Hi everyone,
I had a circuit in which there is a crystal oscillator generating a frequency of 1Mhz.The frequency is then divided by 16
and then by 8 using 74ls93 ripple counters.The output frequuency is approximately 8kHZ.
Uptil now,everything is fine.But now the frequencyof 8kHz is given to monostable multivibrator 74ls122.
The output frequency of this multivibrator is also 8kHz,which is then given to other ic's
The difficulty i am facing is that what is the use of monostable multivibrator,if you can directly have the 8kHz frequency
from the output of 74ls93 ripple counter.The output of 74la122 is given to 74ls165 which is used as parallel to serial
converter.
Thanks in advance for any help
Hi everyone,
I had a circuit in which there is a crystal oscillator generating a frequency of 1Mhz.The frequency is then divided by 16
and then by 8 using 74ls93 ripple counters.The output frequuency is approximately 8kHZ.
Uptil now,everything is fine.But now the frequencyof 8kHz is given to monostable multivibrator 74ls122.
The output frequency of this multivibrator is also 8kHz,which is then given to other ic's
The difficulty i am facing is that what is the use of monostable multivibrator,if you can directly have the 8kHz frequency
from the output of 74ls93 ripple counter.The output of 74la122 is given to 74ls165 which is used as parallel to serial
converter.
Thanks in advance for any help
Hi Eric and Hero,
I've attached the the circuit diagram
Can u plz explain the possible reason for using the 74122 IC.
I've also attached a datasheet of 74165 for your convinience
Hi Eric and Hero,
I've attached the the circuit diagram
Can u plz explain the possible reason for using the 74122 IC.
I've also attached a datasheet of 74165 for your convinience
hi,
On fig#4, the LS164 is a serial input, parallel output shift register.
The digitised voice is the serial input and the parallel output is connected to the LS374 an 8 bit Latch..
The latch is strobed by the 2000nS pulse and so the output of the latch is steady during the DAC conversion period.
As I said earlier the short pulse from the LS122 is required to strobe the LS374 data latch which inputs to the DAC.
Hi Eric,
THanks for giving me your time,but i still have some difficulty regardilng the circuit.
Why is it neceesary to give pulse of width 2micro-sec.Will it not do if i connect 74165 and 74374 directly from the output of second 7493 ripple conter
As Eric has said, the Ls122 is setting up a pulse width to load the shift register with 8 bits. The Load Shift/load pin on the LS165 is level sensitive as opposed to edge triggered as in a D flipflop. This is first weakness in the design as this make the system asynchronous. Second bad choice was using a monostable in a digital circuit where timing is important. The reason is that the analog portions of the LS122 will drift over temperature and timing will shift possibly introducing timing errors. I am sure this works fine for many applications, but in general this sort of design is bad practice.
I think it is good that you are questioning such things.
Hi Eric,
THanks for giving me your time,but i still have some difficulty regardilng the circuit.
Why is it neceesary to give pulse of width 2micro-sec.Will it not do if i connect 74165 and 74374 directly from the output of second 7493 ripple conter
hi,
The monostable 2uSec latches a 'fast, short, slice' of the audio waveform, you dont want the input to the ADC varying while conversion is taking place.
So dont connect the 7493 directly to the latch else you will get severe distortion.
Whats the problem in building the circuit as designed using a LS122.
Because your loading digital bits into the register based on the level output of the 122, over temp, the level timing will shift and possibly lose a bit. With only 8 bit data, 1 bit error will be noticeable. Aside from the analog front end, I was always taught to leave analog out of the digital domain.
Because your loading digital bits into the register based on the level output of the 122, over temp, the level timing will shift and possibly lose a bit. With only 8 bit data, 1 bit error will be noticeable. Aside from the analog front end, I was always taught to leave analog out of the digital domain.
The issue I see is shown in my cheesy timing diagram.
The 122 part is temp and voltage dependent which makes it less reliable in a digital circuit. Of course I have not done the timing setup and holds, but drifty circuits as in the 122 can pose a problem. As I said, I am sure it is fine but I avoid such circuits as a general design rule.
The issue I see is shown in my cheesy timing diagram.
The 122 part is temp and voltage dependent which makes it less reliable in a digital circuit. Of course I have not done the timing setup and holds, but drifty circuits as in the 122 can pose a problem. As I said, I am sure it is fine but I avoid such circuits as a general design rule.
Alas we are at an impass as I have failed to make my point. I would say this is a design philosophy issue and I doubt I can turn your thinking to my side.
As one can never precisely predict performance of an analog part from batch to batch, over temp and over voltage, there may be pitfalls. I did find a supporting author to my point.
Hi Eric,mikebits
I got some idea regarding use of 74122.Your discussion gave me some insight into the circuit.
Now there is a little problem.AD7574 is not available near my place,i just wanted to know that will ADC0808 be compatible with rest of the circuit.I've never used ADC's before,thats why i need some help.
I'll be really thankful for any help
The Adc0808 is a 8 channel part, meaning it accepts 8 analog inputs depending on the address. I suppose you could tie the address lines to channel 1 and set ALE to logic 1 at power up. You should compare the data sheets, I only looked briefly.