Divide by 10 cct needed

[Gyro]

Yay, the saints be praised. Now that wasn't so hard was it? We now have something that we can review and comment on.

My comments:

It would be helpful to others (and yourself later) to include signal names as well as pin numbers so that we don't have to go digging out the datasheet.

You have placed the divide by 2 before the divide by 5. This will give 5MHz and 1MHz as you show ie. overall divide by 10.  If you check the datasheet, you will see that the output of the  divide by 5 (Q_D) does not have a equal mark:space. A divide by 2 always gives you a 50:50 mark space, so if you place the divide by 5 before the divide by 2, you will get a nice even square wave. Of course, configured this way you will get 2MHZ and 1MHz. If 5MHz is important to you, you can use the unused other half of the chip.

You probably want to experiment with a series resistor, or maybe series capacitor to AC couple the 1MHz output to the 2955A, I can't help you there as I don't know what it's expecting. That's one for experimenting to get most reliable operation.

Edit: @Uunoctium: Snap 

[ME]

Cheers chris , getting/ finding that schematic was a complete twat , luckily a chap from groups io helped, heres the datasheet, i only wanted a schematic as a starting point and that was the point of the post before it was an idea to riddicule me!

[Gyro]

Damn, I was hoping you'd worked it out for yourself from the datasheet and info provided in this thread, that was the idea - that you actually learned something. I could simply have pointed you to a GPSDO schematic that I posted here years ago   Sad that I was so wrong at this late stage.

The replies in this thread weren't intended to ridicule you, they were born out of frustration that you wouldn't put in any work yourself.

I'm glad your groups.io friend wasn't switched on enough to work out the 50:50 squarewave output. It shows the superiority of the knowledge base here. [Edit: Actually, they were]


P.S. https://groups.io/g/electronics101/topic/divide_a_freq_by_10/120437316 for context.

[SteveThackery]

The replies in this thread weren't intended to ridicule you, they were born out of frustration that you wouldn't put in any work yourself.

Bloody hell. It seems like the preceding 128 messages have been a waste of everyone's time. The OP didn't do even a lick of work, or embrace any of the extensive advice offered.

[free_electron]

This just screams raspberry pi , with a full blown linux and touch screen UI that shows an application to notify the user the division count is still set to 10. number is not alterable of course.
the rpi talks to an ESP32 wirelessly to maintain galvanic isolation, that controls an arduino with appropriate code that sets '10' as the compare value of a 74688 channel B , while channel A is attached to a chain of 7474 flipflops. when the count reaches the compare value the 688 resets the flipflop chain. use the CMP output of the 688 as the DIV output. you can additional logic to sync the output pulse to be exactly the length of 1 input pulse using an additional flipflop that keeps being set by the input and reset by the 688 output.
of course the software stack needs to make sure that the '10' from the rpi is properly passed to the esp32 , to the arduino , which sets it on the i/o pins.

Bonus points if you shove an FPGA in there to implement all the logic. Of course you need to model the FPGA code in Matlab , have it spit out VHDL and then synthesize that. Bonus points if you involve Claude to write some AI code running on the RPI.

looks good on the resume....

[floobydust]

The replies in this thread weren't intended to ridicule you, they were born out of frustration that you wouldn't put in any work yourself.

Bloody hell. It seems like the preceding 128 messages have been a waste of everyone's time. The OP didn't do even a lick of work, or embrace any of the extensive advice offered.

On another forum, OP's method also frustrated members and there it devolved into silence, no replies.
You can't change/improve/fix people here, they are as they are. Neuroplasticity is a myth.


The Marconi 2955 1MHz rear panel EXT STD input has a two-transistor preamp, into a 74LS90. Not sure why the -12V rail is used.
I would say it will be overloaded with a 5V square wave and would add a resistor divider down to say 1V. It autoswitches over to EXT REF when input amplitude is high enough.

[David Hess]

The Marconi 2955 1MHz rear panel EXT STD input has a two-transistor preamp, into a 74LS90. Not sure why the -12V rail is used.
I would say it will be overloaded with a 5V square wave and would add a resistor divider down to say 1V. It autoswitches over to EXT REF when input amplitude is high enough.

It uses -12 volts at the collector of the PNP follower simply because that is the only negative supply available.  -5 volts would have worked just as well.

I would have used a differential pair configured as a limiting amplifier, but the Marconi PNP follower is driving the common emitter NPN into saturation and cutoff, producing a TTL compatible output signal.