• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 49 subscribers
  • Views 690 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

THS4551: THS4551 model

alpesh kavia1
Prodigy 210 points
Part Number: THS4551
Other Parts Discussed in Thread: ADS127L01

Hello,

I need a spice model for the THS4551.

The model was updated in July2019. When I try to download the model a new web page is opened with the sub circuit model but there are no carriage-returns so it is not usable.

Can you provide a zip file of the ths4551 model which I can use in LTSpice.

  • 0
    Guru 48395 points

    Well I think you want a .lib file, try this, 

    2703.THS4551.lib

  • 0 in reply to Michael Steffes
    Guru 48395 points

    But you are are right Alpesh, the July 2019 model update is not reachable - I have passed that back into the right people. What I sent you was a .lib file of the original model - you probably do not want that if there is an update. Hopefully, the model listing details what changed - usually not. 

  • 0 in reply to Michael Steffes
    Guru 48395 points

    Actually Alpesh, I was able to get imported - steps

    1. right click on downloaded .tsm file and view in directory 

    2. copy that file to where you want to keep THS4551 things

    3. Go to TINA and create a new circuit

    4. go to insert, and down lower this a macro option

    5. click that and then select the file you just saved

    6. it will open the .TSC file

    7. which you can save for future use

    This is a complete update to the original model using template being widely applied to model updates - not sure it has been applied to FDA's yet, but hopefully it works - we spent a lot of time on the original model making sure it would expose phase margin issues - I have some very sensitive circuits, might be interesting to test the new model to make sure it shows the issues those showed - and they were bench validated. 

    here is the new .tsc file, 

    THS4551_2019.TSC

  • 0 in reply to Michael Steffes
    Guru 48395 points

    So I had actually exercised Figure 78 in the data sheet for this article - I had discovered the original circuit was unstable and stepped through some changes in the latter part in this article to improve phase margin. This is the LG simulation using the original model with some improvements to get about 21deg phase margin, I got more in the next two steps. 

    https://www.planetanalog.com/author.asp?section_id=3404&doc_id=565122

    The new model reverses the supplies and gets rid of the shutdown (that is common, ADI does not supply models with the shutdown included either - many of the High speed amps models try to, but I don't think many folks use it). Making those changes in this same circuit increases the phase margin to 30deg?? I wonder about that. The pieces that went into the original model that were critical to this were the unloaded Aol gain and phase and then detailed open loop output impedance model - one or both changed here. 

  • 0 in reply to Michael Steffes
    Guru 48395 points

    I need to correct some things - 

    Figure 78 came from the ADS127L01 datasheet.With the 1nF across the FDA inputs it is very stable - about 63deg phase margin. I had taken that off to reduce output noise peaking where then the phase margin goes to 20deg with the original model. What i was trying to do in the article was improve the SNR while also keeping >50deg phase margin - here is the original FIgure 78 SNR using a 5Vpp full scale assumption or 1.77Vrms for SNR, about 99.7dB SNR

    Here is the modified circuit from the article Figure 12, improved SNR by 4dB with 55deg phase margin, 

  • 0 in reply to Michael Steffes
    Guru 48395 points

    Now back on the question of model changes, 

    one of the subtle things is the SSBW for the CM loop, a good test of that is the active match circuit - here that is for a 50ohm input match where just looking at input impedance which should be 28dBohms or 25ohms in this test, The original model did real well here, 

    Dropping in the 2019 update, looks like the CM loop SSBW has been reduced a lot, but it is there, just maybe too slow, 

  • 0 in reply to Michael Steffes
    TI__Mastermind 19316 points

    Hi Michael,

    Thank you for your thorough analysis. What is the CM loop bandwidth that you are expecting?

    Sean

  • 0 in reply to Sean Cashin
    Guru 48395 points

    hey Sean, 

    well it is a subtle effect because this is the internal loop bandwidth not the bw from the control pin. I am not sure I ever heard what that was on the physical THS4551, but more in the 100MHz region I think. The way I see it is that active match sim where you could increase the new model bandwidth until the 10MHz input impedance is more like the 28.5dBohm of the original model.