HP 204C losing AC voltage slowly. A problem?

[gooseEL34]

Folks.
Got a HP204C sine wave generator.   Set a 1kHz sine at 2.828 as measured with a Fluke 189 and have a small "leak" of AC.  Not an EE so I am not sure if this is normal.

When I watch the fluke, the mV shows 282.75 for example.  Then I watch it, with alligator clip leads (no movement of the probes) and the AC will slowly decrease by .01mV/S or so

I watched it go from 282.75 to 281.99 over a period of about 8 minutes or so.

Is this a bad component in the HP?  Is this normal?    Probably not critical for my applications as I usually just pass a 100mV sine into a guitar amp and watch the wave to set bias on the power tubes, but I am more curious than concerned.

Leaky cap in the HP204C?  Something with the amplitude pot on the front of the unit?   Something to do with heat?   The period of the wave stays rock solid 1.0000 kHz to 4 digits.

Thanks in advance for your thoughts
Tim

[Rerouter]

how long a warm up time did you give it? its not uncommon for test gear to change its output during a warm up cycle,

[amspire]

I think back when the HP204C was designed, HP would have been extremely happy with an amplitude change less then 1% over time. Minimizing distortion would have been a higher priority then stability. It will probably stabilize if you leave it running long enough, but back in the 60's, the common meters were moving coil and the sort of changes you are talking about are hardly noticeable on a moving coil meter. I think you are suffering from a condition "Fluke189ivitis" which is worry triggered by the extra resolution of your excellent multimeter.

Lucky you do not have an 8 digit meter - the HP204C would look even worse then.

If HP were using resistors with a 200ppm/C temp coefficient, then if a resistor warms up 10 degC, its value can change by 0.2% - that is the variation of a single component. I am trying to remember how they stabilize the amplitude of that oscillator - I think it is a zener and JFET circuit, but it may be a light bulb. Both will have a temperature coefficient. If it is a light bulb, then as the temperature in the HP204C case rises, then amplitude will fall.

Richard

[BravoV]

...... "Fluke189ivitis" which is worry triggered by the extra resolution of your excellent multimeter.

Quote worthy ... 

[gooseEL34]

OK.  Thanks guys.  I got the Fluke 189 today and it is my first high resolution meter.  I think I had not let the signal gen warm up at all when I noticed the issue.  Will let it run a bit and see if that slows the leak.....
Also... On the back of the 204C there is a set of banana females that according to the manual, can be linked with a piece that makes the internal ground an earth-ground.   There is nothing connecting the two currently.  Should I make that connection?
With the work I do on tube guitar amps, mostly everything I do is earth ground referenced. 
Thanks
Tim

[amspire]

You could use it either way, but you might find less rectified mains hum gets injected into the amplifier without the earth-ground link connected.

[retiredcaps]

You change the default setting on the 187/189 from 50000 count to 5000 count in the setup menu.