Ultram LO modification before connecting to Digilite 5.7

[G8HCB]

Hi All,

My first post!

I have received my Ultram Oscillator today. It is set for 1249Mhz and does not look the same on the G8AJN site here:- http://www.g8ajn.tv/dlother1.html

I am assuming the MMIC on my board is just a different package. It has a number E00TG 0015 on it and I am not sure what part this is. I am guessing I just remove it, then take the output for the digilite L.O. from the first pin on the left of the E00TG 0015 package.

Picture included (next to a 2p coin for scale). Can anyone confirm if it is just a case of removing and connecting coax. Where should it be connected. Or more so - should I even be removing this E00TG 0015 as importantly the test sheet that came with my Ultram said that the output was already tested at +10dBm


I have no test equipment to measure such levels, so would appreciate some advice from anyone with experience regarding this board and the Digilite.

73 de Chris
G8HCB

[G8HCB]

*** Update - It's actually marked E001G and I have found out a component marked like this is an ECG001B-G which is a DC-6Ghz InGaP HBT Gain Block. (or MMIC)

Applications
• Mobile Infrastructure
• CATV / FTTX
• W-LAN / ISM
• RFID
• WiMAX / WiBro

According to the specs this should deliver 12.5 dBm P1dB (Output Power at 1 dB compression) at 1Ghz. I have read elsewhere the modulator chip needs 5-10 dBm and my test sheet for the ultram shows 10+dBm. Is this in P1dB (Output Power at 1 dB compression) or +25dBm OIP3 at 1Ghz - whatever that is? not sure. Also the quoted dB gain is 22 dB Gain at 1 GHz.

So now I am thinking it is not so straight forward as to just remove this. Maybe I need to attenuate in some manner to give a guaranteed between 5 and 10 dBm.

Any help still appreciated re P1dB and OIP3 dBm measurements and what to do here. I have the spec. sheet if anyone would like it.

And if I am overcomplicating things and it is just a case of removing the E001G then please someone with experience who has also done this, tell me it's OK to do this and go straight in to the LO on the Digilite and advise re best way to connect. i.e. where Pin 1 was - connect microwave coax inner, outer to nearest ground...

73
de
Chris
G8HCB

[G4EWJ]

The AD8346 spec sheet says the LO input should be between -6dBm and -12dBm. Others have found that having it nearer -6dBm improves the noise performance. There's a PI attenuator just before the balun, so this gives some leeway to vary the input at the board edge.

Without the ECG device, there will be about -10dBm coming out of the Ultram, so bypassing the PI attenuator (remove R28, R38 and make R27 zero) should get you in the ball park.

I don't use the Ultram myself, so wait for advice from someone who has one. I'm looking at someone else's DigiLite at the moment which has an Ultram and I've connected it as above. It works, but I need to increase the DigiLite power output to see how good it is on an analyser.

It's better not to generate 10mW of LO if you can help it, as the modulator output is much lower and you don't want any LO finding its way to the output.

Brian

[G8HCB]

Without the ECG device, there will be about -10dBm coming out of the Ultram, so bypassing the PI attenuator (remove R28, R38 and make R27 zero) should get you in the ball park.

OK Brian,

I have nothing on R38 or R28 so OK there. I have the optional 82R on R27 which I will bypass (link).

I will hold fire for a while in case anyone else replies with their thoughts.

Thanks for the info!

P.S. Has anyone got a photo of how they connected the coax to the Ultram board?

Chris
G8HCB

[G0KTD]

Digilite9.jpg (729.08 KiB) Viewed 12228 times

Hi I have removed the MMIC and linked with 0 ohm res and used SMA connector
to the board also on the out put I have a G4DDK amplifier
plus a 2 Watt Pa before the final PA Which should give me about 10/15 out?
I am also doing the VFO for it as well with LCD display then to Box it all up
as you can see all mounted on a piece of Board hoping to fit it in a Serial switch
Box with The Final PA on the top.

Keith G0KTD

[G8HCB]

Keith,

Excellent. I like the idea of bridging gap with the 0 Ohm SMT Resitstor. Nice and tidy. Was it a 1206 or 0805?

And I see you also used a 0 Ohm for R27.

Thanks for the post. Now I am starting to feel easier re connecting the LO thanks to input from Brian and yourself.

Chris
G8HCB

[G3GJA]

I've left all 3 of my Ultram oscillators intact and used 3 SM resistors to set the attentuation of R28 etc. to 12dB. Much easier than risking damge to the Ultram.

I've used SMA end launch female connectors on the Ultram output and the DigiLite input & output. I can null the carrier using the bias adjustments on the DigiLite board to -70dBc; in other words there's no leakage / radiation from the unmodified Ultram LO to be worried about if decent connectors are used.

[G4OHV]

Hi

I bridged the gap left after removing the micc with a 1206 0R resistor, swapped the next component on the output side for a 0805 0R, and removed another component at this junction, so SMA socket direct output

LO.jpg (221.06 KiB) Viewed 12197 times

Regards

Col

[g4bbh]

I use an unmodified Ultram in a very small diecast box. I used a long sma connector on the Ultram so it could be mounted through the box with nuts either side of the box wall and also fitted a feedthrough capacitor on the power feed. There is no detectable leakage of RF. However the high level of RF around the Digilite input connector and R27, R28, R38 and the balun can leak through to the ouput. I have place a small shield over this area (from input SMA connector up to the mod chip) and this has meant I can reliably set bias for carrier null and it stays nulled with no sign of carrier leakage.
Dick G4BBH

[G8HCB]

Hi,

After removing mmic and bridging with a 0 Ohm 1206 SMT - thats all...

I ended up welding the sma connector with a vector flame lighter and mounting in an old tobacco tin with a feedthrough cap. Yes smoking may be antisocial, but comes in handy sometimes.




Cheers

Chris
G8HCB