how to implement a delay line in Capture CIS?

[Kicchan]

Delay line with Capture

Hi everyone,

I need to implement a delay line in Capture CIS environment. My circuit provides an analog signal that must be delayed of 30ns. The delay elements present in the libraries are digital and cannot fulfill my needs. Do I have to use a transmission line?
I hope someone can help me!
Regards

 

[keith1200rs]

Delay line with Capture

Yes, you can use a transmission line for delaying analogue signals. You can define a transmission line by the impedance and delay you want for the "lossless" one, so it is pretty easy. Bear in mind you will lose half the signal. Remember to use the correct source & termination resistance.

Keith.

 

[Kicchan]

Re: Delay line with Capture

Thanks a lot for your help, Keith.
Do I have to choose a transmission line from the tline.olb library? Which part do you suggest to use?
Thank you in advance.

 

[keith1200rs]

Delay line with Capture

I don't use Orcad, so I am not sure - just set up a very simple simulation to check it - an AC source etc as shown. Run an AC analysis & plot the group delay. You could also try putting a pulse through it . The syntax for the transmission line I would use is

T1 IN REFIN OUT REFOUT Z0=50 TD=30n

Keith.

 

[Kicchan]

Re: Delay line with Capture

Thank you very much for your help!

 

[karesz]

Delay line with Capture

Hi,
How do you wish to realize this delay line pls, than per PCB/cable it means ca 6.5m...?
K.

 

[Kicchan]

Re: Delay line with Capture

Hi karesz,

is your question for me? I'm doing some simulations about an already existent circuit (CFD for a sensor readout). In the schematic I have they used an SMP1410 but I don't have any model and must use something else to simulate the daley.

 

[keith1200rs]

Re: Delay line with Capture

Hi karesz,

is your question for me? I'm doing some simulations about an already existent circuit (CFD for a sensor readout). In the schematic I have they used an SMP1410 but I don't have any model and must use something else to simulate the daley.

In which case you might want to use a lumped transmission line equivalent - it will be far closer to the real thing. This is the sort of circuit - but not the values for a 50 ohm 30ns delay. If you cannot find the equations for RLC let me know and I will look them up - I cannot remember them off the top of my head.

Keith.

 

[keith1200rs]

Delay line with Capture

and if you want it to be more realistic it should have some R in series with the L - I posted the wrong image! Sorry.

Keith.

 

[Kicchan]

Re: Delay line with Capture

Thank you a lot for your help, Keith! I'm having a look about some lumped t-line theory in order to model the L(R)C network you suggested.
Thank you again!!

 

[karesz]

Delay line with Capture

HAllo,
Are you sure pls that the values of 220mH & 220nF are the right for a delay of 50nsec?
K.

 

[keith1200rs]

Re: Delay line with Capture

HAllo,
Are you sure pls that the values of 220mH & 220nF are the right for a delay of 50nsec?
K.

No, they most certainly aren't! Around 2.2ms. However, I did say that in my post "... but not the values for a 50 ohm 30ns delay" - you must have missed that bit.

The equations are Z=sqrt(L/C) and the delay is sqrt(LC). These are normally values per unit length. So for a 30ns delay with 10 taps you would design for a 3ns delay per section.

Keith.

 

[Kicchan]

Re: Delay line with Capture

Thank you for your further explanation, Keith!