Using CD4017 to divide input for LM4017

Thread Starter

evo.motorsport

Joined Jun 20, 2007
21
Please help.

I'm dividing frequncy using decade counter and converting divided frequency to voltage. For some reason frequency to voltage converter does not work with the output signal from divider.

I'm attaching images with schematic, original fq signal and divided fq signal.

What am I doing wrong?
 

Attachments

Papabravo

Joined Feb 24, 2006
21,157
Capacitor U10 is probably preventing any meaninful DC signal from the CD4017 from making it to th LM2917. Did you consider that? The effect of the voltage divider and the capacitor might be to confuse the chip into thinking it is seeing a much higher frequency.

Your output circuit looks funny also. Vcc is pulled up to +12V, the collector is connected to +12 Volts and the emitter is open. What gives? You are ignoring the output circuit, and looking only at the charge pump output, is that correct? Meanwhile the inputs to the opamp/comparator are floating open so that amplifier should be beating up and down like a scalded duck. I'm not surprised you are having trouble - did you even read the datasheet?
 

Thread Starter

evo.motorsport

Joined Jun 20, 2007
21
Capacitor U10 is probably preventing any meaninful DC signal from the CD4017 from making it to th LM2917. Did you consider that? The effect of the voltage divider and the capacitor might be to confuse the chip into thinking it is seeing a much higher frequency.

Your output circuit looks funny also. Vcc is pulled up to +12V, the collector is connected to +12 Volts and the emitter is open. What gives? You are ignoring the output circuit, and looking only at the charge pump output, is that correct? Meanwhile the inputs to the opamp/comparator are floating open so that amplifier should be beating up and down like a scalded duck. I'm not surprised you are having trouble - did you even read the datasheet?
I'm using LM2917N and according to the datasheet opamp/comarator is completely disconnected from F/V functionality. I'm using comparator elsewhere. Yes I'm using only charge pump output, but only to measure output with digital voltmeter. Have you look on this image?
It's measured after U10 capacitor, directly on LM2917N PIN#1.

The output LM2917N is fine when I'm using undivided frequency:

The problem exists when I'm using divided friquency:
 

Papabravo

Joined Feb 24, 2006
21,157
When you say it doesnt work with the divided frequency what exactly does that mean? Is the chargepump output high, or low or somewhere in between?

If you cut the input frequency by a factor of 10 won't the output voltage go down by a corresponding factor of 10 without some change in the charge pump components. If you were getting 2V out before the divider you should get 200 mV when you use the divider.
 

Ron H

Joined Apr 14, 2005
7,063
Don't you still need to connect Vcc directly to +12V. Going through a resistor is suspicious.

When you say it doesnt work with the divided frequency what exactly does that mean? Is the chargepump output high, or low or somewhere in between?
LM2917 has a zener regulator. It needs a current limiting resistor. Didn't you even read the datasheet?:D
I think the problem is with capacitor U10. The time constant of that circuit is RC=20k*0.01uF=200usec. The low time for the signal is 0.9*15msec=13.5msec. The input hysteresis is on the order of 30mV. If you have more noise than that on your input signal, you will be seeing strange results. Change cap U10 to 10uF and it might work.
 

Papabravo

Joined Feb 24, 2006
21,157
LM2917 has a zener regulator. It needs a current limiting resistor. Didn't you even read the datasheet?:D
I think the problem is with capacitor U10. The time constant of that circuit is RC=20k*0.01uF=200usec. The low time for the signal is 0.9*15msec=13.5msec. The input hysteresis is on the order of 30mV. If you have more noise than that on your input signal, you will be seeing strange results. Change cap U10 to 10uF and it might work.
Yes I did eventually see the zener regulator, just not on the first pass. He said his scope traces were on the 2917 and they show no evidence of slow rise or fall. I was thinking that the series C - parallel R was acting like a differentiator and what should be on the input is an alternating series of positive and negative spikes. There is a great deal about this post and the pictures and the circuit diagram that are ahhh... to say the least, a bit murky.
 

JoeJester

Joined Apr 26, 2005
4,390
I also question the postings of those pictures. A series RC circuit would be both positive and negative triggers.

Even though the OP stated he followed the datasheet, I didn't see a 2917 with a series capacitor on the input.

My question to the OP would be ... when testing the F - V, did they include the input circuitry of that 10 k, C, and 10k resistors taking the input after the C?
 

Thread Starter

evo.motorsport

Joined Jun 20, 2007
21
Thanks to all.

All working now. LM2917N didn't understand the divided signal. You can see on the posted image decade counter output. I fixed the problem by replacing 2 resistors (U5 and U7) with 2 capacitors.
 
Top