BATC Forum

The updated prototype pcb is working and all being well, it should be going to the manufacturer in the next few day after a few minor tweaks.

I've used the adjustable version of the MCP1826 instead of an SMD type, as it's easier to work with. The sockets for the NIM module are optional - the NIM will fit directly onto the pcb. It may need some extra heatsinking if it's going to be used at high symbol rates with the low SR box unticked.
I went over to Warrington ARC's ATV day this afternoon, which gave me the chance to test the pcb at 125k symbols / second with a DATV-Express transmitter. The tx power was adjusted until the MER reached 30dB. Does a signal strength of -68dBm suggest that the power supplies are low noise? I wonder what others have been getting.

Brian

It is difficult to give an answer with just this test.
using Samsung Minitiouner I have this result : 30 dB MER for -80 dBm input ( in reality -95 dBm input if we know that the booster add 15 dB) Jean-Pierre F6DZP

Is that through an antenna or directly connected to the transmitter with attenuators?

Do you have a similar test for an Eardatek or Sharp? How much boost is there in the Eardatek?

Brian

same signal at same time ==> Eardatek same signal at same time ==>TT S2-1600 (no booster)
Modulator is a Dektec DTA107 : -27 dBm
attenuator HP 394A (-58dB)
4 way splitter about -12dB attenuation
signal is sent at same time to all receivers:
-Tutioune1600 N1
- Minitioune Samsung
- Minitioune Eardatek
-Tutioune1600 N2

I have no Sharp built to try at same time.

Jean-Pierre F6DZP

Very interesting. Does that suggest that the Samsung is a better receiver?

Could you do two more tests please:

On the Eardatek, increase the signal level until the MER reaches 30dB. Then I can compare it directly with my test.

On the Eardatek, decrease the signal level until the MER reaches 3dB and see what the Samsung is showing.

Brian

my result.
we cannot compare the NIM, only the whole Minitiouner, construction and power supply are different.

Jean-Pierre F6DZP

Thanks. Very informative. My Eardatek seems very similar to F9ZG's, so I'll assume it's working reasonably well.

Brian

Very interesting results,
So far I have had best performance from the Eardatek, but my Samsung is a DNBU105121ST not a DNBU10711IST , it has been imposable to find the Samsung DNBU10711IST ,( I wonder what equipment was using that one?). However I was using a switching regulator with two much ripple on both for the initial tests.
Will redo the power supply on my Samsung and test again.
I had already changed the power supply arrangement on the Eardatek, to the same devices as Brian will use on the PCB, ripple is reduced and nothing gets hot either.( no heat sink needed for SR 100-4000 but if you want to use it for Sat TV it will draw more current as has been mentioned before)

Jean-Pierre F6DZP has performed a miracle, a system that performs better than a TT 1600 and costs less très bien

I did some tests the other week with 333ks signal in the different receivers.

Setup connected in series:
DATV Express 333ks -10dBm.
R&S step atenuator -100dB +/-
1dB NF preamp
PYE 2m bandpass filter.
Mar-3 20dB gain mmic
Modified Glencom up converter 146.5 > 1803.5MHz.

This setup assures that there is enough IF signal getting to the receiver card or NIM to rule out the different card/NIM noise figure.

This tests purely how the Rx works with fixed S/N level.
All Receivers are near -60dBm input level, well above the noise floor with no tx switched on.

I set the attenuator so I had full lock and 5dB MER using the USB rx.
(For reference this was -107dBm into the 1dB NF lna)

The Eardatek NIM, 1600 and 3200 cards were tested and all achieved very similar result less than 0.5dB away from 5dB MER, the vber on each receiver was also very close between 7 and 15.

The only difference was that the 3200 card took longer to lock.. but once locked it held in just the same as the others with same vber.

So my tests conclude the current receivers have virtually no difference on weak signal performance when you have an lna in front of them.

Who runs their rx without an lna?!

I hope this is also useful information

Rob