winterhill case?
Re: winterhill case?
they are handy. But remember you will have an insertion loss of approx. 3.5db to about 7db
Re: winterhill case?
The loss shouldn't be a problem for satellite use. I use the 4 output version proSPL409P with 11dB loss. I get MER 10.9 on the beacon from a 120cm dish through 40m of coax. It's the same on the FTS4335 NIM without the internal preamp.
The LNB is an Octagon OTLSO with a 26MHz TCXO, which brings QO-100 out around 1100MHz. I get reduced performance with an unmodified LNB and I've never been able to pinpoint why. Coincidentally, the Winter Hill TV transmitter has channel 55 on 743MHz, but it is 45km away. It's probably more to do with the 3G, 4G and Tetra about 300m away.
Brian
http://www.farnell.com/datasheets/2623701.pdf
The LNB is an Octagon OTLSO with a 26MHz TCXO, which brings QO-100 out around 1100MHz. I get reduced performance with an unmodified LNB and I've never been able to pinpoint why. Coincidentally, the Winter Hill TV transmitter has channel 55 on 743MHz, but it is 45km away. It's probably more to do with the 3G, 4G and Tetra about 300m away.
Brian
http://www.farnell.com/datasheets/2623701.pdf
Re: winterhill case?
Here is my case
I recommend the Labgear F types (Screwfix) but you need a compression tool - I originally made my own from a slotted pipe, but the tool is best.
and ventilation holes
Mike
I recommend the Labgear F types (Screwfix) but you need a compression tool - I originally made my own from a slotted pipe, but the tool is best.
and ventilation holes
Mike
Re: winterhill case?
In reply to Brian, G4EWJ, on performance of modified vs unmodified Octagon LNBs.
Early on, I performed a test with an LNB using a 10GHz input signal to see what the IF output looked like. The LNB used was a PLL Octagon with a 27MHz crystal, generating 9750MHz as the LO. To me, it seemed that there was a significant gain loss (possibly 20-30dB) when the output signal appeared to be much below what looked like the lower cut-off point of the IF filter/amp. Why this HPF effect exists, I am not sure, but one reason could be the DC decoupling choke in the LNB and its decoupling capacitor forming a HPF for the IF.
Hence in this case, for an unmodified LNB (LO 9750MHz), input signals in the 10500MHz (or 10368MHz) range will suffer an LNB gain loss, which will have a direct influence on the effective noise temperature of the system and G/T which in turn will affect the C/No at the demodulator. The LNB or LNA gain is usually kept high so that down-stream noise makes little contribution to the overall system noise, but if this reduces then the tuner noise figure (10-12dB in this case) and cable losses between LNB and tuner become significant. Rough calculations show a 2-3 dB loss of C/No can result for an unmodified vs modified LNB.
Just to say that I use a similar mod. to you but change the crystal to 26MHz, locking if necessary, giving an LO of 9388MHz. In this way the output is in the normal IF range (950 to 2150MHz) for QO-100 input signals. I have modified 4 LNBs in this way and my 1.2m PF antenna gives MER 10.7dB on the beacon (it hasn't had its pointing optimised in a long time).
A way to compensate for LNB gain reduction, would be to add a low noise amp directly after the LNB to makeup for any gain loss.
Attached are a couple of spectrum analyser photos of my non-to-scientific test results at the time.
73 all
Jen G4HIZ
Early on, I performed a test with an LNB using a 10GHz input signal to see what the IF output looked like. The LNB used was a PLL Octagon with a 27MHz crystal, generating 9750MHz as the LO. To me, it seemed that there was a significant gain loss (possibly 20-30dB) when the output signal appeared to be much below what looked like the lower cut-off point of the IF filter/amp. Why this HPF effect exists, I am not sure, but one reason could be the DC decoupling choke in the LNB and its decoupling capacitor forming a HPF for the IF.
Hence in this case, for an unmodified LNB (LO 9750MHz), input signals in the 10500MHz (or 10368MHz) range will suffer an LNB gain loss, which will have a direct influence on the effective noise temperature of the system and G/T which in turn will affect the C/No at the demodulator. The LNB or LNA gain is usually kept high so that down-stream noise makes little contribution to the overall system noise, but if this reduces then the tuner noise figure (10-12dB in this case) and cable losses between LNB and tuner become significant. Rough calculations show a 2-3 dB loss of C/No can result for an unmodified vs modified LNB.
Just to say that I use a similar mod. to you but change the crystal to 26MHz, locking if necessary, giving an LO of 9388MHz. In this way the output is in the normal IF range (950 to 2150MHz) for QO-100 input signals. I have modified 4 LNBs in this way and my 1.2m PF antenna gives MER 10.7dB on the beacon (it hasn't had its pointing optimised in a long time).
A way to compensate for LNB gain reduction, would be to add a low noise amp directly after the LNB to makeup for any gain loss.
Attached are a couple of spectrum analyser photos of my non-to-scientific test results at the time.
73 all
Jen G4HIZ
- Attachments
-
- Low IF OP G4HIZ 20181025.jpg (1.54 MiB) Viewed 4594 times
-
- Normal IF OP G4HIZ 20181025.jpg (1.6 MiB) Viewed 4594 times
Last edited by g4hiz on Wed Apr 07, 2021 8:39 pm, edited 1 time in total.
Re: winterhill case?
Interesting. Many years ago I changed the Octagon crystal to 27.5MHz, to bring 10368MHz out on the 70cm band. I recall that the LNB output was reasonably flat until about 500MHz. It didn't have an LNB connected though, so maybe it is a filter effect.
I'll go out to the dish with an unmodified LNB tomorrow and see what I get. There is something else going on, as I can get the BBC transponder in the shack on the modified LNB, but only on an FTS4335, without the preamp. I can get it at the dish with an FTS4334L. I don't have any filtering, which might be an idea, possibly both at the LNB and the NIM, in case any of the strong local signals are mixing in the LNB.
Brian
I'll go out to the dish with an unmodified LNB tomorrow and see what I get. There is something else going on, as I can get the BBC transponder in the shack on the modified LNB, but only on an FTS4335, without the preamp. I can get it at the dish with an FTS4334L. I don't have any filtering, which might be an idea, possibly both at the LNB and the NIM, in case any of the strong local signals are mixing in the LNB.
Brian
Re: winterhill case?
I have a PLL LNB here that I used with a suitable(ly changed) GPS locked oscillator to generate a 433 Mhz IF.
Basically, it does not work very well at all, presuming to be because of the fall off in IF gain, as Jen says.
I did not pursue this option. I try to keep the IF where its supposed to be. For NB modes I down convert AGAIN to 144 MHz.
On the other hand, you can now buy so called "wideband" LNBs for Sky use. Perhaps these are worth trying out?
Mike
Basically, it does not work very well at all, presuming to be because of the fall off in IF gain, as Jen says.
I did not pursue this option. I try to keep the IF where its supposed to be. For NB modes I down convert AGAIN to 144 MHz.
On the other hand, you can now buy so called "wideband" LNBs for Sky use. Perhaps these are worth trying out?
Mike
Re: winterhill case?
I haven't heard anything about wideband LNBs for a while. I recall they had an LO of 10.41MHz, so would need a crystal change.
Nice job on the case. I'm not sure which colour to go for now.
Brian
Nice job on the case. I'm not sure which colour to go for now.
Brian
Re: winterhill case?
Hi Brian,
There are many colours, Black being one I never knew about - you could then receive Es'hail2 in total darkness!
Three case tips I have are
a) I used a pair of dividers to transfer the PCB component offsets from the PCB in-the-case to the front panel.
b) I had some Red nail vanish here (its for my motorbike tail light) so you can eliminate the glare from the exposed aluminium cuts-out sides using that.
c) Measure at least twice before you cut and use tape to protect the nice coloured anodized finish.
And yes, they do appear to be 10.41 GHz.
Here is the data from a SystemSat one:
Input Frequency Range: 10.7 ~ 12.75 GHz
Output Frequency Range: Vertical - 290 ~ 2340 MHz, horizontal: 290 ~ 2340 MHz
L. O. Frequency: 10.41 GHz
L. O. Frequency Stability: +/-0.5 MHz (Max) @ Room Temp., +/- MHz (Max) @ -40 °C ~ +60 °C
Local Oscillator Frequency Phase Noise: -50dBc/Hz @ 1kHz (Max.), -75dBc/Hz @ 10kHz (Max.), -100dBc/Hz @ 100kHz (Max.)
Noise Figure: 0.1dB (Typ.)
The 0.1 dB NF value is very low in my view. Probably very optimistic.
Here is the actual device used in my "Black buster" LNB for example: NE3503M04
• Super low noise figure and high associated gain
NF = 0.45 dB TYP., Ga = 12.0 dB TYP. @ VDS = 2 V, ID = 10 mA, f = 12 GHz
Anything under 0.3 is a miracle, before you divide the total NF of the remainder of the RF path by the front end gain, and add, to get overall NF.
Mike
There are many colours, Black being one I never knew about - you could then receive Es'hail2 in total darkness!
Three case tips I have are
a) I used a pair of dividers to transfer the PCB component offsets from the PCB in-the-case to the front panel.
b) I had some Red nail vanish here (its for my motorbike tail light) so you can eliminate the glare from the exposed aluminium cuts-out sides using that.
c) Measure at least twice before you cut and use tape to protect the nice coloured anodized finish.
And yes, they do appear to be 10.41 GHz.
Here is the data from a SystemSat one:
Input Frequency Range: 10.7 ~ 12.75 GHz
Output Frequency Range: Vertical - 290 ~ 2340 MHz, horizontal: 290 ~ 2340 MHz
L. O. Frequency: 10.41 GHz
L. O. Frequency Stability: +/-0.5 MHz (Max) @ Room Temp., +/- MHz (Max) @ -40 °C ~ +60 °C
Local Oscillator Frequency Phase Noise: -50dBc/Hz @ 1kHz (Max.), -75dBc/Hz @ 10kHz (Max.), -100dBc/Hz @ 100kHz (Max.)
Noise Figure: 0.1dB (Typ.)
The 0.1 dB NF value is very low in my view. Probably very optimistic.
Here is the actual device used in my "Black buster" LNB for example: NE3503M04
• Super low noise figure and high associated gain
NF = 0.45 dB TYP., Ga = 12.0 dB TYP. @ VDS = 2 V, ID = 10 mA, f = 12 GHz
Anything under 0.3 is a miracle, before you divide the total NF of the remainder of the RF path by the front end gain, and add, to get overall NF.
Mike
Re: winterhill case?
HDMI:
I have added a female full-sized HDMI socket to the rear panel on my case so I can use my WH stand alone.
Or for anything else requiring no added PC control, just a monitor and mouse and KB onto the Pi.
Or anything else you may require.
The tricky bit is the sharp right angle bend in the cable coming from the RPi Micro HDMI. This connector is right next to the inside of the case.
I bought a Micro HD-MI Male to HD-MI Female Extension Cable with Screw Panel Mount from Ali Ex then carefully cut back the stain relief that was moulded as part of the Micro end plug.
https://www.aliexpress.com/item/4001128 ... 5cb2alLZI5
It works, but it is tight. Takes some care in bending the cable so there is no residual strain on the Pi then offer it all up.
I have added a female full-sized HDMI socket to the rear panel on my case so I can use my WH stand alone.
Or for anything else requiring no added PC control, just a monitor and mouse and KB onto the Pi.
Or anything else you may require.
The tricky bit is the sharp right angle bend in the cable coming from the RPi Micro HDMI. This connector is right next to the inside of the case.
I bought a Micro HD-MI Male to HD-MI Female Extension Cable with Screw Panel Mount from Ali Ex then carefully cut back the stain relief that was moulded as part of the Micro end plug.
https://www.aliexpress.com/item/4001128 ... 5cb2alLZI5
It works, but it is tight. Takes some care in bending the cable so there is no residual strain on the Pi then offer it all up.
Re: winterhill case?
That's the front panel.
I used a 90 degree micro HDMI cable, fits nicely, with a panel mount HDMI adapter. Don't try to use a straight one as it probably won't fit.
Mike
I used a 90 degree micro HDMI cable, fits nicely, with a panel mount HDMI adapter. Don't try to use a straight one as it probably won't fit.
Mike