Noisy ECG calibration signal problem

I tried the following.

1. Crystal clock 2.048 MHz binary counted down 17 stages to get 16 Hz 5Vpp signal for potential divider.

2. Potentiometer 2M + 100R to divided down  5Vpp to (5 * 100 / 2 000 000) * 1 000 000 = 250 uVpp.

I found the divided down 250 uVpp 216 Hz signal has big spikes of order 50 mVpp, and stable period of 7.2 * 500 uS = 3.6 mS

Noise of 3.6 mS = 1 000 / 3.6 S == 27 Hz

I don't know the source of these spikes.  One possibility is the rippling of the binary counter.  An easy get around is to replace the crystal clock and binary counter by a simple NE555 based 16 Hz oscillator.

Another thing to try is to replace the noisy 1M resistors by metal film resistors.

.END