The A15 power supply

I had setup a measurement run where I used my HP6034A power-supply instead of the battery-backed lab 24VDC system.

For reasons not yet figured out, the 5065A was quite unstable, and I didn’t have time to figure it out before other stuff would keep me occupied for a week.

So I decided to do a quick test for voltage sensitivity by doing one hour runs with different EXT DC voltages.

The first thing to notice is that the HP5065A has mostly linear regulators, which means that the current stays almost constant as the voltage increases:

../../../_images/plot12.svg

At 23V total power consumption is 30W, at 30V it increases to 39W.

(Note that this is without the 1 MHz & 100 kHz dividers, and without the clock and associated 1 Hz divider and PSU)

The voltages are a few hundred millivolts lower than the PSU voltage because it is measured after the anti-reversal diode.

With greater power comes greater temperature:

../../../_images/plot21.svg

This is the temperature measured at the A15CR5 Zener diode which is (normally) the voltage reference for both the +20V supply and the C-field current. In this case the C-field is driven by a Fluke 732A through a Vishay Z201 resistor.

The temperature does affect the Zener:

../../../_images/plot3.svg

Tempco is roughly 22 PPM/K.

According to my math, that would result in a frequency shift of 8e-14/K, assuming there were no other tempcos in the C-field driver current.

And then we get to the fun part:

The +20V regulation circuit on A15 does not like high EXT DC voltages:

../../../_images/plot4.svg

Above roughly +28V EXTDC, regulation suffers. Not that it’s too great in the first place as can be seen from the stair-steps.

Here is a plot against temperature:

../../../_images/plot5.svg

The non-retrace indicates that it is the EXT DC voltage, rather than the temperature which causes loss of regulation.

Running the, or at least this, HP5065A from a 24V lead-acid battery is clearly a bad idea, as the float voltage is 27.6V and the bulk charge voltage around 29V.

Putting an isolating DC/DC converter in the HP5065 as I talked about in the a previous entry is clearly a good idea.

So, back to the question I was trying to answer: With a stable C-field current, is the HP5065 sensitive to EXT DC voltage:

../../../_images/plot6.svg

Probably.

But there are other sources of instability that must be tamed first.

phk