Contact an Expert

Thread: How measure the electrical length with HP 8719C VNA?


Permlink Replies: 18 - Pages: 2 [ 1 2 | Next ] - Last Post: Apr 30, 2012 7:45 AM Last Post By: Dr_joel
neogsj

Posts: 12
Registered: 11/08/11
How measure the electrical length with HP 8719C VNA?
Posted: Nov 9, 2011 12:24 PM
Click to report abuse...   Click to reply to this thread Reply
Hi,

I need to measure a unknown adapter's electrical length with HP 8719C VNA.
Is there any direct way that I can measure the electrical length?

I have a adapter (94.75ps) of 85052D 3.5mm economy calibration kit as a reference. I tried to measure the electrical length of the calibration kit adapter by DELAY in the VNA, but it does not give the same result. I do not know if the group delay is related with the electrical length.

So how could I measure the electrical length correctly with VNA?

Thank you.
Dr_joel


Posts: 2,744
Registered: 12/01/05
Re: How measure the electrical length with HP 8719C VNA?
Posted: Nov 10, 2011 7:44 AM   in response to: neogsj in response to: neogsj
Click to report abuse...   Click to reply to this thread Reply
Group Delay is the same as electrical length.

The 8719C is quite old, and doesn't use the modern calibration methods; still this should be no problem

What is the kind of adapter that you are using? Is it male-to-male, M-F, or F-F? Ifit is M-F, it relatively easy. If it is F-F or M-M, you need to do apdater removal of some kind. What type to do you have.?
neogsj

Posts: 12
Registered: 11/08/11
Re: How measure the electrical length with HP 8719C VNA?
Posted: Nov 10, 2011 2:41 PM   in response to: neogsj in response to: neogsj
Click to report abuse...   Click to reply to this thread Reply
Thank you for your reply.

I have two adapters, the unknown one is SMA F-F, the one as a reference is SMA F-M.
I have 85052D calibration kit for the VNA calibration. There are open, short and load standards with male and female SMA connector.
Dr_joel


Posts: 2,744
Registered: 12/01/05
Re: How measure the electrical length with HP 8719C VNA?
Posted: Nov 10, 2011 3:22 PM   in response to: neogsj in response to: neogsj
Click to report abuse...   Click to reply to this thread Reply
If your firmware can do adaper removal, that is the thing to use. I don't know the number for the 8719,it was 6.12 or greater for the 8753.

Otherwise, the simpest way is to do a one port cal on the test port, then add a short from the calkit. It is really an offset short, so it will not be a dot, but then do data->mem and data/mem to get a dot. Then add your adapter and put on the other short from the calkit. These shorts, even though they have delay, are offset shorts with the same offset. Data/mem removes the short offset. Read the delay number and it is 2x times the delay of your adapter. Turn on trace statistics to show the mean value of the delay, and you can remove the effect of residual ripple or noise even more. Be sure to first put the VNA in stepped sweep mode.
neogsj

Posts: 12
Registered: 11/08/11
Re: How measure the electrical length with HP 8719C VNA?
Posted: Nov 11, 2011 8:25 AM   in response to: neogsj in response to: neogsj
Click to report abuse...   Click to reply to this thread Reply
Thank you very much~ That works!!!
neogsj

Posts: 12
Registered: 11/08/11
Re: How measure the electrical length with HP 8719C VNA?
Posted: Apr 23, 2012 3:56 PM   in response to: Dr_joel in response to: Dr_joel
Click to report abuse...   Click to reply to this thread Reply
Dr_joel wrote:
If your firmware can do adaper removal, that is the thing to use. I don't know the number for the 8719,it was 6.12 or greater for the 8753.

Otherwise, the simpest way is to do a one port cal on the test port, then add a short from the calkit. It is really an offset short, so it will not be a dot, but then do data->mem and data/mem to get a dot. Then add your adapter and put on the other short from the calkit. These shorts, even though they have delay, are offset shorts with the same offset. Data/mem removes the short offset. Read the delay number and it is 2x times the delay of your adapter. Turn on trace statistics to show the mean value of the delay, and you can remove the effect of residual ripple or noise even more. Be sure to first put the VNA in stepped sweep mode.

Hi, Dr joel,
I used this method to obtain the delay of my adapter. I read the delay number (144 ps) from VNA and divide by 2, I have 72 ps. In fact, my adapter consists of two same SMA connectors back to back. For having the delay of one SMA connector, I divide the delay of the adapter again by 2, so the delay for one SMA connector is 36 ps.
With this number 36 ps, I want to compensate the SMA connector by setup electrical delay in Scale for S11. I fabricated one passive circuit with this SMA connectors (two ports with the same SMA connectors) and connected between two port of VNA. The objective is to measure the S11 of my circuit. However, I cannot have similar phase result as the simulation (magnitude is ok). Then I tuned the electrical delay by myself. The best result is found when the electrical delay is 18 ps.

Should I input half of the electrical delay of connector for electrical delay if I want to measure S11? I tried to find some document which can support my assumption, but I did not find.

So is this method correct for removing the connector from the circuit, so that I can measure the S11 phase correctly?

Thank you.
Dr_joel


Posts: 2,744
Registered: 12/01/05
Re: How measure the electrical length with HP 8719C VNA?
Posted: Apr 23, 2012 10:43 PM   in response to: neogsj in response to: neogsj
Click to report abuse...   Click to reply to this thread Reply
For your initial measurment, did you compensate the value for the delay of the calibration standard short?

Did you use a short or open to terminate the first measurement? What cal-kit short or open did you use.
neogsj

Posts: 12
Registered: 11/08/11
Re: How measure the electrical length with HP 8719C VNA?
Posted: Apr 24, 2012 7:59 AM   in response to: Dr_joel in response to: Dr_joel
Click to report abuse...   Click to reply to this thread Reply
Dr_joel wrote:
For your initial measurment, did you compensate the value for the delay of the calibration standard short?

Did you use a short or open to terminate the first measurement? What cal-kit short or open did you use.


For my initial measurement
I first did a one port cal on the test port, then added a short from the 85052D 3.5mm economy calibration kit. As you said, it is an offset short, so the S11 was not a dot, then I did data->mem and data/mem to get a dot.
Then added my adapter and put on the short from the 85052D 3.5mm economy calibration kit.
Then Turn on trace statistics to show the mean value of the delay, and I can have 144 ps delay number and it is 2x times the delay of my adapter. (The VNA in stepped sweep mode)
Dr_joel


Posts: 2,744
Registered: 12/01/05
Re: How measure the electrical length with HP 8719C VNA?
Posted: Apr 24, 2012 8:09 AM   in response to: neogsj in response to: neogsj
Click to report abuse...   Click to reply to this thread Reply
That all sounds right, and 144 psec for S11 delay does imply 72 psec one-way delay. That means your pair of SMA adapters were on the order of 21-30 mm long (about 1 inch). I get this from c*delay and presume a velocity factor of .75 - 1. If it is SMA, it will typically have a PTFE dielectric, so the that sets the majority of the velocity factor. Is you pair of SMA connectors on this order of physical length?

When you do the single connector measurement, exacly how do you terminate the single connector? Open or short?

Can you post a picture? You might be seeing some fringing effects.
neogsj

Posts: 12
Registered: 11/08/11
Re: How measure the electrical length with HP 8719C VNA?
Posted: Apr 24, 2012 9:38 AM   in response to: Dr_joel in response to: Dr_joel
Click to report abuse...   Click to reply to this thread Reply
Dr_joel wrote:
That all sounds right, and 144 psec for S11 delay does imply 72 psec one-way delay. That means your pair of SMA adapters were on the order of 21-30 mm long (about 1 inch). I get this from c*delay and presume a velocity factor of .75 - 1. If it is SMA, it will typically have a PTFE dielectric, so the that sets the majority of the velocity factor. Is you pair of SMA connectors on this order of physical length?

When you do the single connector measurement, exacly how do you terminate the single connector? Open or short?

Can you post a picture? You might be seeing some fringing effects.


For the connector, the dielectric is PTFE which has a dielectric constant of 2. The pic of the SMA adapter is attached. I did the calculation the physical length of pair SMA connector is about 598 mil, so for one connector, it is 299 mil. I use the calculation: physical length = delay*c/√2. The calculated physical length is match the measured physical length for this connector.

Then, I re-calibrate the VNA with 2-port, make S11 and S22 as a dot by using the data->mem and data/mem when port 1 and 2 connected with short/open. I fabricated a PCB circuit containing a shunt open-circuited microstrip line with quarterwave length at 4.28GHz. I added two SMA connectors at the input and output of the circuit as shown in the attached fig. (circuit with SMA connector.jpg).
Then I measure the phase of S11 at 4.28 GHz with this circuit connected between two ports.
By changing the electrical delay in Scale to 18 ps (not 36 ps), I can have the result similar to the momentum result. It is just a coincidence or the method is wrong?
The objective is removing the connectors from the circuit by using the electrical delay in Scale, so I can measure the phase correctly at harmonic frequency.
Dr_joel


Posts: 2,744
Registered: 12/01/05
Re: How measure the electrical length with HP 8719C VNA?
Posted: Apr 24, 2012 9:51 AM   in response to: neogsj in response to: neogsj
Click to report abuse...   Click to reply to this thread Reply
neogsj wrote:
Then, I re-calibrate the VNA with 2-port, make S11 and S22 as a dot by using the data->mem and data/mem when port 1 and 2 connected with short/open.

BZZZT! There's your problem. After you calibrate, the reference plane is already set correctly. The open and short are not really open and short, but they are offset-open and offset-short

when you normalize to them, you are resetting the reference plane by about 17.5 psec (as I recall). So that explains your results. Just do the cal and don't normalize.

In the first instance (reflection) the normalization is good as you use the short on both the test port and the end of your DUT, so the offset in the measurement is compensated by the offest on normalization at the test port. In the second instance, you don't use the short for measurement, so no offset compensation is needed.
neogsj

Posts: 12
Registered: 11/08/11
Re: How measure the electrical length with HP 8719C VNA?
Posted: Apr 24, 2012 12:07 PM   in response to: Dr_joel in response to: Dr_joel
Click to report abuse...   Click to reply to this thread Reply
Dr_joel wrote:
neogsj wrote:
Then, I re-calibrate the VNA with 2-port, make S11 and S22 as a dot by using the data->mem and data/mem when port 1 and 2 connected with short/open.

BZZZT! There's your problem. After you calibrate, the reference plane is already set correctly. The open and short are not really open and short, but they are offset-open and offset-short

when you normalize to them, you are resetting the reference plane by about 17.5 psec (as I recall). So that explains your results. Just do the cal and don't normalize.

In the first instance (reflection) the normalization is good as you use the short on both the test port and the end of your DUT, so the offset in the measurement is compensated by the offest on normalization at the test port. In the second instance, you don't use the short for measurement, so no offset compensation is needed.


So when I measure my circuit, I just calibrate the VNA with the 85052D 3.5mm economy calibration kit. After calibration I do not need to make S11 and S22 as a dot when port 1 and 2 connected with short/open. Then connect my circuit between 2 ports, set 32 ps electrical delay in Scale for S11. Is that it?

I will try it. Thank you.
neogsj

Posts: 12
Registered: 11/08/11
Re: How measure the electrical length with HP 8719C VNA?
Posted: Apr 24, 2012 10:21 PM   in response to: Dr_joel in response to: Dr_joel
Click to report abuse...   Click to reply to this thread Reply
Dr_joel wrote:
neogsj wrote:
when you normalize to them, you are resetting the reference plane by about 17.5 psec (as I recall). So that explains your results. Just do the cal and don't normalize.

Hi, Dr joel,

I measured my circuit without normalizing. After I set the electrical delay 36 ps in S11, I still cannot have similar result as Momentum simulation (about 70 degree difference). I tried the same procedure with 8720ES VNA, I still have about 70 degree difference. (Simulation result is verified)

And when I normalize the VNA with a OPEN from 85052D cal kit, the reference plane is reset by about 63 ps (measured data).

So I do following steps to figure out how could I use electrical delay in Scale:
1. Calibrate the VNA by 85052D cal kit.
2. Without normalizing, I measure S11 phase for my circuit at 4.28 GHz (port1 of the VNA - DUT - port 2 of the VNA), I have 142 degree.
3. I add an electrical length known adapter from 85052D cal kit (94.75ps) between port 1 of the VNA and the input port of my circuit (port1 of the VNA - adapter - DUT - port 2 of the VNA). I read the S11 phase at 4.28 GHz. Because of the added adapter, I have another phase at 4.28 GHz: -150 degree.
4. For removing the adapter added at the input of my circuit by using electrical delay in Scale, I tuned the electrical delay for S11 to have -142 degree. When the electrical delay reach 189.95 ps, I can have 142 degree. 189.95 ps is approximately 2 times of the electrical length of the added adapter.

Based on that I assume maybe I should try 72 ps instead of 36 ps, the error reduced but still I have 14 degree phase difference at
4.28 GHz. I can have similar phase when I tune the electrical delay to ≈83 ps if I do not normalization.

If I do the normalization with the open from 85052D cal kit after fabrication, when I measure my circuit, I have to tune ≈18-20 ps to have similar result.

If I do the normalization without connecting any standard component after fabrication, just open the port 1 of VNA, when I measure my circuit, I have to tune 72 ps to have similar result. <<<==based on the result, this method seems correct.

I do not know which one is right way to measure the phase of circuit correctly if I want to remove the connector's effect.

Thank you.
Dr_joel


Posts: 2,744
Registered: 12/01/05
Re: How measure the electrical length with HP 8719C VNA?
Posted: Apr 25, 2012 12:15 AM   in response to: neogsj in response to: neogsj
Click to report abuse...   Click to reply to this thread Reply
The trouble here is you are using electrical delay to do a "port extension" function, which is the cause of your problems. You can use electrical delay to remove delay from a measurement, and it will have 2:1 phase response for reflection measurments and 1:1 for transmission. But that is not the proper function to use. If you use port extension to remove delay from your measurement then it automatically accounts for relfection vs. transmission measurment. Stop using electrical delay and try port extensions for this purpose. I think you will find your troubles go away.
neogsj

Posts: 12
Registered: 11/08/11
Re: How measure the electrical length with HP 8719C VNA?
Posted: Apr 25, 2012 7:33 AM   in response to: Dr_joel in response to: Dr_joel
Click to report abuse...   Click to reply to this thread Reply
Dr_joel wrote:
The trouble here is you are using electrical delay to do a "port extension" function, which is the cause of your problems. You can use electrical delay to remove delay from a measurement, and it will have 2:1 phase response for reflection measurments and 1:1 for transmission. But that is not the proper function to use. If you use port extension to remove delay from your measurement then it automatically accounts for relfection vs. transmission measurment. Stop using electrical delay and try port extensions for this purpose. I think you will find your troubles go away.

I tried to measure the DUT without normalization with using port extension. By input 36 ps or 18 ps in Extension input A, I still do not have correct phase result for S11.

Then, I measured the DUT with normalization. By input 18 ps in Extension input A, I can have correct phase result for S11 now.

But I am still confused by the normalization, electrical delay and port extension even I have the correct phase result now. Is the last method by which I can have good result is correct?

Edited by: bharti bajaj on Apr 26, 2012 6:48 AM

Point your RSS reader here for a feed of the latest messages in all forums