Tektronix TDS3000 "Low resolution" warning ????

[darkstar49]

Hi,

does anyone know why a TDS3054 shows a 'Low resolution' warning from around 315MHz onwards for the frequency measurement ?

10K points at 5GS/s, that should be more than enough for a frequency measurement up to 500MHz, no ?
I'm using a sine wave, 500mV pk-pk, 50 ohm, real-time sampling

I first thought my siggen was at fault (maybe it was, but it died in the meantime...), but now I've got a calibrated R&S SML, so that won't be the issue... then I thought "well, it's because I hacked it from 100 to 500MHz", but I just checked with a 'native' TDS3054, and it's exactly the same .... 

I'm a software guy, so sorry if I'm missing the basics here...  ;-)

[Jwalling]

I don't know the why, but it is normal for these scopes

[David Hess]

I do not know what threshold Tektronix uses but one cycle at 315 MHz is only 16 sample points at 5 GS/s.  Frequency is the reciprocal of the time between the 1st and 3rd mesial measurements.

[darkstar49]

I do not know what threshold Tektronix uses but one cycle at 315 MHz is only 16 sample points at 5 GS/s.  Frequency is the reciprocal of the time between the 1st and 3rd mesial measurements.

Could you elaborate a bit ?? Are you saying the 5GS/s are the problem ? A DPO4054 with half that sampling rate doesn’t have that problem...

[MarkL]

The TDS3000 user manual says the frequency is computed as the "Reciprocal of the period of the first cycle in the waveform."  So, even if you have the measurement set to the whole capture record, it apparently doesn't use it.

If you put the scope in dots mode and then do a single shot capture, perhaps it will better illustrate the warning.  There just aren't many data points for computing an accurate frequency in one cycle.

If you want a more accurate and stable frequency measurement, you could turn on measurement statistics and look at the mean frequency over a number of captures.  Or, turning on waveform averaging will do the same thing.  But neither will get rid of the warning.

[David Hess]

I do not know what threshold Tektronix uses but one cycle at 315 MHz is only 16 sample points at 5 GS/s.  Frequency is the reciprocal of the time between the 1st and 3rd mesial measurements.

Could you elaborate a bit ?? Are you saying the 5GS/s are the problem ? A DPO4054 with half that sampling rate doesn’t have that problem...

Is that real time sample rate or equivalent time sampling rate which is probably at least 50 GS/s on the TDS3054 at 1ns/div?  The DSO generates the low resolution warning based on the number of available sample points in the processing record.  This should only be a problem on the TDS3054 when making single shot captures where the same rate really is limited to 5 GS/s.