ADC values too large

[DavidAdey]

The values for x/yview_adc (which I think corresponds 1:1 to PE) seems too large, ranging from tens to thousands. The energy scale value seems Ok (70/MeV) so either the dE is wrong or the summing of the hits has a problem.

[tunnell]

Could it be a proton or neutron stopping? 1000 is 14 MeV... is that crazy? I agree though that electrons would saturate. It should be PE. All it does it multiply what I think is energy in MeV by 70 (or something similar).

[DavidAdey]

Maybe. I'd like to distinguish the hits by their PID, but that's a different issue. The obvious tracks look more sensible (1-4MeV / plane), so you're probably right. I'm just beginning to keep track of curious things so feel free to close the issue.

[tunnell]

Track away.... we at least know that we need to cap out the ADC values. Maybe cap at 1024?

On 14 Nov 2012, at 17:11, DavidAdey <[email protected]mailto:[email protected]>
wrote:

Maybe. I'd like to distinguish the hits by their PID, but that's a different issue. The obvious tracks look more sensible (1-4MeV / plane), so you're probably right. I'm just beginning to keep track of curious things so feel free to close the issue.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/4#issuecomment-10374460.

[tunnell]

Also: the PID isn't necessarily unique.

On 14 Nov 2012, at 17:11, DavidAdey <[email protected]mailto:[email protected]>
wrote:

Maybe. I'd like to distinguish the hits by their PID, but that's a different issue. The obvious tracks look more sensible (1-4MeV / plane), so you're probably right. I'm just beginning to keep track of curious things so feel free to close the issue.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/4#issuecomment-10374460.

[DavidAdey]

With the ADC we first need to multiply the PE by the gain of the electronics (the sample SiPM plots in the LOI show about 30/PE). We can probably go higher than 10bits for the limit, but the details I'm not sure on.

[tunnell]

How expensive do ADCs get? I thought after 10 or so they were pricey.

On 14 Nov 2012, at 18:05, DavidAdey <[email protected]mailto:[email protected]> wrote:

With the ADC we first need to multiply the PE by the gain of the electronics (the sample SiPM plots in the LOI show about 30/PE). We can probably go higher than 10bits for the limit, but the details I'm not sure on.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/4#issuecomment-10376666.

[DavidAdey]

Yeah, that's what I mean by details. They do jump in price, but it depends on how its balanced against rate, channels, and so on. I'll ask around.

[DavidAdey]

Ok, so the high values are too large - there's a correction to the linear scaling of light yield with energy.

On the ADCs, they will probably be 12bits but in effect only 10bits, with around 5 ADC counts per PE. Apparently this can be got around by using multiple chips with varying gains.

So, I've amended the light yield calculation, but will approximate the limit at 1000PE for now.