Ka B/C overlap (Hughes) Thread

Llew · Jul 27, 2014

It's all I had at the time VS. I'm sure a lower spec regulator will do fine, the osc takes ~30mA, the comparator maybe sinking 20mA max I would have thought, but it's not going to be anything like that to the input of the osc.

Llew · Jul 27, 2014

Just checked with the Avo -

30mA with switch on. 15mA with switch off.

Good to go

Vipersan · Jul 27, 2014

Llew said:
Just checked with the Avo -

30mA with switch on. 15mA with switch off.

Good to go

Cheers Llew ..
No changes needed ...much as I suspected.
Hopefully my new osc will be here tomorrow and I can continue the build ..
rgds
VS

Vipersan · Jul 28, 2014

Vipersan said:
Cheers Llew ..
No changes needed ...much as I suspected.
Hopefully my new osc will be here tomorrow and I can continue the build ..
rgds
VS

Done some more today ..
almost built but I'm still missing a couple of components ..
On order but not arrived ..
Here it is so far ..
rgds
VS

Llew · Jul 28, 2014

Very neat job so far VS. Just be sure your 100nF osc. decoupling capacitor is placed as close to the component as possible (pads 4 & 2). Spurious HF noise from oscillators and all that. Mine could be closer, but I suspect it's doing the job OK.

Vipersan · Jul 28, 2014

Ok ..here is a scan of Ka band A with the SL3 selected pointing at 42E ..
Currently the switch and osc are activated ..producing 27mhz which IS switched on and off on V/H scans..
BUT currently no 27mhz injected back into the IF line ..so active only inside the die cast box ..
This was to test the effect of the Comparator /osc device/ and psu connected to the 75 ohm breakout inside the box ..
So ALL connected but even though 27mhz is being generated it currently goes nowhere..
Hope that makes sense ..
Certainly the scan is clean and useable though lifts the noise floor at the higher frequencies.
rgds
VS

Vipersan · Jul 28, 2014

..and this scan is just the TBS card connected to the HV inline switch with NOTHING connected on the output side ..
ie ..Un-terminated
rgds
VS

Vipersan · Jul 28, 2014

So this is where it gets tricky ..
I have simply cap coupled the output of the osillator to the IF line ..
I did an unterminated scan at 42 E ..and one with the SL3 connected ..
You can see that it crushes the IF making the previously 10db turksat signal unusable ...
and standing or travelling waves evident in the unterminated scan.
this is simply loading the IF line via a 15pf capacitor to the osc and NOT entirely related to the 27mhz signal as a similar result can be seen in both Vertical and Horizontal scans ..
As I suspected ..a better method of injecting the 27mhz signal is needed ..
cap coupling is inadequate..
I am hoping a PIN diode might be the answer but may simply make it worse ..or not work at all..
As I have already said ..RF is a tricky unpredicatable beast..

Tomorrow I will try a smaller capacitor ...as there is already some capacitance between the IF line and the osc output as these 2 lines lie side by side on the verro board ..
I'd estimate in the order of 1 to 2pf ....and this does not crush the signal but it is doubtful sufficent 27mhz signal is coupled to the IF at this level of capacitance.
rgds
VS

hvdh · Jul 28, 2014

I think for the 27 MHz injection point, a ferrite choke (such as the 6-hole "pig snout"), or some ferrite beads, should be used in series, as close to the IF line as possible.
Such that 27 MHz gets through sufficiently, and the GHz IF frequencies don't.
Then the series capacitor value is not so relevant, and will only do the galvanic isolation.

Vipersan · Jul 28, 2014

hvdh said:
I think for the 27 MHz injection point, a ferrite choke (such as the 6-hole "pig snout"), or some ferrite beads, should be used in series, as close to the IF line as possible.
Such that 27 MHz gets through sufficiently, and the GHz IF frequencies don't.
Then the series capacitor value is not so relevant, and will only do the galvanic isolation.

Hi Hans ..
yes indeed I was contemplating winding a few turns on a small ferrite ring ...but just how may turns ...
I'm clueless in this regard..

any links to DIY construction of similar devices would be most useful....
rgds
VS

hvdh · Jul 28, 2014

This calls for (more) experimentation on your side, since I don't know what ferrites you have in your junk box!
Not many turns, if any, will be needed for > 1 GHz suppression.
A few days ago you mentioned that just a short piece of wire at the tap point, already had serious effect on the freq.scan curve.
So try the ferrite beads etc. you have on that piece of wire, when the curve looks good again, try that one to inject the 27 MHz. Via a series capacitor (1 nF or so), of course.

Llew · Jul 28, 2014

Something amiss there. Unterminated, it looks like no filtering of oscillator's harmonics. I think you need to get rid of those with a decent filter. Try what I use, an L filter, 100pF to ground at the output of the osc, and 100R out to LNB.

Can't think why you've lost the Turksat carrier, if it's just a straight through from the TBS to LNB (?)

I'll put my Norsat on the Gibby on 42E to compare my scan with yours tomorrow.

Vipersan · Jul 28, 2014

..found some 6 hole ferrites on ebay ..
£1 each ...
seems reasonable so ordered a couple..
400273798707
more waiting ..
lol

Llew · Jul 28, 2014

Just seen Hans's possible solution. Yes, worth trying if you don't try the filter I suggested.

Vipersan · Jul 28, 2014

Llew said:
Something amiss there. Unterminated, it looks like no filtering of oscillator's harmonics. I think you need to get rid of those with a decent filter. Try what I use, an L filter, 100pF to ground at the output of the osc, and 100R out to LNB.

Can't think why you've lost the Turksat carrier, if it's just a straight through from the TBS to LNB (?)

I'll put my Norsat on the Gibby on 42E to compare my scan with yours tomorrow.

I suspect it's a mismatch impedance wise ...
The loading effect of the oscillator device through the 15pf cap..?
I will of course include a filter as well ..but transformer coupled via ferrite may well be the way to go..
we shall see ..
but am happy with the construction thus far..
rgds
VS

Vipersan · Jul 29, 2014

Gotta rush I'm late opening the shop ..
but hope you are as impressed with this as I am Llew..
A scan at 9E using the hughes with the H port blocked with alluminium tape ..
the only changes made were the introduction of 2x ferrite beads onto the breakout legs on the IF cable under the pcb ..
still using a 15pf cap direct coupling ..
will try it on the SL3 at 42E later but not really relevant as it is only for use with the Hughes.
I cannot get the switch to operate inside at the TBS/PC output ..but not surprising considering the amount of standard switching I do ..
So the switch is currently in line outside connected directly to the Hughes ..with a short cable.
Photo to be added later..
here is the scan

Llew · Jul 29, 2014

That's a nice scan VS. No gremlins

Look forward to the next stage.

Vipersan · Jul 29, 2014

Ok ..a bit more testing ..
This time the switch was placed in line with the XMW and carl pointed at 7E ..
H/V switching makes no defference to this single polarity LNB - Ka band D - but I'm looking for signs of baseband gremlins ..
So the test was valid ..
A blind scan was performed and captured ..
Then the test repeated with the HV inline switch removed
Hard to detect the changes between the two scans ...on first glance.
The Switch is outside and close to the XMW ..connected by about 12 inches of sat cable..
rgds
VS

Vipersan · Jul 29, 2014

As promised a photo of the switch outside on the Hughes ..
Not finished yet as still testing and modding ..but eventually will be sealed and made waterproof..
rgds
VS

Llew · Jul 29, 2014

Do rou reckon the ferrites are making a difference? I've a mind to try them myself .

Ka B/C overlap (Hughes) Thread

cerca trova...

cerca trova...

Emmett Browns Ghost

Emmett Browns Ghost

Attachments

cerca trova...

Emmett Browns Ghost

Attachments

Emmett Browns Ghost

Attachments

Emmett Browns Ghost

Attachments

Specialist Contributor

Emmett Browns Ghost

Specialist Contributor

cerca trova...

Emmett Browns Ghost

cerca trova...

Emmett Browns Ghost

Emmett Browns Ghost

Attachments

cerca trova...

Emmett Browns Ghost

Attachments

Emmett Browns Ghost

Attachments

cerca trova...