Hi Sir,

I have a load measured by network analyzer as below, at 500kHz its impedance is 234.72 - j*1173.2, and I am targeting to match its impedance to 50 Ohm through impedance matching technique.


as the first step, I want to transfer the real part of impedance to 50 Ohm, and in the Smith chart tool I found using a 1:0.46 transformer can transfer the impedance from 234-j*1173(point : DP1) to 50-j*253(point : TP2)


since 1:0.46 is about 2:1, I found a 2:1 transformer "Coilcraft UA7962" ( https://forum.allaboutcircuits.com/forums/general-electronics-chat.5/post-thread , used in DC/DC switching converter), insert the transformer between network analyzer and the load then measure again, but the load impedance is transferred to 6+j*273, real part is "6" which is far from the expected "50".



then I tried another transformer "Mini-circuits TC4-6TG2+" ( https://www.minicircuits.com/WebStore/dashboard.html?model=TC4-6TG2+ , used in the RF field, note its Ohm ratio is 4:1 means turns ratio is 2:1)
but the impedance is transferred to 156+j*92, again, the real part is far from the target 50 Ohm



I know there must be error between theoretical simulation and the real world, however, the above result shows not only the error is large, but also the behavior of a transformer is far from simulation.

Please give me some instruction about what's wrong in my tests, am I using the wrong transformers ? is the error caused by unavoidable test setup error ? or the other reasons ?

I have no idea about what would be the next step, any suggestion would be highly appreciated, thanks.

 

Coilcraft UA7962 is not suitable for frequencies above 300kHz.

 

Your load has a reactive impedance and thus it requires more than just matching the resistive portion of the impedance. IN addition, the match will be frequency sensitive no matter what arrangement you use.
DC to DC switching converters usually operate in a very narrow frequency range and thus, if they are correctly designed, the various impedance points are well matched. ( That is a big part of why I buy switcher supplies instead of designing them. They are more complex than they would seem to be.

 

"Mini-circuits TC4-6TG2+"

How long are the coax cables? Where is the termination resistor? What value? Which winding are you driving?
---edited----
I would start out by using the center tap of the transformer to make a 1:1 transformer. Have 50 termination. Test.
Maybe even go back to no transformer, just solder the coax together and test.

 

How long are the coax cables? Where is the termination resistor? What value? Which winding are you driving?
---edited----
I would start out by using the center tap of the transformer to make a 1:1 transformer. Have 50 termination. Test.
Maybe even go back to no transformer, just solder the coax together and test.

the coax cables is about 10cm long at the two sides of transformer, and the transformer is inserted between network analyzer and the load. Before inserting the transformer (ie. network analyzer measure the load directly), at 500kHz impedance is 234-j*1173(reactive), and after inserting the transformer the impedance changes to 156+j*92(inductive). According to the theory, after inserting a transformer the impedance should still be reactive, but the measured value is inductive, that's why I am posting this thread and asking for help, thanks.

 

The theoretical solution is clearly incorrect because matching a reactive impedance requires a reactive matching arrangement, which a simple transformer is not going to provide.
I suggest a review of the AC Circuit theory textbook concerning AC impedance.

 

The theoretical solution is clearly incorrect because matching a reactive impedance requires a reactive matching arrangement, which a simple transformer is not going to provide.
I suggest a review of the AC Circuit theory textbook concerning AC impedance.

Hi MisterBill2,

using a transformer is the "first step" to transform the real part of impedance to 50 Ohm(DP1 to TP2), then my second step is placing an inductor to cancel the reactive part of the impedance(TP2 tp TP3), see the figure below



but test result of the first step deviate with the theory a lot, this is why I am asking help here, thanks for your reply !

 

the coax cables is about 10cm long at the two sides of transformer, and the transformer is inserted between network analyzer and the load. Before inserting the transformer (ie. network analyzer measure the load directly), at 500kHz impedance is 234-j*1173(reactive), and after inserting the transformer the impedance changes to 156+j*92(inductive). According to the theory, after inserting a transformer the impedance should still be reactive, but the measured value is inductive, that's why I am posting this thread and asking for help, thanks.

View attachment 322133

Hi ronsimpson,

I tested the 2:1 transformer with a 50 Ohm load, if turns ratio "1" connects to 50 Ohm load and "2" connects to network analyzer, I am expecting 200 Ohm on all frequencies and below is the result



and if turns ratio "2" connects to 50 Ohm load and "1" connects to network analyzer, I am expecting 12.5 Ohm on all frequencies and below is the result


looks like measurement result is even less as expected at low frequencies (ie. 500kHz), and referring to the datasheet of Mini-circuit TC4-6TG2 the guaranteed operation frequency range is 1.5~600MHz, so maybe RF transformers is not suitable for my application ?

I guess I need to find another transformer, but I have no idea where to find a transformer off-the-shelf that may meets my requirement, my requirement :

1. operation frequency : 100kHz~1MHz
2. able to sustain 100Vpp impulse transmit waveform (for example, 500kHz/5 cycle/100Vpp pulse, pulse repetition rate about 100us)

Or maybe I need to make a transformer by myself ? Is there any learning resources on the Internet ?

Any suggestion would be appreciated, thanks.

 

The TC4-6TG2 transformer is rated for 1.5mhz to 600mhz.
You are getting 215 ohms @ 1.5mhz and 201 ohms @ 2mhz close to 200 you expected.
You are getting 13 ohms @ 1.5mhz and 13 ohms @ 2mhz close to 12.5 you expected.

my requirement :

1. operation frequency : 100kHz~1MHz
2. able to sustain 100Vpp impulse transmit waveform (for example, 500kHz/5 cycle/100Vpp pulse, pulse repetition rate about 100us)

100khz... you need a transformer with more inductance.
100Vpp that will take a bigger transformer. You need a transformer that can handle 500 volt uSeconds.
--------------edited----------------
Wrong. 100Vpp is 50Vpeak for about 5uS = 250VuS. 5uS is 1/2 cycle of 100khz.
The TC5-6TG2 transformer is often used in the input of a TV set or radio. I don't see a voltage x time number, but it is probably very low.
----------------------------------------

It looks like you are sending power through the transformer. Here are something to look at. It is a Gate Driver Transformer. It is not rated for bandwidth. GateDriver
Here is another transformer. 13mH This transformer lists inductance and capacitance so you can find the self-resonant frequency.
Another 1024C
Base/Gate Driver Transformers (coilcraft.com)
Hope this helps. What are you building?

 

Certainly the frequency handling ability of the transformer must include the frequency being transformed. My low frequency matching experience was matching the 8 ohm output of a band's amplifier to the "70.7 volt" speaker array in the bar they were playing at. It was a rather emergency situation. I used a 250 watt "variac" transformer backwards and the results were good. The transformer frequency rating was 60 Hz and the music was well above that. ( This was in the days before "mega-bass" sound,)
My point being that certainly both the frequency rating and the power handling capability must be adequate for the application.
IN ADDITION, matching of a reactive impedance load and source is frequency sensitive, so I suggest starting with a much narrower frequency range.

 

The TC4-6TG2 transformer is rated for 1.5mhz to 600mhz.
You are getting 215 ohms @ 1.5mhz and 201 ohms @ 2mhz close to 200 you expected.
You are getting 13 ohms @ 1.5mhz and 13 ohms @ 2mhz close to 12.5 you expected.

100khz... you need a transformer with more inductance.
100Vpp that will take a bigger transformer. You need a transformer that can handle 500 volt uSeconds.
--------------edited----------------
Wrong. 100Vpp is 50Vpeak for about 5uS = 250VuS. 5uS is 1/2 cycle of 100khz.
The TC5-6TG2 transformer is often used in the input of a TV set or radio. I don't see a voltage x time number, but it is probably very low.
----------------------------------------

It looks like you are sending power through the transformer. Here are something to look at. It is a Gate Driver Transformer. It is not rated for bandwidth. GateDriver
Here is another transformer. 13mH This transformer lists inductance and capacitance so you can find the self-resonant frequency.
Another 1024C
Base/Gate Driver Transformers (coilcraft.com)
Hope this helps. What are you building?

Hi Ronsimpson,

I got a muRata 1025C with Er = 570Vus


and now the matching result is much more close to my expectation, real part is close to 50Ohm(54-j*178@500kHz), and then I can further add a series inductor to cancel the imaginary part (56-j*35@500kHz).



after the matching the receive signal is improved a lot, thanks for your valuable suggestion !