Why are there 2 transformers in a phase shifter?

[kevinj]

This is a circuit from mini*circuits. You can get any phase from 0 degrees to 180 degrees by applying 0V to 15V at the control input. Put your signal at the input and it will be phase shifted at the output. The device works from 700MHz to 1GHz.

The black things are varactors(not sure what range is) but they can operate from 0V to 15V.

The blue things with the red dot are inductor (8nH).

There are resistors (202 marking)

Caps are small.

Questions : there are 2 transformers, why? Why not just 1? Are they baluns? Does anyone know how this circuit works?

The windings are all in the same direction. There ratio is 4:4.

In the first picture there are components circled. That's because they are coupled together to act as the transformer. The yellow are coupled to each other and the red are coupled to each other. There coupling factor is 1.

If you look closely at the 3rd picture you'll notice that they switched the red and the green wires. If they didn't do this then they would have a different layout, but I don't think that would matter much. If there is another reason let me know. Thanks.

 

[flatulent]

varactor phase shifter

You should draw out the schematic. I suspect this works by combining two paths in different ratios. The two paths have fixed phase shifts and the ratioing is done with a diode attenuator.

 

[kevinj]

bridge phase shifter

I don't really think the circuit uses a diode attenuator because there is only 1.75dB insertion loss. Also, it seems that the 2 black things are varactors because they are in series with inductors. It looks like a LC filter that actually changes phase. What do you think?

 

[Davood Amerion]

differential phase shifter all pass transformer

Dear kevinj
Yes black things are varactors.
I think your drawing is wrong.
Your mistake is in L10-13 pins.
Anyway, you said control volate is 0-15V, and i think that varactor capacitance for 15/2 = 7.5Volt is about 5-10PF .
with these value L0,C10 and L1,C9 resonance frequency will be about 800-1000GHz.
if control voltage is lower than 7.5 volt resonance circuit act as inductor.
in the other word inductance effect will be dominant.
and if V control is higher than 7.5V capacitance will be dominant.
I dont know why it uses 2 transformer.
with this explanation , depend on control voltage one capacitor or one inductor will be series in input-output pass.

Regards,
Davood Amerion.

 

[kevinj]

Re: Phase shifter

Sorry, I uploaded the circuit drawing because it should have had varactors (reversed biased diodes) instead of caps, C10 & C9.

Davood, I think you are right about the varactor values. What formula are you using to figure 5pF to 10pF?

I've been using this model, zv952v2, and produced these results. I can't get the 180 degrees and my insertion loss is too large.

I resimulated with this newer circuit drawing and produced the same results just to make sure my original circuit is correct. The reason I don't switch L10 & L13 like you suggest is because I wanted it to look just like the circuit layout from the company.

I couple to transformers using a 'mind'. If you look on the left of the schematic you will see them.

Also, I only simulate to 12V because that is the breakdown of the varactor, and I only have that model file.

Let me know about why there is so much insertion loss and about the transformers if you can. Thanks.

 

[Davood Amerion]

Re: Phase shifter

the resonance formulla: Fr=1/(2Π√LC)

see drawings [phaseshifter1]:

second drawing [phaseshifter2]is equvalent circuit

and i must to correct my sentens:
"if control voltage is higher than 7.5 volt resonance circuit act as inductor.
in the other word inductance effect will be dominant.
and if V control is lower than 7.5V capacitance will be dominant.
"

 

[flatulent]

Re: Phase shifter

There are some hidden grounds missing. You should remove the parts and measure the resistance between all pairs of traces.

You cannot get +-90 degree phase shift with a tuned circuit without having high attenuation beyond the +-45 degree points. I suspect that the network is an all pass type. These are either bridge form or T form with a center tapped inductor.

 

[kevinj]

Re: Phase shifter

Davood,

I'm not sure about your schematic. The 2 transformers are connected strangely. Why didn't you just draw 2 straight lines? The circuit is the same if you cross connect the 2 transformers together.

Flatulent,

I took all the components off and the circuit looks the same as I have it posted. I didn't find a center tap. The circuit seems like a resonant circuit like Davood said, but I do agree that there is a lot of loss after +/- 45 degrees--just like the plot I posted earlier.

Not sure who is correct?

 

[flatulent]

Re: Phase shifter

I still think that there are some hidden connections. An all pass network will produce the +-90 degree phase shift with no amplitude attenuation and there are the exact number and type of parts in the circuit to make a bridge all pass network. One transformer connects the top and bottom nodes of the bridge to the input. One transformer connects the left and right nodes of the bridge to the output. The bridge has two L and two C elements. The left side has L top and C bottom. The right side has C top and L bottom. Then there is the bias network for the variable C elements.

 

[kevinj]

Re: Phase shifter

Flatulent, I looked up bridged circuits and found what I think you are talking about. I am pretty sure that you ARE right. I am going to try to simulate the circuit this coming week. The transformers are needed, I believe, to make a signal single ended to differential at the input and output. Thanks a lot!!