Picoscope 3404D MSO strange vertical step size

Post any questions you may have about our current range of oscilloscopes
Post Reply
kc55
Newbie
Posts: 0
Joined: Sun Dec 06, 2015 10:52 am

Picoscope 3404D MSO strange vertical step size

Post by kc55 »

Hi, I just got my new scope but I observed a strange behavior. Someone please help me on this. I need to know if the unit I just received is faulty.
I applied a square wave at 1 kHz 4 V p-p (from the unit's function generator output) to channel A. All other channels are off. Channel setup: 10 V full scale, 200 us/div, number of samples at 1MS, no resolution enhancement, probe and software both at 1X.
I zoomed in to take a look at the noise. Due to the scope's 8-bit vertical resolution, the step size is (10/(256-1))*1000 = 39.2157 mV. But, when I measured the actual step size (it looks like a step to me), it was 47.4 mV, about 20.9% more than the expected step size How can this be possible?
I own an old (discontinued model) Picoscope 3204. With the exact setup, I get 39,64 mV, very close to the expected value.
Attachments
3404D MSO
3404D MSO
3204
3204

Martyn
Site Admin
Site Admin
Posts: 4491
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: Picoscope 3404D MSO strange vertical step size

Post by Martyn »

Can you check what happens when you have all four channels enabled.
Martyn
Technical Support Manager

kc55
Newbie
Posts: 0
Joined: Sun Dec 06, 2015 10:52 am

Re: Picoscope 3404D MSO strange vertical step size

Post by kc55 »

It's 39.83 mV with all channels switched on, with nothing connected to the other channels.
I tried some combinations and found that I need to switch on at least two other channels to get this value.
With only one other channel switched on, I get approximately 44.7 mV.
Hmmm, my older 3204 scope does not exhibit this behavior.
Attachments
3404D MSO - all channels switched on
3404D MSO - all channels switched on

Martyn
Site Admin
Site Admin
Posts: 4491
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: Picoscope 3404D MSO strange vertical step size

Post by Martyn »

That is the expected behaviour, it is due to the use of digital gain, applied to the ADC values to ensure that the readings in the software are within the published specifications for the unit. Digital gain was not present on the older units.

The reason it varies depending on number of active channels is due to the setup of the ADC chip cores and the loading this creates.
Martyn
Technical Support Manager

kc55
Newbie
Posts: 0
Joined: Sun Dec 06, 2015 10:52 am

Re: Picoscope 3404D MSO strange vertical step size

Post by kc55 »

Hi Martyn, thank you for your reply.
I was told this is due to ADC interleaving.
The specification states when up to 4 channels are used, the max sampling rate is 250 MS/s. Does that mean each ADC is able to sample at up to 250 MS/s?
If so, is it possible to disable this interleaving when only 1 or 2 channels are used, and when sampling rate per channel is <= 250 MS/s? This should bring back the almost 255 vertical steps.
With interleaving active, I visually counted the number of vertical steps is only 212, which is consistent with calculated result with a step size of approximately 47.4 mV (10,000 / 47.4 = 211).
So, the number of vertical steps is reduced by 255 - 212 = 43, or almost 17%. Loosing 43 steps for an 8-bit scope seems a little too much.

Regards,
kc

Martyn
Site Admin
Site Admin
Posts: 4491
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: Picoscope 3404D MSO strange vertical step size

Post by Martyn »

Unfortunately it will not be possible to do this at the present time due to space limitations in the available code space on the device.
Martyn
Technical Support Manager

Post Reply