Difference between revisions of "800 MHz Receiver"

From W9CR
Jump to navigation Jump to search
 
(9 intermediate revisions by the same user not shown)
Line 23: Line 23:
 
There is a image filter (FL2030) which filters the input to the double balanced mixer to prevent the RF radiating out the antenna port.
 
There is a image filter (FL2030) which filters the input to the double balanced mixer to prevent the RF radiating out the antenna port.
  
<gallery heights="200px" widths="200px" mode="packed-hover">
+
<gallery heights="300px" widths="300px" mode="packed-hover">
 
800 MHz Receiver - Mixer filters.png | 800MHz Receiver Image Filters
 
800 MHz Receiver - Mixer filters.png | 800MHz Receiver Image Filters
 
800 MHz RX Image Filter 2.jpg| 800 MHz RX Image Filter
 
800 MHz RX Image Filter 2.jpg| 800 MHz RX Image Filter
Line 32: Line 32:
 
In the second 800 MHz board I looked at, (TRF6551G39) there is another filter (FL2300) which is on the input to the mixer from the Local Oscillator output.  This is not present in the first board I converted (TRF6551J), and just has a 50 ohm line in place of it.  It remains to be seen what this filter does.  It needs to be removed and swept for response.
 
In the second 800 MHz board I looked at, (TRF6551G39) there is another filter (FL2300) which is on the input to the mixer from the Local Oscillator output.  This is not present in the first board I converted (TRF6551J), and just has a 50 ohm line in place of it.  It remains to be seen what this filter does.  It needs to be removed and swept for response.
  
<gallery heights="200px" widths="200px" mode="packed-hover">
+
<gallery heights="300px" widths="300px" mode="packed-hover">
 
800 MHz Receiver - Mixer filters.png|800 MHz Receiver TRF6551G39 - Image Filter
 
800 MHz Receiver - Mixer filters.png|800 MHz Receiver TRF6551G39 - Image Filter
 
800 MHz Receiver TRF6551J - 0305.JPG|800 MHz Receiver TRF6551J - Image Filter
 
800 MHz Receiver TRF6551J - 0305.JPG|800 MHz Receiver TRF6551J - Image Filter
Line 54: Line 54:
 
The VCO in the 800 differs from the 900 unit in that the 800 VCO does not have a soldered on shield over it.  I'm not sure what difference this makes in practical use.  Note on the TRF6551J unit there is a solder mask which would fit a shield.   
 
The VCO in the 800 differs from the 900 unit in that the 800 VCO does not have a soldered on shield over it.  I'm not sure what difference this makes in practical use.  Note on the TRF6551J unit there is a solder mask which would fit a shield.   
  
<gallery heights=200px widths=200px mode="packed-hover">
+
<gallery heights=300px widths=300px mode="packed-hover">
 
900 MHz RX VCO shield.jpg|900 MHz RX VCO shield
 
900 MHz RX VCO shield.jpg|900 MHz RX VCO shield
 
800 MHz Receiver - 0252.JPG|800 MHz VCO no shield
 
800 MHz Receiver - 0252.JPG|800 MHz VCO no shield
Line 133: Line 133:
 
As this is a low side injection receiver the VCO runs 73.35 MHz below the intended signal.  This is different from the high side injection systems used on VHF and UHF.  Low side injection is less of an issue on 800/900 as traditionally there are no users on the image frequencies (+73 MHz) of any real ERP to cause interference.
 
As this is a low side injection receiver the VCO runs 73.35 MHz below the intended signal.  This is different from the high side injection systems used on VHF and UHF.  Low side injection is less of an issue on 800/900 as traditionally there are no users on the image frequencies (+73 MHz) of any real ERP to cause interference.
  
Picture of VCO goes here, and diagram of the VCO
 
  
Mask off the VCO using Krapton tape and/or aluminum foil.  The electrolytic caps don't do well with excess heat, especially as most are 10-15 years old at this point.
 
  
Setup the board with an under board pre-heater set to 460f. This is important as you want the board to warm up to close to the melting point of solder. Then the hot air wand will heat it the rest of the way.
+
* Mask off the VCO using Krapton tape and/or aluminum foil.  The electrolytic caps don't do well with excess heat, especially as most are 10-15 years old at this point.
 +
 
 +
<gallery widths=300px heights=300px mode="packed-hover">
 +
800 MHz Receiver vco diagram.png|800 MHz Receiver vco diagram
 +
800 MHz VCO taped off.JPG|800 MHz Receiver VCO taped off for desoldering
 +
</gallery>
  
[[File:Board_Preheater.jpg|thumb|Underboard pre-heater]]
+
* Setup the board with an under board pre-heater set to 460f. This is important as you want the board to warm up to close to the melting point of solder.  Then the hot air wand will heat it the rest of the way.
 +
<gallery widths=300px heights=300px mode="packed-hover">
 +
Board_Preheater.jpg|Underboard pre-heater
 +
</gallery>
  
Once you have this preheated for a few minutes, use the hot air wand with no tip 15 L/m and 650f on the coax line (L2202).  Move it in a circular pattern while waiting on it to reflow.  Once it's flowed, pull the resonator up and allow it to cool on an insulated pad.  You don't want to thermal shock it, as it will crack being it's ceramic.
+
* Once you have this preheated for a few minutes, use the hot air wand with no tip at 15 L/m and 650f on the coax resonator (L2202).  Move it in a circular pattern while waiting on it to reflow.  Once it's flowed, pull the resonator up and allow it to cool on an insulated pad.  You don't want to thermal shock it, as it will crack being it's ceramic.
  
 
Careful you don't disturb any of the VCO parts, as they will most likely be reflowed. If you do, don't panic, just look at the before picture and put everything back to how it was on the VCO. It's a pretty easy circuit to figure out.  
 
Careful you don't disturb any of the VCO parts, as they will most likely be reflowed. If you do, don't panic, just look at the before picture and put everything back to how it was on the VCO. It's a pretty easy circuit to figure out.  
  
While the VCO is still Warm, remove L2201 and set asside.  This will allow us to power the VCO with out powering the rest of the receiver.
+
* While the VCO is still Warm, remove L2201 and set asside.  This will allow us to power the VCO with out powering the rest of the receiver.
 +
 
 +
* Change the following parts in the VCO area.
 +
 
 +
{| class=wikitable
 +
|+ style="text-align: left;" | 800 to 900 parts differences.
 +
!Part number!!800!!900!!Notes
 +
|-
 +
|C2207||3.3 pF||2.5 pF||INFO ONLY don't change, steering range
 +
|-
 +
|C2211||12 pF|| 6.8 pF||Tank
 +
|-
 +
|C2209||8.2 pF||5.6 pF||Tank
 +
|-
 +
|L2202||430 mils||350 mils||Resonator
 +
|-
 +
|-
 +
|||||||
 +
|-
 +
|R2815||2100||4700||ID resistor
 +
|}
 +
 
 +
<gallery widths=300px heights=300px mode="packed-hover">
 +
800 MHz Receiver VCO layout.png|800 MHz Receiver VCO layout
 +
</gallery>
  
Now the VCO should be exposed and you can tack on some test wires to it while it cools down.
+
* Now the VCO should be exposed and you can tack on some test wires to it while it cools down.
  
PUT TEST POINT PICTURE HERE
+
* Hook up to the VCO as shown.
  
Hook up to the VCO as shown, we'll do the Low VCO first.
+
<gallery widths=300px heights=300px mode="packed-hover">
 +
800 MHz Receiver VCO test points.png|800 MHz Receiver VCO test points
 +
</gallery>
  
 
* 10.0 V power to VCO on P2200 Pin #3
 
* 10.0 V power to VCO on P2200 Pin #3
 
* 0-12 V on the steering line, P2201 #1
 
* 0-12 V on the steering line, P2201 #1
* output probe near the C2219 VCO output stripline.  This doesn't need to be tacked on, just a probe near it.  
+
* output probe near the C2219 VCO output stripline.  This doesn't need to be tacked on, just a probe near it so you can pickup the output on the spectrum analyzer.  
 +
 
 +
==== Cutting down the Coaxial resonator ====
  
 +
*The next step is to cut the coaxial resonator (L2202) down from the back side (the shorted, non leaded side) till it's 0.350" long.
 +
You will need a diamond grinder to do this. 
  
MODIFY THE VCO Inductor
+
Once it's 350 mils long (5 mils longer is ok, shorter will require more trimming of C2205), fashion a copper tape strip and apply over the end of the resonator.  This will need to be soldered in place around the edge and then soldered to the inside center conductor.  This replaces the shorted out end which was ground off. 
CHANGE PARTS PER TABLE
 
  
=== Tuning VCO ===
+
<gallery widths=300px heights=300px mode="packed-hover">
 +
Resonator trimmed and shorted.png|Resonator trimmed and shorted
 +
</gallery>
  
Tuning is needed to ensure we cover the proper ranges with the VCO.  If we can't align the station due to the VCO not being locked in the commercial band, being locked in the ham band does not count.
+
* Reinstall the coaxial resonator
 +
 
 +
<gallery widths=300px heights=300px mode="packed-hover">
 +
800 MHz Receiver VCO resonator installed.jpg|800 MHz Receiver VCO resonator installed
 +
</gallery>
 +
 
 +
To arrive at the correct lengths the following measurements and parts changes were tried.
 +
 
 +
{| class="wikitable"
 +
!!Parts changed !! Resonator length (mils) !! 0.5V !! 5V !! 9V !! Delta
 +
|-
 +
|Resonator only||390||739||766||785||46
 +
|-
 +
|C2211 & C2209||390||755||786||807||52
 +
|-
 +
|C2211 & C2209||350||785||820||841||56
 +
|-style="background:Pink"
 +
|C2211/C2209 & trim C2205||350||809||849||874||65
 +
|}
 +
 
 +
It was found a combination of changing C2211, 2209 and grinding the resonator to 0.350" got the VCO to within 10 MHz of where it needed to be to cover the entire ham band and the commercial band just below it.  This does make the coverage of the receiver larger than before, which is aventagious for amateur use, as it now covers the entire ham band.  Sensitivity was found to drop off above 925 MHz however.
 +
 
 +
 
 +
 
 +
==== Tuning VCO ====
 +
 
 +
Tuning is needed to ensure we cover the proper range with the VCO.  If we can't align the station due to the VCO not being locked in the commercial band, being locked in the ham band does not count.
  
 
A few points about VCO tuning:
 
A few points about VCO tuning:
  
* The VCO will be 73.35 MHz higher than the intended receiver frequency
+
* The VCO will be 73.35 MHz lower than the intended receiver frequency
 
* '''the sweet spot for the VCO steering voltage is 2.5-7.5 v'''
 
* '''the sweet spot for the VCO steering voltage is 2.5-7.5 v'''
 
* each VCO is designed to cover half the range of the quantar.
 
* each VCO is designed to cover half the range of the quantar.
* with the shield on they move up about 2 MHz
+
* with the shield on they move up about 1.5 MHz
* from hot to cold Fr changes about 1.5 to 2 MHz. Hotter makes Fr go down, cool makes it go up.
+
* from hot to cold Fr changes about 1 to 2 MHz. Hotter makes Fr go down, cool makes it go up.
 
* removing capacitance makes Fr go up.
 
* removing capacitance makes Fr go up.
 
* removing inductance makes Fr go up.
 
* removing inductance makes Fr go up.
* tune a bit below the frequency (5 MHz) and adjust the tuning cap and a grinder tool (C2205)
+
* tune a bit below the frequency (5-10 MHz) and adjust the tuning cap and a grinder tool (C2205)
 
* Get it close, I'm a perfectionist, but really 1-2 MHz from idea will not matter.
 
* Get it close, I'm a perfectionist, but really 1-2 MHz from idea will not matter.
 
* at higher steering voltages (>9v) the VCO may get dirty.  This is normal.
 
* at higher steering voltages (>9v) the VCO may get dirty.  This is normal.
Line 180: Line 243:
 
Put 5.0v on the steering line and note the frequency on the spectrum analyzer.  You can sweep it from low to high and check the coverage is correct per the table above (it will be about 2MHz higher with the case off).
 
Put 5.0v on the steering line and note the frequency on the spectrum analyzer.  You can sweep it from low to high and check the coverage is correct per the table above (it will be about 2MHz higher with the case off).
  
We're going to align it for a center frequency of XXXXX MHz. The reason for this is the lower VCO covers XXX, making the center 519, but with the shield off we will see the CF be ~2 MHz lower. The shield adds capacitance to the circuit and this ups the resonate frequency.   
+
We're going to align it for a center frequency of '''835''' MHz. The reason for this is the VCO needs to cover 815-860 (888-933 MHz receiver freq), making the center 838, but with the shield off the resonate frequency lower by ~2 MHz, and we want to be able to tune this out. The shield adds capacitance to the circuit and this ups the resonate frequency.   
  
  
while monitoring the VCO output on the spectrum analyzer, use a diamond tip Dremmel grinder and remove some of the high impedance end tuning cap on the VCO. Go a very little at a time, it makes a big difference. Use this to bring the VCO frequency up to XXXXXXX on the dot. It's normal for the VCO to jump a couple hundred KHZ or so depending on the stability of your power supply. There is no PLL running to lock it.
+
While monitoring the VCO output on the spectrum analyzer, use a diamond tip Dremmel grinder and remove some of the high impedance end tuning cap on the VCO. Go a very little at a time, it makes a big difference. Use this to bring the VCO frequency up to 835 on the dot. It's normal for the VCO to jump a couple hundred KHZ or so depending on the stability of your power supply. There is no PLL running to lock it.
  
Do a sweep according the the service manual and ensure it covers the intended range for the Low VCO (again subtract 2 MHz from the intended range with the case off).  You want the coverage to be in the sweet spot of 2.5v to 7.5v, if it's not re-adjust it.
+
Do a sweep according and ensure it covers the intended range, 815-860 MHz.  You want the coverage to be in the sweet spot of 2.5v to 7.5v, if it's not re-adjust it, but keep in mind it's easier to bring the resonate frequency up by removing material than make it go lower by adding material.
  
Clean it with some solvent and let it dry.
+
Clean it with some solvent and let it dry.
  
 +
=== Removing the Preselector ===
  
Below is real data from a 800 to 900 conversionThe before data was taken with the shield off making the frequency about 2 mhz lower.  The pink represents the sweet spot of the VCO tuning voltages.  Note this covers both VCO ranges acceptably.  
+
The preselector is a large heat sink on the board, and will require advanced SMD experience to removeOnce it's removed it will be jumped out using a section of miniature coax (RG-393).
  
 +
* The first step is to mask off the rest of the board to prevent damage due to the heat required to reflow the shield and preselector
 +
 +
* re-install the shields on the cases near the preselector assembly and tape the few exposed compounits
 +
<gallery widths=300px heights=300px mode="packed-hover">
 +
800 MHz Receiver Preselector Removal - 0268.JPG|800 MHz Receiver Preselector Removal
 +
</gallery>
  
{| class="wikitable"
+
* tape aluminum foil heat shield over the exposed board
!!!colspan="2" style="text-align: center;"|Before Mod (shield off)!!colspan="2" style="text-align: center;"|After mod, Shield on
 
|-
 
|Volts||Low||High||Low||High
 
|-
 
|0||450||465||491||508
 
|-
 
|1||456.5||472||498||516
 
|-style="background:Pink"
 
|2||462.5||478||504||521
 
|-style="background:Pink"
 
|3||467.5||483||509||527
 
|-style="background:Pink"
 
|4||472.5||488||514||531
 
|-style="background:Pink"
 
|5||477.5||492||518||536
 
|-style="background:Pink"
 
|6||482||497||522||539
 
|-style="background:Pink"
 
|7||487||501||526||543
 
|-
 
|8||491||505||530||547
 
|-
 
|9.2||496.5||510||535||552
 
|-
 
|10||500||513||537||555
 
|-
 
|Delta||50||48||46||47
 
|}
 
  
=== Image filter ===
+
<gallery widths=300px heights=300px mode="packed-hover">
 +
800 MHz Receiver Preselector Removal - 0270.JPG|800 MHz Receiver Preselector Removal - 0270
 +
800 MHz Receiver Preselector Removal - 0272.JPG|800 MHz Receiver Preselector Removal - 0272
 +
</gallery>
  
The image filter is a low pass filter which prevents the Local Oscillator from radiating out the receiver antenna port.  This may be debatable as to the necessity of this modification as the cutoff of the R1 image filter is still above the maximum operating frequency of 470 MHz.
+
* turn on your pre-heater under the board and let it warm up for 5-10 minutes, as you don't want to thermal shock the preselector
 +
<gallery widths=300px heights=300px mode="packed-hover">
 +
800 MHz Receiver Preselector Removal - 0274.JPG|800 MHz Receiver Preselector Removal - 0274
 +
</gallery>
  
{| class="wikitable"
+
* using the 1000W heat gun and tip, reflow the shield. keep the heat gun moving, this can '''DESTROY''' the board if done improperly. Once it's reflowed, remove the shield and set it asside.
|+ style="text-align: left;" | Sensitivity for 12 dB SINAD of R0 to R2 receiver before and after image filter.
 
!Frequency!!Befor mod!!After Mod
 
|-
 
|483.000||-123.2 dBm||-123.5 dBm
 
|-
 
|454.000||-122.7 dBm||-123.1 dBm
 
|-
 
|454.250||-122.7 dBm||-123.1 dBm
 
|-
 
|470.000||-121.7 dBm||-122.4 dBm
 
|}
 
  
This is a spectrum plot of the R1 and R2 traces over laid on each other. R1 is in yellow and R2 is purple.
+
<gallery widths=300px heights=300px mode="packed-hover">
As can be seen the R1 filter doesn't start attenuating the signal until 490 MHz, and since this is past the first amplifier stage, loss here is not a huge problem.
+
800 MHz Receiver Preselector Removal - 0277.JPG|800 MHz Receiver Preselector Removal - 0277
 
<gallery heights=200px widths=200px mode="packed-hover">
 
UHF R1 to R2 RX FILTER no table.png|UHF R1 to R2 RX FILTER no table
 
UHF R1 to R2 RX FILTER.png|thumb|UHF R1 to R2 RX FILTER
 
 
</gallery>
 
</gallery>
  
There is no difference in sensitivity, but all units which I convert I modify this for completeness.   
+
* Now while keeping heat on the preselector evenly pull it up buy using a pick or tongs.  DO NOT grip it by the exposed ceramic material or cause undo pressure to be applied to it, or it may crack.   
  
Below are the parts differences.  All parts are NPO 0805 size ceramic parts.  I remove all the "Discard" parts first and then move the couple parts since the board is already hot. 
+
<gallery widths=300px heights=300px mode="packed-hover">
{| class=wikitable
+
800 MHz Receiver Preselector Removal - 0281.JPG|800 MHz Receiver Preselector Removal - 0281
|+ style="text-align: left;" | R1 to R2 parts differences.
+
</gallery>
!Part number!!R1!!R2!!Notes
 
|-
 
|C2050||9.1 pF||8.2 pF||Discard
 
|-
 
|C2051||16pF||13pF||move to C2053
 
|-
 
|C2053||18 pF||16 pF||Discard
 
|-
 
|C2054||8.2 pF||6.8 pF||Move to C2050
 
|-
 
|C2057||18 pF||16 pF||Discard
 
|-
 
|C2059||18 pF||13 pF||Discard
 
|-
 
|C2070||5.6 pF||3.3 pF||Move to C2076
 
|-
 
|C2076||7.54 pF||5.6 pF||Discard
 
|-
 
|C2079||9.1 pF||5.6 pF||Discard
 
|-
 
|C2083||47 pF||22 pF||Discard
 
|-
 
|||||||
 
|-
 
|R2449||0 Ohm||0 Ohm||ID Leave
 
|-
 
|R2450||0 Ohm||3300 Ohm||ID Replace
 
|}
 
  
Some parts (C2050-C2054) are under a soldered on shield cage between the preamplifer and the 1st mixer.
+
* Set the Preselector aside an allow it to cool naturally, as it may crack if thermal shocked.
<gallery heights=300px widths=300px mode="packed-hover">
+
<gallery widths=300px heights=300px mode="packed-hover">
UHF Image Filter case on.jpg| Shield on
+
800 MHz Receiver Preselector Removal - 0290.JPG|800 MHz Receiver Preselector Removal - 0290
UHF_RX_Image_filter_case_off_R1_to_R2.png | Shield off R1-R2 parts
 
 
</gallery>
 
</gallery>
  
Using a board pre-heater and hot air wand remove the shield.  It does not go through the board, but be careful you do not over heat it as the inductors will melt.  It's much easier than the VCO shield.  Once off change the parts per the table and reinstall the shieldNote the orientation of the shield, it's designed with cutouts on the input and output strip-lines.
+
* note some the preselector "carrier" parts may have reflowed and movedThis is only a concern if you want to re-install it at further date.
  
* Change the parts under the 1st mixer shield per the table and picture below
+
<gallery widths=300px heights=300px mode="packed-hover">
<gallery heights=400px widths=400px mode="packed">
+
800 MHz Receiver Preselector Removal - 0299.JPG|800 MHz Receiver Preselector Removal - 0299
UHF_RX_First_Mixer_board_R1-R2_mod.png | UHF 1st mixer parts moved
+
800 MHz Receiver Preselector Removal - 0297.JPG|800 MHz Receiver Preselector Removal - 0297
 
</gallery>
 
</gallery>
  
=== Change the Module ID ===
+
* Allow the board to cool.  It's optional if you want to clean up the solder on the underside of the board, I normally leave it.
  
There is one resistor, R2450 by the PLL synth which must be changed to a 3300 Ohm chip to set the RX type to R2.
+
* once cooled, install a coax jumper across the pads of the preselector
<gallery heights=400px widths=400px mode="packed">
+
 
UHF_RX_R2450_location.jpg| R2450 location
+
<gallery widths=300px heights=300px mode="packed-hover">
 +
800 MHz Receiver preselector jumper TRF6551J - 0306.JPG|800 MHz Receiver preselector jumper TRF6551J - 0306
 
</gallery>
 
</gallery>
  
=== Preselector ===
+
==== Preselector modification ====
  
The preselector modifications are covered [[UHF_Preselector|elsewhere]].
+
There have been a couple attempts at modifying the preselector by the same method used on the VCO resonator.  A couple have been broken due to the thermal shock.
  
=== Final testing ===
+
the 800 MHz preselector is 36 MHz wide centered on 814.5 MHz. 
 +
 
 +
The height including the "chassis" is 0.585"
 +
 
 +
removing .050" making it .535" total, moved the CF to 890 MHz which might have worked.  More research needs to be done.
 +
 
 +
=== Results ===
  
Once the preceding modifications have been done it's time to test the receiver module with out the preselctor.  
+
The following results were obtained from a unit converted which did not have the extra filter between the VCO amp and the Mixer.  A conversion of one of these extra FL3200 units has not been attempted at this point.  
  
With the power off install the RX board back in the metal case and slot it into the Quantar chassis.
+
This was using a 12dB SINAD for NBFM 1.5 KHz 1kc tone deviation.  The voltages were measured from the test screen in the quantar software.  I've noted this as the quantar reads it's steering voltage about 0.5v low.
  
Power the station on and program a test code plug with the following frequencies.
+
There is a pronounced roll-off above 915 MHz, which may be due to the image filter (FL2030).  As this covers both 25 and 12 MHz split repeaters in the ham band, further experimentation was not attempted.  If you find out what is limiting this and how to overcome it for full ham band performance please update this wiki.
  
{| class="wikitable"
+
{| class="wikitable"  
!
+
|+ style="text-align: left;" | Sensitivity for 12 dB SINAD of 800 to 900 receiver for NBFM
!Lock?
+
!Frequency!!VCO voltage!!After Mod
!Steering voltage
+
|-
!Sensitivity
+
|885||Unlock||NA
 
|-
 
|-
!425.000
+
|890||1.06v||-121.0 dBm
|
 
|
 
|
 
 
|-
 
|-
!430.000
+
|895||1.57v||-121.0 dBm
|
 
|
 
|
 
 
|-
 
|-
!433.000
+
|900||2.04v||-121.0 dBm
|
 
|
 
|
 
 
|-
 
|-
!435.000
+
|902||2.27v||-121.2 dBm
|
 
|
 
|
 
 
|-
 
|-
!438.000
+
|905||2.63v||-121.5 dBm
|
 
|
 
|
 
 
|-
 
|-
!454.000
+
|910||3.18v||-121.5 dBm
|
 
|
 
|
 
 
|-
 
|-
!454.250
+
|915||3.84v||-121.5 dBm
|
 
|
 
|
 
 
|-
 
|-
!470.000
+
|925||5.26v||-119.7 dBm
|
 
|
 
|
 
 
|-
 
|-
!473.000
+
|927||5.53v||-119.2 dBm
|
 
|
 
|
 
 
|-
 
|-
!475.000
+
|930||6.00v||-117.7 dBm
|
 
|
 
|
 
 
|-
 
|-
!478.000
+
|935||6.79v||-116.2 dBm
|
 
|
 
|
 
 
|-
 
|-
!480.000
+
|940||7.57v||-114.7 dBm
|
 
|
 
|
 
 
|-
 
|-
!485.000
+
|948||8.55v||-112.6 dBm
|
 
|
 
|
 
 
|-
 
|-
!490.000
+
|950||Unlock||NA
|
 
|
 
|
 
 
|}
 
|}
  
Use the control and metering screen to check the oscillator voltage at each frequency.  Ideally you should be between 2 and 7.5 volts for 438.00-454.00 and 454.25-470.  The other frequencies are there just to find the range of the VCO's as installed.  If your minimum lock voltage is under 1.5v you may want to re-adjust the VCO in question.  If the high frequency is above 8v you may wish to adjust as well.
+
=== Final testing ===
  
Check the sensitivity of the receiver at all locked frequencies. With out the preselector installed the sensitivity should be better than -121.5 dBm or .2μV for 12dB SINAD.
+
Once the preceding modifications have been done it's time to test the receiver module.  
 +
 
 +
With the power off install the RX board back in the metal case and slot it into the Quantar chassis.
  
Once this is verified, button it up, install the modified preselector and mark the unit as a R2 modified.
+
Power the station on and program a test code plug.
  
Congratulations!
+
Use the control and metering screen to check the oscillator voltage at each frequency. Ideally you should be between 2 and 7.5 volts for 896.00-940.00.  If your minimum lock voltage is under 1.5v you may want to re-adjust the VCO in question.  If the high frequency is above 8v you may wish to adjust as well.
   
 
  
== R3 to R2 ==
+
Check the sensitivity of the receiver at all locked frequencies.  The sensitivity should be better than -120.5 dBm or .225μV for 12dB SINAD in NBFM.  Motorola quotes the WBFM sensitivity for 900 MHz as .35μV which is -116 dBm or better, but this includes the preselector (1.5 dB loss).
  
No data on this yet. Should be doable, VCO will be a problem.
+
Once this is verified, button it up, install the modified preselector and mark the unit as a 800 modified to 900.
  
== R4 to R2 ==
+
Congratulations and please report back with your findings!
  
No data on this yet, may be doable VCO will be a problem.
+
[[Category:Quantar]]

Latest revision as of 22:49, 30 November 2016

This page covers the 800 MHz receiver covering 806-826 MHz. The 900 Receiver is almost identical to this as well and covers 896-902 MHz

Basics

The receiver is pictured below. The 800/900 receiver differs from all others as it's half width and includes an internal preselector. The preselector is soldered on to the board and not tune-able as in the VHF/UHF modules.

The receiver is a low side injection with the 1st Lo operating +73.35 MHz lower than the receiver frequency. There is no CPU on board like there is in the exciter as the SCM talks to the PLL chip (U2401) directly.

There is a 7 pole preselector in the front end prior to the first RF amplifier. This has about a 1.5 db insertion loss with a very good roll off. I'm actually surprised it's as good as it is, 1.5 dB is low for a ceramic filter.

There is a image filter (FL2030) which filters the input to the double balanced mixer to prevent the RF radiating out the antenna port.

In the second 800 MHz board I looked at, (TRF6551G39) there is another filter (FL2300) which is on the input to the mixer from the Local Oscillator output. This is not present in the first board I converted (TRF6551J), and just has a 50 ohm line in place of it. It remains to be seen what this filter does. It needs to be removed and swept for response.

Programing ROM

The SCM knows what type of module is inserted by reading voltage divider resistors on the u2600 A/D Converter. Some of these are on ports used for other things, Change Frequency and Lock and then A8 input is used for ID. The R2816 and R2815 form a voltage divider feeding the A8 input with the computed voltage in volts.

Range R2422 Chg Freq R2414 Lock SPACE R2816 R2815 A8 Volts
900 0 0 1200 4700 4.0
800 0 0 1200 2700 3.5

VCO

The VCO in the 800 differs from the 900 unit in that the 800 VCO does not have a soldered on shield over it. I'm not sure what difference this makes in practical use. Note on the TRF6551J unit there is a solder mask which would fit a shield.

As the VCO runs at 73.35 MHz lower than the intended receive frequency the VCO will expect to lock over 726.65-752.65. Typically this will have extra range at either side of the VCO. C2205 is a laser cut tuning cap which is tuned at the factory to bring the unit up to a proper resonate frequency. Removing material (removing capacitance) raises the frequency of the VCO.


Service Manual

The service manual excerpt is in PDF and below.

800 MHz Receiver Model TRF6551 Service Manual Excerpt

In PNG format.

Pictures

Conversions

Converting to the 800 MHz to 900 MHz receiver works best if you plan to use an external preselector/pre-amp as the losses on 900 MHz are high. 100' of 7/8" heliax has just under 1.2 dB of loss, and 6' of RG-400 is 1 dB. It's completely possible to have 6' of interconnect cable between the duplexers and filters, and installing a pre-amp as close to the preselctor as possible will negate most of this loss.

800 to 900 MHz

The basic premise of converting the 800 to a 900 is four major parts:

  • Convert the VCO for the proper frequency coverage
  • Remove/Modify the ceramic preselector
  • Change the module ID
  • Test the unit.

It's important to have a known good working unit on 800 before starting work. If it is not working at it's intended frequency, fix it first.

It would be good to become very familiar with the schematic and service manual before attempting this.

Tools and test equipment you should have (not an exhaustive list)

  • Hot air station
  • Under board pre-heater
  • 1000W heat gun and tips (for preselector removal)
  • High quality soldering iron (metcal)
  • Dual voltage variable DC supply
  • Spectrum analyzer and probes
  • Service monitor
  • Experience working with SMD and reflow techniques

Converting VCO

There is a single VCO for both the 800 and 900 bands. This is in contrast to the UHF and VHF modules which use two VCO to cover the entire band. The VCO (Q2202) consists of a tuned coax resonator shorted on one end (L2202) in resonance with parallel capacitors (C2211 & C2209) and laser cut tuning capacitor (C2205).

As this is a low side injection receiver the VCO runs 73.35 MHz below the intended signal. This is different from the high side injection systems used on VHF and UHF. Low side injection is less of an issue on 800/900 as traditionally there are no users on the image frequencies (+73 MHz) of any real ERP to cause interference.


  • Mask off the VCO using Krapton tape and/or aluminum foil. The electrolytic caps don't do well with excess heat, especially as most are 10-15 years old at this point.
  • Setup the board with an under board pre-heater set to 460f. This is important as you want the board to warm up to close to the melting point of solder. Then the hot air wand will heat it the rest of the way.
  • Once you have this preheated for a few minutes, use the hot air wand with no tip at 15 L/m and 650f on the coax resonator (L2202). Move it in a circular pattern while waiting on it to reflow. Once it's flowed, pull the resonator up and allow it to cool on an insulated pad. You don't want to thermal shock it, as it will crack being it's ceramic.

Careful you don't disturb any of the VCO parts, as they will most likely be reflowed. If you do, don't panic, just look at the before picture and put everything back to how it was on the VCO. It's a pretty easy circuit to figure out.

  • While the VCO is still Warm, remove L2201 and set asside. This will allow us to power the VCO with out powering the rest of the receiver.
  • Change the following parts in the VCO area.
800 to 900 parts differences.
Part number 800 900 Notes
C2207 3.3 pF 2.5 pF INFO ONLY don't change, steering range
C2211 12 pF 6.8 pF Tank
C2209 8.2 pF 5.6 pF Tank
L2202 430 mils 350 mils Resonator
R2815 2100 4700 ID resistor
  • Now the VCO should be exposed and you can tack on some test wires to it while it cools down.
  • Hook up to the VCO as shown.
  • 10.0 V power to VCO on P2200 Pin #3
  • 0-12 V on the steering line, P2201 #1
  • output probe near the C2219 VCO output stripline. This doesn't need to be tacked on, just a probe near it so you can pickup the output on the spectrum analyzer.

Cutting down the Coaxial resonator

  • The next step is to cut the coaxial resonator (L2202) down from the back side (the shorted, non leaded side) till it's 0.350" long.

You will need a diamond grinder to do this.

Once it's 350 mils long (5 mils longer is ok, shorter will require more trimming of C2205), fashion a copper tape strip and apply over the end of the resonator. This will need to be soldered in place around the edge and then soldered to the inside center conductor. This replaces the shorted out end which was ground off.

  • Reinstall the coaxial resonator

To arrive at the correct lengths the following measurements and parts changes were tried.

!Parts changed Resonator length (mils) 0.5V 5V 9V Delta
Resonator only 390 739 766 785 46
C2211 & C2209 390 755 786 807 52
C2211 & C2209 350 785 820 841 56
C2211/C2209 & trim C2205 350 809 849 874 65

It was found a combination of changing C2211, 2209 and grinding the resonator to 0.350" got the VCO to within 10 MHz of where it needed to be to cover the entire ham band and the commercial band just below it. This does make the coverage of the receiver larger than before, which is aventagious for amateur use, as it now covers the entire ham band. Sensitivity was found to drop off above 925 MHz however.


Tuning VCO

Tuning is needed to ensure we cover the proper range with the VCO. If we can't align the station due to the VCO not being locked in the commercial band, being locked in the ham band does not count.

A few points about VCO tuning:

  • The VCO will be 73.35 MHz lower than the intended receiver frequency
  • the sweet spot for the VCO steering voltage is 2.5-7.5 v
  • each VCO is designed to cover half the range of the quantar.
  • with the shield on they move up about 1.5 MHz
  • from hot to cold Fr changes about 1 to 2 MHz. Hotter makes Fr go down, cool makes it go up.
  • removing capacitance makes Fr go up.
  • removing inductance makes Fr go up.
  • tune a bit below the frequency (5-10 MHz) and adjust the tuning cap and a grinder tool (C2205)
  • Get it close, I'm a perfectionist, but really 1-2 MHz from idea will not matter.
  • at higher steering voltages (>9v) the VCO may get dirty. This is normal.

Put 5.0v on the steering line and note the frequency on the spectrum analyzer. You can sweep it from low to high and check the coverage is correct per the table above (it will be about 2MHz higher with the case off).

We're going to align it for a center frequency of 835 MHz. The reason for this is the VCO needs to cover 815-860 (888-933 MHz receiver freq), making the center 838, but with the shield off the resonate frequency lower by ~2 MHz, and we want to be able to tune this out. The shield adds capacitance to the circuit and this ups the resonate frequency.


While monitoring the VCO output on the spectrum analyzer, use a diamond tip Dremmel grinder and remove some of the high impedance end tuning cap on the VCO. Go a very little at a time, it makes a big difference. Use this to bring the VCO frequency up to 835 on the dot. It's normal for the VCO to jump a couple hundred KHZ or so depending on the stability of your power supply. There is no PLL running to lock it.

Do a sweep according and ensure it covers the intended range, 815-860 MHz. You want the coverage to be in the sweet spot of 2.5v to 7.5v, if it's not re-adjust it, but keep in mind it's easier to bring the resonate frequency up by removing material than make it go lower by adding material.

Clean it with some solvent and let it dry.

Removing the Preselector

The preselector is a large heat sink on the board, and will require advanced SMD experience to remove. Once it's removed it will be jumped out using a section of miniature coax (RG-393).

  • The first step is to mask off the rest of the board to prevent damage due to the heat required to reflow the shield and preselector
  • re-install the shields on the cases near the preselector assembly and tape the few exposed compounits
  • tape aluminum foil heat shield over the exposed board
  • turn on your pre-heater under the board and let it warm up for 5-10 minutes, as you don't want to thermal shock the preselector
  • using the 1000W heat gun and tip, reflow the shield. keep the heat gun moving, this can DESTROY the board if done improperly. Once it's reflowed, remove the shield and set it asside.
  • Now while keeping heat on the preselector evenly pull it up buy using a pick or tongs. DO NOT grip it by the exposed ceramic material or cause undo pressure to be applied to it, or it may crack.
  • Set the Preselector aside an allow it to cool naturally, as it may crack if thermal shocked.
  • note some the preselector "carrier" parts may have reflowed and moved. This is only a concern if you want to re-install it at further date.
  • Allow the board to cool. It's optional if you want to clean up the solder on the underside of the board, I normally leave it.
  • once cooled, install a coax jumper across the pads of the preselector

Preselector modification

There have been a couple attempts at modifying the preselector by the same method used on the VCO resonator. A couple have been broken due to the thermal shock.

the 800 MHz preselector is 36 MHz wide centered on 814.5 MHz.

The height including the "chassis" is 0.585"

removing .050" making it .535" total, moved the CF to 890 MHz which might have worked. More research needs to be done.

Results

The following results were obtained from a unit converted which did not have the extra filter between the VCO amp and the Mixer. A conversion of one of these extra FL3200 units has not been attempted at this point.

This was using a 12dB SINAD for NBFM 1.5 KHz 1kc tone deviation. The voltages were measured from the test screen in the quantar software. I've noted this as the quantar reads it's steering voltage about 0.5v low.

There is a pronounced roll-off above 915 MHz, which may be due to the image filter (FL2030). As this covers both 25 and 12 MHz split repeaters in the ham band, further experimentation was not attempted. If you find out what is limiting this and how to overcome it for full ham band performance please update this wiki.

Sensitivity for 12 dB SINAD of 800 to 900 receiver for NBFM
Frequency VCO voltage After Mod
885 Unlock NA
890 1.06v -121.0 dBm
895 1.57v -121.0 dBm
900 2.04v -121.0 dBm
902 2.27v -121.2 dBm
905 2.63v -121.5 dBm
910 3.18v -121.5 dBm
915 3.84v -121.5 dBm
925 5.26v -119.7 dBm
927 5.53v -119.2 dBm
930 6.00v -117.7 dBm
935 6.79v -116.2 dBm
940 7.57v -114.7 dBm
948 8.55v -112.6 dBm
950 Unlock NA

Final testing

Once the preceding modifications have been done it's time to test the receiver module.

With the power off install the RX board back in the metal case and slot it into the Quantar chassis.

Power the station on and program a test code plug.

Use the control and metering screen to check the oscillator voltage at each frequency. Ideally you should be between 2 and 7.5 volts for 896.00-940.00. If your minimum lock voltage is under 1.5v you may want to re-adjust the VCO in question. If the high frequency is above 8v you may wish to adjust as well.

Check the sensitivity of the receiver at all locked frequencies. The sensitivity should be better than -120.5 dBm or .225μV for 12dB SINAD in NBFM. Motorola quotes the WBFM sensitivity for 900 MHz as .35μV which is -116 dBm or better, but this includes the preselector (1.5 dB loss).

Once this is verified, button it up, install the modified preselector and mark the unit as a 800 modified to 900.

Congratulations and please report back with your findings!