10Ghz Transverter
10Ghz Transverter
10.368 GHz Transverter VK3XDK
(These Transverters work very well at 10.240Ghz!!)
The Transverter series was deigned for good performance and ease of build/Faultfinding. The Modules are kept small and have facilities for bias tees allowing remote mounting. All units have intentionally been kept to a minimum with an "add sections to suit" philosophy. (i.e. RF preamp/filter, IFamp/filter).
RX
The 10.368 GHz RX section features a NE32584 HJfet on the input, biased at 10mA and about 2V drain-source. The FET uses microstrip matching on the input and output for a gain of around 10-14db and Noise figure of less than 1db.
The second stage uses a Minicircuits MMIC era3 with 3.5V (35mA) fed to its output. These MMICS feature reasonable gain at 10368 GHz (around 10db?) and a fairly low Noise figure.
Following the second amplification stage is a microstrip (edge coupled) filter with a band pass of less than 400 MHz (loses of less than 4db) and a Hittite hmc220MS8 Double balanced mixer. This particular mixer likes an LO drive of over 7dbm for low conversion loses (10-13dbm Mixer).
With an LO input of >7dbm at 9936 MHz, Overall (conversion) gain on the receive section has been found to be over 10db (Result will vary board to board)
TX
The 10.368 GHz TX section also uses the hmc220MS8/edge coupled filter combination.
Following the filter is an ERA3 MMIC with 3.5V (35mA) on its output. This device has good gain (around 10db) and is easily obtainable.
The final output stage Uses an RFMD nlb310 MMIC with 4.6V (50mA) on its output. This device features good gain and good output at 10.368 GHz
With an LO drive of >9dbm and IF drive of 0dBm at 432 MHz, output has been found to be over 7dBm
With an LO drive of >9dbm and IF drive of 5dBm at 432 MHz, output has been found to be over 12dBm (Results will vary board to board)
LO image typically better than -40dbc @12dbm output.
LO rejection is the biggest variable board to board, with a bit of experimenting an image of better than -55dbc should be obtainable.
(info regarding optimising LO rejection will be in "construction and general help" soon)
CAPACITORS
All coupling capacitors are 0.8pf ATC100 (10.368 GHz series resonant)
Other capacitors (not so critical) are standard smd (0805) types and 10uf/16V tantalums.
RESISTORS
0805 and leaded
OTHER COMPONENTS
NE32584 (NEC) HJfet biased at 2V 10mA.
NLB310 (RFMD) MMIC (Good gain, good output)
ERA3 (MINICIRCUITS) MMIC (Good gain, reasonable noise figure)(X2)
HMC220MS8 (HITTITE) 10-13dbm double balanced mixer (x2)
ICL7660 Negative voltage generator (HJfet bias)
SMA Connector Edgemount (X4)
MCX Connector Topmount (X2)
Zener Diode (4.6-5.6V) Lowers voltage to ICL7660
PCB
The substrate is glued to an aluminum carrier (using silver epoxy or standard epoxy (ie araldite) if low frequency/dc grounding screws are used)
VK3XDK Transverter (construction notes)
The following is a guide to construction. (I use the same techniques on all transverters)
Use very fine solder and use sparingly!
Tweezers, Liquid flux (as well as fluxed core solder), fine iron and magnifying glass are recommended
PCB MOUNTING.
Using a strong solvent, clean both the PCB and aluminum carrier thoroughly prior to gluing.
Mix just enough silver epoxy (or standard “araldite” with mounting screws) for an even thin spread across the aluminum.
Press the PCB down onto the glue and use a bit of gentle rotation to sit the PCB flat and even on the epoxy.
Place assembly on a flat even surface (under a bit of weight) until the epoxy is at least half set.
Clean up mess using a strong solvent (I use paint thinners, but acetone should work).
PCB SOLDERING.
Clean top surface ready for soldering.
MCX connectors need their centre pin trimmed to sit flat (surface mounted), SMA center pins may also need shortening (depending on track length)
Solder on MCX connectors (do this first as it creates a bit of heat which could damage the fragile mixers)
-Sit the connector in position and solder the centre pin to its pad first; this will hold the connector in place while soldering the body!
-Use a hot iron (or two small irons) and focus most heat into the MCX body!
Solder on mixers quickly and carefully (The mixers are FRAGILE and hard to fault find. BE CAREFUL!)
-Very fine solder, very fine iron tip, magnifying glass and some brush on liquid flux makes this job a lot easier (hardest part of the board construction)
-Soak up solder bridges ect. With solder wick (be gentle, don’t use too much heat!)
-Double check soldering with a magnifying glass!
Solder on SMD's (semiconductors last)
-First melt a bit of solder on one side of the PCB pad only, use tweezers to position the component then re-melt the solder to secure the component in place.
-There are several paralleled components (I align the two next to each other on the bench. Then using tweezers grip together and solder in the above manner!)
-Liquid flux helps lots with this type of soldering/resoldering
Form and trim leaded components for surface mounting
SMA CONNECTORS.
Temporarily solder the SMA's in position (center pin only!) Once mounted, the bodies can be rotated (gently) for alignment.
Mark out SMA hole positions using fine tip marker
Remove SMA’s
Center punch and drill to suit mounting screws (I use small self tapers)
-Be very careful while drilling into the edge of the aluminum (for obvious reasons!)
Refit SMA’s and solder permanently.
BEFORE POWERING UP
Attach dc supply wires for +8V RX and +8v TX (boards are designed for 8vdc and ARE NOT REGULATED!)
-You will need to supply an "off board" Regulated +8Vdc supply rail.
-If you don’t have any 7808's, the simplest way is to "UP" a 7805 or similar TO-220 Flat Pack regulator.
Check all work! Check for shorts!
And then.....check all work AGAIN (After 20 or more boards like this, I still make silly mistakes)
Adjust 20K trimmer to about 800ohm to Earth, this should supply enough "safe" bias for the HJFET for the moment!
RX POWER UP
Low noise FETS can be unstable, although I've not had problems with these designs, it is good practice to apply 50ohm loads to all RX ports before power up.
Adjust Bias trimmer for 2.0V Drain-Source
Apply an Oscillator Injection Signal of appropriate level (typically +6 to +7 dBm) into the Multiplier. If you have access to a Spectrum Analyser or Power Meter, check the final Multiplier Output level. There should be greater than +6 dBm on each Port.
Connect multiplier output to "LO" input of Transverter.
Check IF output, you should hear the IF Noise floor rise!!
If you have access to a Sig Gen, apply a relatively “loud” RF Signal of say -80 dBm and check the IF Output. For the most part, Snowflaking will NOT be necessary but you may recover a few extra dB of Sensitivity.
There should be at least 8 dB of Conversion Gain for these transverters. If not, check the DC levels (marked on schematic). If these are all OK, there may be a problem with the Mixer. They are fragile and cannot easily be tested. Keep an eye out for stability issues although from past experience, there shouldn’t be any. However, if you suspect instability, check all grounding etc. The SMD capacitors or resistors could also be damaged.
As a last resort, try a bit of microwave absorbent material.
TX POWER UP
Apply the Oscillator Injection and IF Injection signals of appropriate levels into the Multiplier and IF Ports. These are typically +6 to +7 dBm for both Oscillator and IF.
There should be at least +7 dBm Output available but greater than +10 dBm is more typical with the correct Injection Levels applied. If not, check your DC levels (marked on schematic). If these are all OK then you may have a suspect Mixer. They are fragile and cannot easily be tested.
Keep an eye out for stability issues although from past experience, there shouldn’t be any. However, if you suspect instability, check all grounding etc. The little SMD capacitors or resistors could also be damaged.
As a last resort, try a bit of microwave absorbent material.
FINDINGS AND CHANGES
Snow Flaking
Although Snowflaking isn’t always required, there is often some dbs to be picked up!
RX
On the 10.368 Transverter a (2mm by 2mm) snowflake at about 2mm in front of the ne32584 gate will often pick up some dbs
TX
On the last element of the TX filter (closest to the ERA3) a (3mm by 2mm) snowflake has been shown to increase output by a few dbs.
This snowflake closes the gap between the 5th and 6th filter elements and is soldered about halfway along the bottom section of the “L” shaped element.
Im hoping to have some photos available for more clarity soon.
ICL7660 shunt Zener
The current design uses a shunt Zener (through a 1.2K resister) to drop the voltage to the ICL7660. This has been shown to cause a slow bias for the ne32584 (due to a r-c time constant with the 10uf capacitor).
This hasn’t yet been a big problem, but I think changing the series 1.2K resistor to about 560R would be worthwhile.
VK3XDK Transverters and multipliers (Test results)
With Gippstech (VK3 microwave conferrence) coming up on the weekend, I quickly put together 3 kits (In one day!) for display at the event.
3.4 GHz Transverter and multiplier
5.76 GHz Transverter and multiplier
10.368 GHz Transverter and multiplier
Bear in mind these kits were built quickly and results will vary for each kit construction (better or worse)
All dbm measurements are taken with a 50mm rigid coax section and a sma to female n adaptor connected between the actual port and a Boonton 4B power head.
10.368 GHz Transverter TX (RX not yet tested on this board) Without Snowflaking
Transverter Band pass found to be (forgot to test!!)
Tested with an LO injection of 8dbm @ 9936 MHz (10dbm would be better but my signal generator wouldn’t get there!)
Power out results,
0dbm @ 432 MHz (IF) for 5dbm Output @ 10368 MHz
5dbm @ 432 MHz (IF) for 8dbm Output @ 10368 MHz
10dbm @ 432 MHz (IF) for 9dbm Output @ 10368 MHz
LO image better than -34db with 9dbm output.
It was found LO rejection with this board and the 5.7 GHz Transverter board was very sensitive to surroundings. (Microwave absorb and or careful mounting in their final enclosure will help LO rejection).
Tested with 10224 MHz LO (for 2M IF) found to be unsuitable!
There may be improvements with a higher level LO? And/or microwave absorb/ tuning??
10.368 GHz Transverter TX (RX not yet tested on this board) With Snowflaking
Transverter Band pass found to be (forgot to test!!)
Tested with an LO injection of 8dbm @ 9936 MHz (10dbm would be better but my signal generator wouldn’t get there!)
Power out results,
0dbm @ 432 MHz (IF) for 7dbm Output @ 10368 MHz
6dbm @ 432 MHz (IF) for 10dbm Output @ 10368 MHz (unusual, 12-13dbm is normally easy to obtain!!)
LO image better than -38db with 10dbm output.
It was found LO rejection with this board and the 5.7 GHz Transverter board was very sensitive to surroundings. (Microwave absorb and or careful mounting in their final enclosure will help LO rejection).
Tested with 10224 MHz LO (for 2M IF) found to be unsuitable!
There may be improvements with a higher level LO? And/or microwave absorb/ tuning??