HP34401A frequency response not as I expected

[Lorenzo_1]

While doing some testing recently, I was surprised to find that my 34401A DMM was seemingly reading frequencies as high as 1.65MHz reliably(?) despite the manual specifying a maximum frequency range of 300kHz.  Here's a sampling of readings across the range I tested - input from an HP3325A that's out of cal but generates stable and pretty accurate frequencies. The DMM hits a brick wall at 1.65MHz but reads stable frequencies very reproducibly for inputs of at least 0.05-5V amplitude and for sine and square waves. DMM was set to 6-digit high speed. 

Is this surprising or unexpected?  I thought there'd be some over-range capability but this seems a bit over the top. Are there some gotchas hidden away here somewhere? I didn't find anything on this question elsewhere on the web.
 
Input (MHz)   Reading
0.1                  99.99978k
0.3                  299.9993k
0.5                  499.9989k
1.0                  999.9978k
1.2                  1.199997
1.4                  1.399997
1.6                  1.599996
1.65                1.649996
1.7                  1.5
1.8                  1.2
2.0                  1.0

[Kleinstein]

The limit for the frequency counter function could depend on the software version. So maybe they have not updated the specs after an improvement there.

Another point is that the sensitivity may go down considerably with higher frequency - so they might need an higher amplitude. Form the the low frequency values it looks like they use a kind of reciprocal counting. This could mean the resolution could go down with higher frequencies. Depending on the implementation there could also be a chance for rare errors with frequencies close to the limit, e.g. if something happens on GPIB.

[HighVoltage]

I have tested this before and saw a brick wall at around 700 kHz.

Are you using a sin wave output of your generator?

[Lorenzo_1]

Thanks for the feedback. I tried testing frequency response down to 1mV (lowest amplitude my 3325A will output) and the DMM readout is stable and right on the button from 3Hz to 1.6MHz.  And it's the same regardless of whether it's looking at a sine or square wave.  Not sure that the result is of any real practical value given uncertainty about how/when you can rely on the readout.   

[K1JOS]

I have a 34410A and it reads frequency at 2 mVpp pretty accurately up to 800 KHz

jerry

[awallin]

On my 34401A under SYSTEM > REVISION I get "11-05-02"
and with a 100 mV sine-wave from a Siglent SDG2042X it reads OK up to about 1.35 MHz
with other amplitudes the limit seems to vary either a bit higher or lower around 1 MHz.

[DavidAzulay5000]

i test  my hp 34401a
With the help of the HP 3325B.
The goal was to check the accuracy of my 3325 in ACV
And I happened to try high frequencies and the 34401B
It was accurate down to the last hertz.

And to my surprise it is a frequency measurement
From 0.5 Hz to 2.63 GHz
I was really shocked I thought the spec was 300kHz
The measurements were accurate
after 2.63 giga hertz is no longer a meter. The high  frequency measurements were accurate at voltages from 1 up to 4 volts RMS.
Apparently the specification is a frequency without attenuation in the high voltages