New Mixer
Developments
Millimeter Wave Mixers
Oleson Microwave Labs
(OML) introduced a line of single diode, unbalanced mixers covering all
waveguide bands from 18 to 325 GHz in November, 1996.
In February, 1997, Tektronix contacted OML regarding this type of
product. Tektronix had decided to
discontinue the manufacture of millimeter wave harmonic mixers including the
their 782 series of mixers. Tektronix
inquired as to whether they could arrange to recommend to their customers the
use of the OML mixers in replacement of the Tektronix mixers.
At that time, OML forwarded data and procedures to Tektronix for their
examination. A short while later the Tektronix field sales force began to send
customers to OML to purchase millimeter wave harmonic mixers.
The OML data sheet for
the unbalanced mixers did not address conversion loss, but rather specified
sensitivity in the form of “equivalent average noise level” in the same
manner as had Tektronix for their 490 and 780 series of mixers.
Included with the OML Millimeter Wave Mixers data sheet is a Test Procedure detailing the test set up and methodology used to
check the mixers functionality. A
frequency vs. conversion loss plot taken with this “scalar” type test set
was included with each mixer shipped. Each
mixer was also tested at several different frequencies to ensure that it met the
“equivalent average noise level” specification. A “Certificate of Conformance” was included with each mixer to
attest to the compliance of that mixer to the OML specifications.
Early versions of the OML data sheet (current revision is “J” dated
10-14-97) refered to the fact that a more advanced test set was being developed
and that more specific test data would be available as an extra cost option in
the future. Much work has now been completed and significant insight in the
testing of these mixers has been obtained (more details be given later in this
paper). As a result of these
investigations, the OML single diode harmonic mixer will continue to be
specified with an “equivalent average noise level” specification, conversion
loss will not apply as a specification.
During the above investigations OML discovered that the manufacturer’s specifications applying to the spectrum analyzer interfaces serving the harmonic mixer have been typically quite loose with considerable variation experienced from one spectrum analyzer to another for a given manufacturer and model. In previous generations of spectrum analyzer design the physical properties such as L.O. power variation across the band and the inexact manner in which adjustment was made to the bias made it virtually impossible to set up a test emulation for many of those spectrum analyzers. The single diode unbalanced harmonic mixer will become OML’s utility type of mixer with use for spectrum analyzers lacking the precision interfaces for harmonic mixers and for general scientific use. A new two diode, balanced even harmonic mixer design is being phased in as OML’s premium spectrum analysis mixer.
New
Products
OML will begin to accept
orders for two diode balanced even harmonic mixers starting June 1, 1998
(earlier if possible) for delivery scheduled 60 days later. Initially mixers will be available covering only the following
bands: WR-42 (18 to 26.5 GHz), WR-28 (26.5 to 40 GHz), WR-22 (33 to 50 GHz),
WR-19 (40 to 60 GHz). The design
of the mixers has been completed and final characterization is in process.
Additionally, OML has completed it’s new harmonic mixer test set with
operation to 60 GHz. This new test set
will be employed to produce final test data for each mixer where the spectrum
analyzer type has been specifically indicated on the purchase order.
This testing and deliverable test data will be included in the individual
mixer price. In the same manner as the
single diode unbalanced harmonic mixer, these new two diode balanced even
harmonic mixers begin to yield best conversion loss performance when the L.O.
level is above +12 dBm. The maximum L.O. input level is +17 dBm.
The flattest conversion loss will occur when the L.O. is maintained at a
level value, with no more variation than +/- 1.5 dB.
OML testing is accomplished with +/- 0.8 dB.
Data derived from this
testing will be in graphic format. The
graphic data will consist of a frequency vs. conversion loss plot representing
401 data points equally spaced across the specific waveguide band.
An example data sheet is attached. A
tabular data option will consist of a listing of frequency vs. conversion loss.
The number of tabular data points can be specified by the purchaser as:
51,101,201 or 401 points. The 401 point
tabular data can be supplied in addition to the graphic data at no extra charge.
The test set is calibrated based on the number of test points. A number of test points other than 401 will require recalibration
and thus an extra charge. The test set
has been designed to emulate different specific spectrum analyzers.
The data sheet, either graphic or tabular, will identify the specific
type of spectrum analyzer being emulated.
The following three OML
emulated specification items will be specifically identified on the data sheet
as there most likely will be variances from unit to unit of that type of
spectrum analyzer: 1) the input test
power, typically -30 dBm, 2) nominal L.O.
power used, as specified by the spectrum analyzer manufacturer,
3) the specific current limited bias voltage applied (if bias is used as
is the case when operating the balanced even harmonic mixer as an odd harmonic
mixer or when using the single diode unbalanced mixer).
The test set R.F. and L.O. sources are operated in the leveled mode with
variation of less than +/- 1 dB across the band of interest.
The outer limits of the variance of these two power sources will be
measured and their impact will be taken into account in the “uncertainty”
data along with other contributors that can be identified. One source of “uncertainty” is the cable used to connect the
harmonic mixer to the spectrum analyzer. OML
has selected a one meter long high quality cable manufactured by Semflex as the
standard cable for the test set. A plot of the cable insertion loss and VSWR is attached.
The cable part number is S119BRFSS10039. If
the manufacturer plans on supplying its own cable with the mixers to their
customers, the cable supplied must be equivalent in performance to this Semflex
cable for the OML data to be valid. This
Semflex cable is optionally available from OML. The
final “uncertainty” data will be listed on the data sheet.
The test set uses a
millimeter vector network analyzer (VNA) as single frequency (I.F.), calibrated
power measurement receiver. The VNA
processor is used for controlling a large portion of the spectrum analyzer
emulation. Files on the VNA hard disk
will contain the frequency range, multiplier (n), power level and step size for
the R.F. and L.O. synthesizers. Such a
file will be dedicated to each waveguide band for each specific spectrum
analyzer to be emulated. The VNA system
to be employed consists of current model Anritsu Lighting VNA’s and El Toro
Synthesizers. A phase locked Anritsu or
HP source with excellent power stability will be used as the absolute power
reference in the system. Power levels will be measured using Anritsu and HP power meters.
All equipment carries current calibration.
Future
Products
OML is currently
developing two diode even harmonic mixers to cover the waveguide bands from 50
to 110 GHz. Until that development is
completed the OML single diode mixers will be the only available product, sold
subject to the conditions previously described. Two diode, even harmonic mixers covering:WR-15 (50 to 75 GHz),
WR-12 (60 to 90 GHz), WR-10 (75 to
110 GHz) are scheduled to become available in the September time frame.
In addition to the completion of the mixer designs, OML must finish the
construction of multiplier assemblies to extend the frequency range of our
current 60 GHz synthesizer. The
multipliers themselves have been OML products for over four years. The
element needed is the completion of a high quality R.F. signal power leveling
system. OML’s goal is to achieve a R.F. power flatness of +/- 1 dB over
each of the above waveguide bands.
Emulation
Data Base
In order to support you, the manufacturer, and your customer, certain data is needed to properly configure the OML harmonic mixer test set. OML has determined that each one of the following items must be included to achieve a credible emulation for each waveguide band. The items number shown here correspond to the number blanks on the Emulation Data Sheet:
1) L.O. frequency range 2) Nominal L.O. power 3) L.O. equation, i.e., = (n +/- I.F.) 4) Does the user have choice of harmonic to be utilized? if so what are the choices? 5) I.F. frequency 6) Bias voltage and current range 7) Bias source impedance (very important) 8) Is the bias indicated on the analyzers readout?in voltage or current? 9) Is the diplexer internal or external? 10) Any other specific conditions that should be observed.
You can download this application
note, including a form,
that you
are requested to complete and return by FAX. All
information will be held confidential. This information is needed so that our emulation of your spectrum
analyzer products will be as correct as possible. This
will allow OML to support you and your customers with a device that will enhance
your product. Without this manufacturer
confirmed information from you OML will have to construct it’s emulation based
on catalog data and information gained by using the actual spectrum analyzer
when one becomes available. The
1) L.O. frequency range, can
usually be found in the manufacturers catalog but should be confirmed. The 2) Nominal L.O. power,
is very important as the catalog data may list a minimum power
specification that is significantly lower then the “in factory” or typical
power level. OML cannot emulate a power
characteristic that changes across the band, such as a downward slope.
Our L.O. synthesizer is specified by the manufacturer to be flat +/-0.8
dB, a specification that we have found it meets through periodic verification.
OML must use a single value that represents the average L.O. power output
of the typical spectrum analyzer of that model, so that the data generated will
be as valid as possible. The
3) L.O. equation, will give our
emulation the correct multiplier number and tell us if the L.O. is above or
below the R.F. Some spectrum analyzers
allow the user the 4) Choice of harmonic
to be utilized. If your spectrum analyzer
has this feature, OML needs to know the default or preferred harmonic. Also some of your customers may want to use a more optimum harmonic
over a narrow portion of the band so we need to know if your spectrum analyzer
allows this. OML must know the
5) I.F. frequency, for it’s
emulation. The adjustment limits of your
spectrum analyzers 6) Bias voltage and
current must be known. For OML to provide
correct bias information to the customer 7)
The bias power supply
Price
and Delivery
Price and delivery
quotations for WR-42, WR-28, WR-22, and WR-19 mixers are available at this time.
The quoted prices will include the testing outlined above.
Orders will not be accepted until the requested spectrum analyzer
emulation data has been received. The
quote will not include the test cable unless so requested.
OML desires to support the manufacturers effort in every manner possible.
Please do no hesitate to contact us with any questions that you might
have.
For a quote or with
questions from outside of the United States, please contact:
Radar Systems TechnologyTel # 650 969 5534 FAX
# 650 969 7078.
300 Digital Drive, · Morgan Hill, CA 95037
Tel. (408) 779-2698 · FAX: (408) 778-0491 · email: info@omlinc.com
|
Home | Introduction | Vector
| Sources | Scalar | Spectrum
| Power |
©
1999-2006 OML, Inc. (Oleson Microwave Labs.) All Rights Reserved.