RTC with 1 Hz output as default (no serial communication wanted)

[eV1Te]

I am currently designing an analog high impedance amplifier circuit, where i need to generate a built-in calibration signal that goes through all the amps stages to verify that the gains etc. are correctly set.

The calibration signal will be connected to the input of the main electrometer-type amplifier (LMP7721) via a high-insulation-resistance reed-relay with guarded internal screen.
For this application it makes sense to have a calibration square-wave of very small amplitude and a low frequency of ca 1 Hz. The amplitude will be handled separately by an additional op-amp stage.

Now to my question:  How to generate a reasonably accurate 1 Hz square wave?
1. Crystal with RTC IC is probably best... But without a micro I need a RTC chip that has 1 Hz output enabled by default. Is there any small (8-pin) RTC ICs that do that? In all chips I have seen its disabled by default, also enable/reset pin is desirable so it can be turned off when not used, for reduced noise on power-rails but it's probably not required.

2. Schmitt trigger oscillator or 555 timer comes to mind, but for 1 Hz you need large caps, you might not get 50 % duty cycle, and it might inaccurate/drift which could be confused by error or low quality by the end user.

3. A micro with crystal (or built in RC) would also work, but the extra complexity of having to write software and add programming headers to the PCB is not wanted.

4. Are there better ways?


Thanks

[schmitt trigger]

Probably your best bet would be to hack an “analog” clock controlled by a 32768 Hz tuning fork crystal. These devices output a bi-phase 1 Hz signal, and work from a single 1.5 volt battery.

If you can find the raw IC, the part number is MC14450, although it may be obsolete.

[blackdog]

Hi eV1Te,

Use a cheap Arduino klok module with a DS3231 IC on it.
Program one of the outputs to 1Hz, and i think one of the outputs is standard 1Hz, its a open source output.


https://www.aliexpress.com/item/DS3231-AT24C32-IIC-High-Precision-RTC-Module-Clock-Timer-Memory-Module/2037934408.html
Ofther setting the klok, you dont need the arduino any more.

Kind regarts,
Bram

[SilverSolder]

Perhaps go for a tiny micro...  and find some more work for it to do,  in order to justify its presence?  -  it could manage the whole test, control relays, etc., and make the whole calibration cycle more precisely repeatable.

[Ian.M]

A 74HC4060 + a 32.768KHz watch crystal will give you a 2Hz output.  Either redefine your spec for a 2Hz cal signal, or add another /2 stage (one D flipflop) after it.

However, I'd go with the small MCU.   Many have internal oscillators of +/-2% or better accuracy, which is likely to be superior to the 555.  Use  a PIC or AVR and after you've finalised your code, you can order them pre-programmed from MicrochipDirect^*, so can treat them as a custom chip, no production line programming required, so no ICSP header either.  For prototyping, patch on fly-wires to a loose ICSP header.   

Another option would be Analog Devices TimerBlox.  A LTC6995-1 or -2 + three resistors to configure it to directly produce a 1Hz output.

MicrochipDirect's price per programmed chip is quite reasonable, however the setup fee isn't worth it for very small runs, and as it applies to every code change, your code needs to be 100% complete, tested and frozen.

[Quarlo Klobrigney]

That's the way I do it. Cheap and easy. You could even use a radio type for $10 American.

Quote from: schmitt trigger on Today at 18:15:31

Probably your best bet would be to hack an “analog” clock controlled by a 32768 Hz tuning fork crystal. These devices output a bi-phase 1 Hz signal, and work from a single 1.5 volt battery.

[Datman]

Use an ATtiny with an ICSP USBASP programmer. The software will be very simple. If you start using microcontrollers, they will be very useful in your future works.

[soldar]

How to generate a reasonably accurate 1 Hz square wave?

If you can hack a clock or watch that seems pretty straightforward.

A xtal oscillator does not even need an IC as you can design an oscillator with just a transistor and then divide with counters. I have done this a few times, even with odd frequencies.

If you have the right xtal then the IC is the 4060 14-stage Counter Divider Oscillator CMOS IC.

I do not think any micro is needed at all.

A 555 is doable depending on the precision you need. If you need high precision then not good. 

Another possibility is a mains frequency divider.

[2N3055]

Just use CD4060 with RC oscillator. It is an amplitude calibrator, frequency within 5% is good enough....

[eV1Te]

Thanks for all the ideas.
I have tried to summarize the different methods here below and some reasons why they will or will not work in my case.

• Since no one knew about a RTC chip with the 1 Hz output enabled by default, I assume that all require serial comms to enable it.
  

• Using a micro with a crystal is a good option that I have considered. But the volumes for this project is not very large hence it would be costly to go for pre-programmed directly from the vendor. But too complex/cumbersome to do manual programming of every chip before soldering (need to build a jig of some sort) and I do not have space to add programming headers/connectors to each board (custom programming jig could solve it, but that becomes to complex).

• Hacking some old clock IC does not work as the source of components is not reliable, (most are obsolete today).

• Using some existing module or separate PCB is too costly and requires more space and extra mounts etc.

• Hence I will probably fall back to using a schmitt trigger (or a NAND gate to have one input as an enable pin) oscillator. The 555 also works, but has more pins and require more components.
  
• - The single NAND SN74AHC1G00 seems to be a good choice:
     Symmetric input levels for close to 50 % duty cycle and low-leakage which enables me to use large resistors (instead of a large capacitor)

[KMoffett]

You didn't say if this is a one-off or a design for a device to go into production. If a one off, attached is a circuit I've used for a 1 second square wave. It used the guts from a 1.5VCD desk or wall clock for the 1 second pulse, and a 555 monostable to get the 50% square wave.

Ken

[Ian.M]

• Hence I will probably fall back to using a schmitt trigger (or a NAND gate to have one input as an enable pin) oscillator. The 555 also works, but has more pins and require more components.
  
• - The single NAND SN74AHC1G00 seems to be a good choice:
     Symmetric input levels for close to 50 % duty cycle and low-leakage which enables me to use large resistors (instead of a large capacitor)

That's very much not a good idea.  The frequency of a single gate Schmitt inverter oscillator is highly dependent on the width of its hysteresis band, which isn't specified for the SN74AHC1G00, and you will certainly violate its max recommended input transition time (100ns @3.3V).  If you use a dual NAND you can implement a two gate oscillator with predictable period, but that needs three passives (see Fairchild AN-118), and you are back up to an 8 pin package. 

You could use a CMOS 555, with Threshold and Trigger tied together (which gives you an inverter with a guaranteed 1/3 Vcc hysterisis band), and to the timing capacitor, and a single feedback resistor from Out to the Threshold + Trigger + TC node.

However that still leaves you needing a fairly large, extremely low leakage timing cap, and a high value feedback resistor, so it will be highly sensitive to board surface contamination and enviromental conditions.

The best option if you want to avoid programmable chips, is an higher frequency oscillator + a divider chain.  74HC4060 if you can afford the board space, or bite the bullet and pay the premium for the LTC6995 timerblox chip if you need something in a smaller package.

[voltsandjolts]

Some EPSON parts will output 1Hz just by setting pins.

e.g. RTC-4543SA (built-in crystal, no external parts required..well, decoupling cap I guess)
https://www.digikey.co.uk/product-detail/en/epson/RTC-4543SA-A0-ROHS/SER3245TR-ND/1021877
FOUT freq set by FSEL pin as 1Hz or 32768Hz, also tie FOE enable pin appropriately.

BTW The above IC is spec'd at 3uA current consumption with FOUT disabled but IIRC this increases significantly with FOUT enabled, like 100uA or something.

[Yansi]

Why complicating stuff so much? Do you even need RTC at all?

Use 4060 + crystal as your 1Hz source.

//EDIT: Forgot 4060 will give you 2Hz, so use 4521 instead

[NivagSwerdna]

Now to my question:  How to generate a reasonably accurate 1 Hz square wave?

Define 'reasonably accurate' and the solution will follow.