Post Go back to editing

TOSC CALCULATION of LT124x

See this calculation given in LT124x data sheet

Say, I need 10.5 use as Tosc. With max d as .9

So Td= .1* Tosc= 1.05 usec

We can also know RC from the charts. At 100kHz, R=30k, C=500pF, RC=15usec, Tr=8.75 usec... approx Td left is 1.25 usec...fair enough.

But,  from  Td to RC relationship,  one gets RC=145.74 usec 

This  is off by factor of 10!

If it is 14.575 , one is closer to reality.

Also, Dead time charts for LT1243  start from 50% yet the output frequency is equal  to oscillator freq unlike '42 whose output is half.

It is a bit confusing.

Appreciate some clarification.

-b

So what is the correct relationship?

I KNOW if R=9k, C=1000pf, it runs approx <200kHZ( we have a system running with this value)

thnx for clarification.

Parents
  • Ok, if you need 95.2KHz and 10% dead time (Tosc=10.5us, tr=9.45us, td=1.05us), I would suggest using the equations. Given tr=9.45us and solving tr = 0.583*RC for RC yields RC=16.2us. Next, given td=1.05us and RC=16.2us, solving td = 3.26* RC /( (0.0164)*R -11.73 ) for R yields R = 4K and therefore RC = 16.2K / 4K = 4.05nF. You can then modify LTSpice's LT1243 "jig" schematic and see that the results are close (file/New Schematic, hit F2, type LT1243, hit “Open this macromodel’s test fixture” button and change R and C values, delete main MOSFET and hit the running man symbol).

    Regarding RC=145.74usec, using R=25K, td=1.25us and solving td = 3.26* RC /( (0.0164)*R -11.73 ) for RC, I also get RC = 145.74us (tr = 0.583*RC = 85us, C = 5.8nF, Tosc = 86us or 11.6kHz which is a valid point on the graph).

    Comment: There are three charts on top of page 7 of the LT1241 Data Sheet. The first one SHOULD be labeled Output Deadtime vs Oscillator Frequency - LT1242, LT1243" (not LT1244). The second one should be labeled Output Deadtime vs Oscillator Frequency - LT1242, LT1244, LT1245" (not LT1243). The third Graph is labeled correctly "Timing Resistor vs Oscillator Frequency".

    Kurk

Reply
  • Ok, if you need 95.2KHz and 10% dead time (Tosc=10.5us, tr=9.45us, td=1.05us), I would suggest using the equations. Given tr=9.45us and solving tr = 0.583*RC for RC yields RC=16.2us. Next, given td=1.05us and RC=16.2us, solving td = 3.26* RC /( (0.0164)*R -11.73 ) for R yields R = 4K and therefore RC = 16.2K / 4K = 4.05nF. You can then modify LTSpice's LT1243 "jig" schematic and see that the results are close (file/New Schematic, hit F2, type LT1243, hit “Open this macromodel’s test fixture” button and change R and C values, delete main MOSFET and hit the running man symbol).

    Regarding RC=145.74usec, using R=25K, td=1.25us and solving td = 3.26* RC /( (0.0164)*R -11.73 ) for RC, I also get RC = 145.74us (tr = 0.583*RC = 85us, C = 5.8nF, Tosc = 86us or 11.6kHz which is a valid point on the graph).

    Comment: There are three charts on top of page 7 of the LT1241 Data Sheet. The first one SHOULD be labeled Output Deadtime vs Oscillator Frequency - LT1242, LT1243" (not LT1244). The second one should be labeled Output Deadtime vs Oscillator Frequency - LT1242, LT1244, LT1245" (not LT1243). The third Graph is labeled correctly "Timing Resistor vs Oscillator Frequency".

    Kurk

Children