strange transmission line

[xshou]

I did some reverse engineering of an optical receiver package. The make is a decent company.

And got some interesting findings:
there are a pair of 50Ohm transmission lines, CPW;
However, the center part of the traces (gold) were removed.

I'm not sure if there is any special coating on top of them.

Asked another experienced engineer; he can't explain it either.

Any clues? I would guess it might help damp possible oscillation; but how?

Thanks

 

[flatulent]

reduced losses

One guess is to reduce losses since gold has a higher resistance than copper. Another guess is to confuse you. I have heard of cases where organizations which reverse engineer things are sent specially modified products.

Back in the early days of microwave tubes individual ones with especially high performance were sent to competitors when they tried to buy samples to discourage them from trying to copy them. They could not match the performance.

 

[mike_bihan]

Excuse me, I am a little confused.
You mean that the gold layer is removed or the central metal trace is removed entirely.

 

[xshou]

Sorry for the confusion it might cause.

I mean:

the middle part of the trace was removed. But the two outter part of the trace remain.
the trace is shown as below
///////////////////////////////////////////
/////////----removed----------- ////////
//////////////////////////////////////////
shadow is gold layer; dash is empty(?), but might have special material underneath.

Groud is just as usual.

What I want to emphasize is: the weird Tline is aimed for broadband signal, virtually 73kHz up to 10GHz.

Thank you.

 

[junction_crosstalk]

Are they far away separated ? Maybe a kinda capacitive coupling is expected, such as filter application.

jct

 

[harkonnen]

Very interesting.
Maybe it is to adjust for eps_eff over that broad frequency range to keep the line constant 50 Ohm.

Have you simulated it? Can you give exact dimensions?

 

[Spasomat]

Hi,

I've also seen the described structure in a commercial broadband amp. I don't remember the exact location but I think it was near the output and there was just one of these obscure structures in the amp. The substrate or the structure weren't coated with anything. The whole backside of the substrate was cooper cladded.

Regards,

Spasomat

 

[xshou]

Spasomat is absolutely right. These Tlines are between the output of TIA and AC couple caps at the output of this receiver. The TIA is supposed to be broadband (40k~8GHz, 58dBOhm).

It's really hard for me to measure the dimension of the structure since it's within the tiny package and just too small.

Any one knows the name of this type of structure? or any reference/paper about it? Hopefully it's not patent related.

More interesting is the ac couple cap; it is also kind weird in terms of value and dimension. I would post measurement of the two caps next week.

 

[hobbymat]

This might be an impedance transformer. I haven't EM simulated this structure, but I'd strongly guess that a CPW line with a split center conductor has a different impedance than a similar line with a "normal conductor".

The reason for using this structure would probably be that the CPW ground plane will stay untouched, and the location and impedance of the "different" tuning line can be easily changed, even with a knife in prototyping phase.

In a normal microstrip line you can add tuning stubs of different width and length quite easily as the ground plane is not on the way. You can (and sometimes even have to!) do this with a small piece of foil and a soldering iron when you want to tune the circuit to a good performance. With a CPW line this method is generally not possible or at least easy.

Hobbie