Comtech Tuner Module Upgrade
#1
Posted 01 January 2011 - 06:52 PM
So, I think this is a good time to talk about the Comtech Tuner Module Upgrade. And I will use the fellow's bad experience to help explain the upgrade process.
I will begin with some of my personal observations. Please keep in mind that my comments apply to 900-1300 MHz video Rx's and NOT the more modern 2.4 - 5.8 GHz models.
1. Comtech is based in Tawain. They introduced their RF modules a couple decades ago and the intended application at the time was consumer satellite TV receivers. They were the only game in town at first, but eventually competition appeared as other suppliers copied their tuner module designs (cosmetically, the copies look very similar). Comtech offers several different modules; the one needed for a 900-1300 MHz A/V Rx is the BSS479LAIDF analog tuner module.
2. Rx's that have a tuning knob do not use a tuner module. So installing a Comtech module in them is not practical.
3. The tuner modules used in the 900MHz through 1.3GHz wireless video receivers are all interchangeable (as long as you observe their pin-outs, which can vary). The only real difference is in the software used on the external baseband circuit board; Its microcontroller sends data to the tuner module via two 12C buss pins that configures the module's operating frequency. The reason I mention this is because I have read more than once on other forums that the 1.2 - 1.3 GHz Rx's are better because their tuner modules have higher performance. Let me say that such a claim is complete nonsense because the tuner modules in them are the same ones used in 900MHz Rx's.
4. Unfortunately, the modules produced by some of the other suppliers have lower performance. One bad sign is an economy priced video Rx that has F-style TV antenna connectors instead of the more expensive SMA type. Another common sign can be found in modules that were hand soldered instead of a reflow soldering machine. Issues like these (and more) have been reported on the low cost Mainland China made 900MHz - 1.3GHz video receivers. The Taiwan brands (for example, Racewood and Lawmate) seem to have avoided these quality problems; It is my hope that these two favorite brands never try reducing their manufacturing costs by using the cheaply made tuner modules. I'm not saying that the tuner module copies are all junk, but sadly I have found that many of the Hong Kong / China brands are poorly designed and/or poorly made, as are many ShanZhai products. Mainland China is a country that is in its industrial infancy, so they still have a lot to learn about designing/manufacturing quality products.
5. Speaking of Racewood and Lawmate video Rx's, I personally do not understand why many folks insist that that they have Comtech branded modules in them. The OEM modules these two companies have been using are good quality and have the same performance as the beloved Comtech. However, all those cheap import Rx's (Hong Kong & eBay specials) are likely candidates for a Comtech module upgrade.
6. Comtech uses positive RSSI scaling. Many clone modules, including the good performers, often have negative going RSSI scaling. So keep this difference in mind if you will be building a custom DiY diversity controller that uses RSSI.
7. If you replace the tuner module in your video Rx you must recalibrate the video levels, even if you install a tuner module with the exact same part number as the original. No exceptions! My opinion is that if you are not able to calibrate the video level, which requires an o-scope, then don't bother to change the tuner module. I'll talk about this again later on.
8. The exact wiring to the tuner module will vary. The non-Comtech module will probably have different pin order than the Comtech module. Plus, the baseband PCB that is connected to the tuner module will have wiring that can be different than other video receivers. So be prepared to destroy your Rx if you are not sure what to do.
That sums up some of big ticket issues. Now we can talk about the fellow that had a bad experience with his two Comtech modules.
#2
Posted 01 January 2011 - 07:11 PM
One of the issues with the Comtech swap is that its wiring is sometimes different than what is found in the similar-looking module brands. Fortunately he knew this and had a wiring photo that he found online. The photo clearly showed him how to connect the wiring from his identical looking baseband PCB. But after carefully reviewing the photo link he sent me I noticed that two wires were swapped. The 8V tuner bias voltage pin was swapped with a nearby I2C pin, which unfortunately was a deadly mistake.
Here is the photo he used, as posted at another popular FPV forum. But I have added red-line correction marks to show the error:
Just to be clear, the wiring is wrong in that photo. If you wire the Comtech Module as shown you will most likely kill it. The red X's show the mistake.
#3
Posted 01 January 2011 - 07:17 PM
First I will show the wiring when the OEM tuner module is used. Keep in mind that the wiring shown here is specific to this particular tuner module.
But if the Comtech module was installed it will look like this:
I also marked the RSSI pin location in case you need to use it.
#4
Posted 01 January 2011 - 07:36 PM
Adjusting video amplitude/level is a chicken or the egg situation. In order to adjust the video Rx you need to have a video Tx that is properly adjusted. Or on the flip side, to adjust the video Tx it is very convenient to have a Rx that is correctly calibrated. In the industry the common method is to first calibrate the Tx to the required video bandwidth (which directly determines the Tx's video level) using the Bessel null method, but this requires a spectrum analyzer.
However, if you have a video Tx that has given you good performance, especially if it has been bench calibrated by your FPV vendor (ask your FPV store if they did this in-house for you), then you can use it to adjust the Rx's video level. Some basic instructions on how to do that are posted here:
http://www.rc-cam.co...e-video-levels/
A less precise method is to measure the video level of the Rx before the tuner replacement. Then measure it again after the swap and set it to be the same as before. At least this way things will be no better or worse than what you started with. But honestly, if you are going to all the trouble of upgrading the Rx to make it perform better, then use a calibrated Tx to help you set the Rx to the correct levels.
Final tips: As mentioned, on some wireless video systems the video level will radically change if the frequency/channel is changed. So check for this nuisance and if necessary re-adjust the video whenever you select a different RF channel. By the way, if your baseband PCB does not have a pot and the video levels are off by more than 15%, then I would suggest that it is time to invest in a better Rx.
Edited by Mr.RC-Cam, 24 February 2012 - 11:12 AM.
#5
Posted 29 July 2011 - 01:56 PM
This is the first test of my homebrew RSSI diversity and PLL controler :
I can also control any transmitter on custom frequency:)
My five Comtech have exactly the same calibration, you can see only somes small clip when the multiplexer switch.
(four tuners are tested on this video)... I use only One baseband PCB!! (3 wire only is needed, GND, baseband and 8V reg), 5V must be external, the onboard 7805 can't power 4 tuners.
It's easy to recover pinout inside any receiver (to swap a crappy tuner with a good Comtech) -> GND, 5V and 8V.
I2C CLK, I2C SDA can be reversed without damage.
On Comtech, there are only two contacts remaining, RSSI, and LNB power (to power the new "EME" Ultra Low Noise preamp from MiniKit:)))
Pascal
#6
Posted 30 July 2011 - 10:02 AM
I fully understand your design goals because your configuration is similar to a project I worked on. But mine only had two Comtech tuners. The tuners were BSM479 which have software selectable 17/27MHz IF filters. Like your project, they shared one baseband processing board (a TR-1500 board harvested from a Racewood A/V receiver). A Maxim OSD IC provided frequency readout, RSSI indication, and spectrum graphing (poor man's spectrum analyzer). Push button switches were used to select the frequency (continuous tuning) or measurement data. But yours is fantastic because it has four receivers!
#7
Posted 30 July 2011 - 10:08 AM
Terry
#8
Posted 05 September 2011 - 12:25 AM
Here is the "1.0 final" stuff :
Features (sorry for me english lol) :
Rotary encoder and fast/easy to use OSD GUI.
Decade Engineering BOB-4 OSD module @ maximum processing speed / maximum display optimisations / 50Hz refresh rate (my SPI bus run @ 1MHz)
4 (up to 16) "17mhz saw modded" tuners diversity
Visual RSSI bargraph for each tuners / diversity selected tuner / usefull diy antenna or low noise preamp comparison tool
Diversity algorithm that takes into account the background noise level /w manual or auto trim.
Live "poor man's spectrum analyser" mode, PLL charge pump configured @ 170uA for spectrum and 50uA for normal tuning (In the future a dedicated tuner with a very narrow bandpass 480MHz IF filter give a better spectrum display / selectivity) Next I test a way faster refresh on an OLED display module.
PPM output for the training cable, this receiver can synchronise any VTX PLL (1 cheap arduino mini pro for 1 modded TX)
Manual frequency hopping + non standard RF usage/non standard CCTV channels = bulletproof/virtually undisturbable video downlink
Individual AFT/AFC Automatic Frequency Control / Fine Tunning on each tuners ->
A closed loop "servo" frequency control with individual bargraph + or - 2MHz -> compensate for tuners components tolerances, usefull @ long range/ for weak signal
And there is still 90% free space in the groundstation ATmega2560 for telemetry / antenna tracking and more DIY stuff !
Next -> double conversion superheterodyne tuner ?
Pascal
Edited by Serveurperso, 05 September 2011 - 04:41 AM.
#9
Posted 06 September 2011 - 09:53 AM
#10
Posted 08 September 2011 - 07:45 PM
#11
Posted 11 September 2011 - 08:08 PM
Sorry, but I have no idea what its pin-out is. There are over a dozen different Chinese tuner modules that show up in these cheap video Rx products so it is mystery to me what is in yours.did know the color code or the pinout on the hobbyking base plate to be able to succesfully connect the module?
Perhaps post some clear close-up photos of the interconnect wires on the tuner module's pins and some photos of the same wire connections on the PCB side. Also post the EXACT voltages measured at each wired pin on the tuner. No promises, but this information may help identify the signals. Also, please study the tuner module's tin can and report the manufacture name and part number that is printed or stamped on it.
#12
Posted 11 September 2011 - 11:15 PM
#13
Posted 11 September 2011 - 11:55 PM
I see that the baseboard has only one trim... hope it is the video level trim (and not the audio)
In most baseband PCB designs, if it is a pot then it is the video level adjustment. If it is a variable inductor then it is the audio subcarrier's filter adjustment. When in doubt just mark the existing position (so you can restore the position) then see what happens when you tweak it.
#14
Posted 13 September 2011 - 03:09 AM
hi, MR RCAM, planning on replacing the module on my 900mhz hobbyking receiver with a comtech module. However wanted to ask if you did know the color code or the pinout on the hobbyking base plate to be able to succesfully connect the module? tks
Hi,
From left to right, antenna on the left side, "ComTech" brand (and white BSS479LAIDFM sticker) on top.
__------------
Antenna-__ |
| |
|___________|
1 234 56789
1 -> LNB or LNA power supply (only needed if you want to feed a preamp powered by coax cable)
2 -> Analog 0-5V RSSI level
3 -> 5V regulated logic power
4 -> Baseband out (mixed video + sound subcarrier)
5 -> I2C DATA
6 -> 9V PLL charge pump power (>9V for >1500MHz!)
7 -> I2C CLOCK
8 -> not needed (PLL programmable out)
9 -> not needed (PLL programmable out)
Ground is the metal case.
Remove the old tuner
Use a multimeter to find 5V and 9V
Use a headphone to find baseband (white noise)
Use a headphone to find I2C data and clock (typical data burst noise only when you change the channel switch)
Reverse I2C data and clock if the RX frequency dont change.
10 minutes operation and no risk for your baseband board IF you're used to "DIY".
Pascal
#15
Posted 21 September 2011 - 09:24 PM
#16
Posted 21 September 2011 - 09:33 PM
Success, HOBBYKING & BEVRC video receivers now have comtech tunners!!!! I used the OEM tunner + downloading the data sheets of the ICs I found on that board, traced it pin into the corresponding "color". Using the Comtech data sheet, soldered each cable back and voula! plug & play... the image came nice and clear! adjusted the pot on the base board, with out showing any effect in the screen so I can only imagine these chinese boards have some sort of auto function on the signal level. Posting the pictures of my setup so future followers welcome of my experience. Saludos a todos!
#17
Posted 21 September 2011 - 10:47 PM
Terry
#18
Posted 22 September 2011 - 07:31 AM
... adjusted the pot on the base board, with out showing any effect in the screen so I can only imagine these chinese boards have some sort of auto function on the signal level.
It is doubtful there is an auto circuit in the baseband board. Instead, from my experience this observation is because the monitor/display has aggressive AGC circuits to automatically control video levels. This will give you the false sense that everything is fine. So, try it again using an older monitor that has "mechanical" contrast and brightness knobs. But keep in mind that a o-scope is needed to correctly calibrate the video amplitude.
#19
Posted 09 December 2011 - 09:45 AM
absolutely amazing info. was just looking for some pinouts on Google and landed here.
TYVM,



