customers using an active
detergent in the water wash process may need to apply a piece of kapton tape
over the SynQor label to prevent the label from fading.
What type of flux do you use?
SynQor uses either Alpha Metals or Koki no clean fluxes.
^ back to top ^
Reliability
What is SynQor's MTBF?
SynQor's standard reliability measure is Mean Time Between Failure (MTBF)
measured in hours of time. SynQor uses the Military Standard MIL217-F to
calculate MTBF based on ambient temperature and load. SynQor's standard
calculations are for 48V nominal input, 300LFM, and 80% load at 25C, 40C, and
55C. SynQor can also calculate MTBF using the Tellcordia (Formerly Bellcore)
Parts Count Method TR332. In addition, SynQor also measures the actual field
reliability in MTBF or FIT by product family. To request any of this
reliability information please submit a request to your
local SynQor sales representative, or email
. Some MTBF is also presented in the product
datasheets.
What are common failure modes?
The most common failure mode for units returned by our customers is NFF, or "No
Fault Found". To prevent the return of NFFs through SynQor's RMA process,
SynQor has evaluation boards available so customers can test the operation of
the converter before returning them. The evaluation boards allow for simple
testing and debugging.
What is SynQor's qualification process?
SynQor has a three-step release process: POD, POM, and Qualification. POD, or
Proof Of Design, is the process during which the converter's performance is
evaluated and characterized over all rated operating conditions and beyond, in
accordance with HALT principles. POD measures component stresses to insure that
design guidelines are met and that no components are over-stressed in both
normal and abnormal conditions. POD insures that long-term reliability and life
targets are achieved. Other tests performed at the POD stage include phase and
stability margins, thermal margin, capacitive load tests, destructive thermal
cycling, DFM, and waveform analysis.
POM, or Proof Of Manufacturing, insures that SynQor has designed a part that
can be manufactured, is reliable, and has proper margins to be run in a high
volume factory. POM insures that target yields are met, that our ATE systems
are optimized, and any new manufacturing equipment or processes constructed are
optimized. SPC analysis at this stage is mandatory, and CPKs are scrutinized to
insure that SynQor has a repeatable and consistent process.
Qualification is the final stage of product release. The purpose of the
qualification process is to insure that SynQor has designed and built a product
that exceeds our customers' expectations. Testing at this stage includes
aggressive thermal shock cycles, extended high temperature life and humidity
tests, vibration and mechanical shock tests in accordance with military
standards, terminal solderability, full mechanical and dimensional compliance,
as well as a thorough examination of the solder and manufacturing quality.
^ back to top ^
Troubleshooting
How do I measure output ripple?
Measuring output ripple and noise is extremely dependent on the test setup. A
good starting point for this is to use SynQor's Evaluation Board which can be
purchased from SynQor. Our evaluation board provides a BNC connector for
measuring output ripple. The industry standard noise measurement requires 20MHz
bandwidth limit which is detailed in our
data sheets. There is also an application
note titled, "Output Voltage Ripple Measurements," which
provides proper methodology and test set-ups to achieve accurate output ripple
readings whether you are using our evaluation board or an application board.
This paper also describes how using a traditional probe with a ground clip will
allow radiated noise to couple into the probe creating a false measurement,
which manifests as large high frequency spikes. These are largely measurement
artifacts which are most often caused by two effects: magnetic coupling through
a large scope ground loop, and/or transmission line effects on unterminated
cables. SynQor's application note
on this topic will show you how to reduce the measurement circuit loop to a
minimum, almost eliminating magnetic noise pickup and allowing a more accurate
measurement of the converter's output voltage ripple.
How do I measure Output Load Current?
You can easily measure the output load current of any SynQor open frame
converter. The application note Output
Load Current Calculations outlines the equations and other information needed
to make this measurement.
^ back to top ^
Thermal
How do I choose a heatsink?
Choosing a heatsink requires knowing several things:
-
The output power required.
-
The power dissipated by the converter at the load current and input voltage.
-
The worst case ambient temperature.
-
The amount of airflow available.
-
The maximum baseplate temperature of the converter, which is 100C for all
SynQor baseplated converters.
Example:
What size heatsink is required for 3.3V quarterbrick at 20A, 48V input, 70C
ambient, and 200LFM of airflow?
Using SynQor's PQ48033QGA25 data sheet figure 3, at 48V input, and with a 20A
load, the power dissipated by the SynQor's 3.3V quarterbrick converter is about
8.5 Watts. The allowable temperature rise is simply the maximum baseplate
temperature minus the worst case ambient temperature. So in this example, 100C
- 70C = 30C allowable rise. The thermal impedance of the heatsink required is
simply the allowable temperature rise, divided by the power dissipated. In our
example this is 30C/8.5Watts, giving a maximum thermal impedance requirement
for our heatsink of 3.53C/W at 200LFM. Using any standard heatsink catalog you
can now select a heatsink. A good choice would be Intricast's (www.intricast.com)
HS1361XM01 heatsink, which is 0.5" tall and has a thermal impedance of 2.60 C/W
at 200LFM.
How do I measure the temperature on your converters?
Where do I place thermocouples?
SynQor's derating curves are created based on a derating of 125C for switching
MOSFETs. Using SynQor's data sheet,
locate the thermal image, which shows the hottest components on the power
supply. Generally you will want to measure the input MOSFETs, the isolation
stage MOSFETs, and the output synchronous rectifiers. Care should be taken not
to block too much of a device's surface area, as this could significantly
reduce its ability to remove heat via convection.
How does SynQor's thermal protection work, and where is
the sensor?
On SynQor half bricks, a thermistor located on the topside of the PCB drives a
comparator circuit. On the quarter brick converters there is an integrated
temperature sensor that will trip the converter off if an over-temperature
condition exists. Both methods work by sensing the PCB temperature. Email
for location details.
Do you have flotherm models?
SynQor has basic thermal models that can be used for customers to develop
flotherm models, but due to the use of multiple power handling devices to
spread the heat generation across the board, it is very difficult to provide a
detailed thermal model of the open frame converter.
What is the optimal orientation of the converter?
In general, half bricks are less sensitive to orientation than quarter bricks.
For half brick converters 2.5V and below, it is best to have the air flowing
towards the output terminals, for half bricks 3.3V and above it is better to
have the airflow towards the input side of the converter. On quarter bricks and
eighth bricks, it is always best to make the airflow perpendicular to the long
side of the converter as this exposes the maximum surface area to a cooling
airflow. Although the differences are minor, it is best practice to review the
derating curves and thermal images found in the
data sheets to determine optimal orientation.
What is the difference in air speed measurements of CFM
vs. LFM?
Designers often need air speed measurements to calculate thermal derating and
power dissipation for their DC/DC converters and for their overall systems.
There are two basic units of measure: CFM (cubic feet/minute) is a measurement
of volume, LFM (linear feet/minute) is a measurement of velocity. Fan
manufacturers use CFM because they rate their fans according to the quantity of
air they can move. Velocity (speed) is more meaningful to heat removal at the
board level. This is what most DC/DC converter manufacturers will specify when
calculating thermal derating curves and other performance specifications. To
convert CFM measurements to LFM, use the following equation:
LFM = CFM/AREA
LFM = Linear feet per minute of airflow
CFM = Cubic feet per minute of air volume
AREA = the area of the opening in square feet.
For example, let's assume you are blowing air through a 6" x 6" opening across
the top of a DC/DC converter with a 100CFM, unobstructed fan.
LFM = 100/ 0.25 sq feet or about 400LFM calculated.
The most accurate way to measure actual air speed is with an anemometer.
Some manufacturers specify airflow in Linear Meters/Second. Use the table below
to convert feet/minute into meters/second:
100 f/m = 0.5 m/s
200 f/m = 1.0 m/s
300 f/m = 1.5 m/s
400 f/m = 2.0 m/s
500 f/m = 2.5 m/s
How do I attach heatsinks?
Use M3 screws with a length no longer than 0.125". Tighten the screws to a
torque of 6.0 in-lb. Minimum screw length is 0.10" for adequate hold. For
more details, please refer to the appropriate
data sheet addendum for baseplated modules, located in the products
section.
What type of potting material do you use?
In the baseplated versions of its modules, SynQor uses a proprietary potting
material. Contact your local SynQor representative,
or email
for details.
^ back to top ^
|
Layout
© Copyright 2009 SynQor