LMC555: LMC555 SNVA703 Application Report question

Hi Sir,

Good morning,
Sorry, we have diffecult to measure 220pF capacitor.
Because our probe have about 100-120PF parasitic capacitance, so we can not do it.
Do you have other suggest for us?

Marco

Hi Sir,

Sorry, we have diffecult to measure 220pF capacitor.
Because our probe have about 100-120PF parasitic capacitance, so we can not do it.
Do you have other suggest for us?

AJ,

A capacitor can be measured using this simple circuit.

Switch can be a switch or a jumper wire. Short cap with switch. Remove short and measure time from 0V to 63.21% of Vtest; Calculate C1 as Time1 / Rtest.

Do this test again with the capacitor disconnected from the circuit. Calculate C2 as Time2 / Rtest.

The capacitance of the capacitor is C1 - C2

Feel free to change supply voltage and Rtest 

Hi Sir,

We use capacitor meter and measure value is 249PF.

Thanks

Marco

Marco,

Based on the measurements in this thread, 3.6k is needed to make 220pF run at 600 Khz.
Accuracy will be a function of the capacitor tolerance and to a lesser extent LMC555 propagation delay time.

Marco,

I built the circuit in the lab, R = 3.65k 1% actual value 3.6596k. C= 220pF, actual value 239pF.

Output was 585 kHz into a roughly 40pF 10M load.

The output high on resistance is greater than the output low resistance,  therefore some ramp voltage is seen on output VOH time. Be sure not to heavily load the output as that may change the output frequency.

The output high on resistance is greater than the output low resistance,  therefore some ramp voltage is seen on output VOH time. Be sure not to heavily load the output as that may change the output frequency.

=> After LMC555, we connect SN74LVC245APWR and I looking for sepc the load are 40PF 500ohm. So our device is not serious load. Does is?

Thanks.

Marco,

That is not the correct table.  Use this table

This typical capacitance is per input. 8 inputs is 44pF plus 8 leakages.

8 leakages could be as 160uA but will likely be significantly lower.

It is not mandatory but it will be better to buffer the LMC555 output before going to SN75LVC245A

Hi Sir,

I did the simple experiment, remove the SN75LVC245A and test the LMC555 output.

But I get frequency is 429.4KHz.

R=5.49K

C=220PF

I think the circuitis is simple, output should be 600KHz.

Does it correct?

Thanks.

Marco,

The circuit is simple but reality is complex.

Even this formula (using pin 7) is not exact.

For pin 3 usage, replace (Ra) with output high (Ron). Replace RB with R

Alternatively uses 3.65k and 220pF like I did.

Hi Michallick,

Good morning, thanks for your explain and provide to us new value 3.65k and 220pF to meet about 600KHz.

Then I have last questions.

(1) If I want to design other frequency like 450KHz, I should refer to what to use LMC555? Because we have other project should use this, so we could not let you provide to us every time.

(2) If our duty is 50%, then I just can refer one formula, like tc(H) or Tc(L)? Does it correct? Do you have more formula parameter information?

Thanks.

Marco

Marco,

This formula is a much better starting point to get (R) and (C) values from a desired frequency (F)

The TC(l,h) formula unfortunately require LMC555 parameters that are not specified in the data sheet.

 

For a normal Ra Rb and C setup (using pin 7) the formula becomes.

The difference is due to propagation delay difference between pin 7 and pin 3. 

Hi Michallick,

Thanks for your reply.

According to value 3.65k and 220pF to meet about 600KHz.

But I use formula as below, I can not get F=600KHz when R=3.65K and C=220PF.

Could you help us to explain? Thanks :)

Marco

Marco,

Don't forget the "1+"

I suggested a slight smaller resistance (2650 ohms) because most capacitors tend to run high (in my experience) and any load on output will slow frequency a little. 

Hi Michallick,

Sorry, I really forget it.

I try experiment and change resistor.

1. R=3.65K C=220PF, then frequency is 638KHz

2. R=3.3K C=220PF, then frequency is 690KHz

3. R=2.87K C=220PF, then frequency is 766KHz

According to experiment, if resistor less more, the frequency be higher.

So if I want to design other frequency, I have difficult to  choose R and C value to meet frequency.

Thanks.

Hi Michallick,
Sorry, update my experiment result.

This experiment that I do not remove buffer(U63).
1. R=3.65K C=220PF, then frequency is 638KHz
2. R=3.3K C=220PF, then frequency is 690KHz
3. R=2.87K C=220PF, then frequency is 766KHz

This experiment that I remove buffer(U63).
5.49K = 436KHz
3.65K = 595KHz
2.87K = 707KHz

It is look like we should remove buffer to check.
before our experiment, you provide to me R=3.65k and C=220PF that is like correct.
But can I understand how do you get it value?
Because the above formula can not meet final values.
Thanks

Marco,

My results show the the frequency decreases with more output capacitance.

Oddly, it seems that you saw 638 kHz with buffer (cap load) and 595k without buffer (less cap load)
Just putting a oscilloscope probe on the output will slow it down a little.

How accurate does your 600kHz need to be?

Hi Michallick,

It is not accurate.

We can accept about 30% tolerance.

It is for PWM power IC frequency used.

Thanks.

Marco

Marco,

LMC555 can easily support 30% tolerance.

Hi Michallick,

Good day.

Thanks for your reply.

But you have not give us an answer.

How do you know R=3.65k C=220pF then frequency=600KHz?

This information is importance for us.

Marco

Marco,

The frequency is a function of the R, C and the load on the output. All the results, I have posted have been lab experiments using LMC555 samples. Because loads are different in each application some fine tuning may be needed to get final values.

Hi Michallick,

Thanks for your reply.

Because I find out frequency is different tolerance for per piece.

Our application is when I have no external frequency, I should need a frequency to device.

When external frequency input to IC, output frequency should need synchronize refer input frequency.

If we don't want to use LMC555, do you have suggested product for our application?

thanks.

Marco

Hi Michallick,

Our power supply IC can accept the tolerance is 5%.
So I want to fix our tolerance as 5%.
Thanks a lot.

Marco

Marco,

A free running clock that syncs to an optional input clock is a phase locked loop (PLL).
www.ti.com/.../products.html
I don't know which one that would have a center frequency with 5%, I suggest starting with CD74HC297

Hi  Michallick,

Sorry late to reply.

Do you know what does fequency tolerance of LMC555?

Our customer want us giving an answer for that.

So I need your help.

Marco

Marco,

This is the only LMC555 accuracy specified in the data sheet.