[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: shaper gain



Hi Fernando,
	Ok, just to check, this will be setting the gain 1/5 of the present 
"anode" shaper gain, is that correct? (If not, please say...)
	If I do have that right, I would recommend to do the following:
	- Set the middle gain stage same as the present "anode" shaper, i.e., 
feedback network to be 499 Ohm || 10pF. [Or for your convenience you can 
start with an "anode" board and then not have to change these parts.] 
This level of gain is probably about the minimum here, unless you start 
changing a lot of components. I.e., I don't think you can use much less 
than 499 Ohms here.
	- Set the output stage to about 1/sqrt(2) overall gain, presently it is 
unity gain. Or use an external 3dB resistive attenuator.  To be 
specific, presently R233 etc. is 49.9 Ohm, R232 etc. is open. Use 69.8 
Ohms, 165 Ohms respectively. Note these are 1/8W leaded resistors, to 
make this change "easier"... You could go down a little more, but 
probably no lower than an overall gain of 0.6 here. I think 1/sqrt(2) is 
almost as good and it gives you the option of just using an external 3dB 
instead of changing these resistors.
	- Take the rest of the desired gain change, (1/5)/(1/sqrt(2))=0.283, 
and apply it in the input attenuator network, i.e., R299-302 & R174 & 
R176, etc. Presently this has a gain of 0.488, so you want to set this 
instead to a gain of 0.138, while of course maintaining the 100 Ohm 
differential input impedance. I calculate that you could replace the 
64.9 Ohm resistors with 110 Ohm and the 61.9 Ohm resistors with 16.9 Ohm 
to do this. [There are other solutions of course, but note that the 64.9 
Ohm "series" resistance also provides some input protection function, I 
don't think it should be ever less than that value.]
	This change will of course increase the noise contribution from the 
shaper/ADC but it will still I think be dominated by the preamp noise as 
it should be. We need to verify this, for instance by a noise comparison 
with the preamp power on/off.
	By the way, you may ask why not to just lower the gain of the input 
stage instead of increasing the input attenuation, the reason is just 
the stability limit for the AD8129, you probably already have seen it on 
the datasheet. I just mention in case... [If you substituted the unity 
gain stable AD8130 you could then take no attenuation on the input stage 
and have roughly a factor 3.6 lower noise from the shaper board, but it 
probably that doesn't make a big difference. Since preamp noise is 
supposed to dominate.]
	I will cc this to the list just for the record. [No need for everyone 
to worry about these details though, please...]

	Gerard


Fernando J. Barbosa wrote:
> Hi Gerard,
> 
> I would like to change the gain of one of the high gain shaper to the
> new lower gain of about 2.16/5 = 0.43. I will get the SMT components for
> the change.
> In this way, we will have a shaper with a gain of 2.16 for the cathode
> strips and another with a gain of 0.43 for the sense wires. It may be
> that the ASIC will still be above the linear region with the Fe-55 but
> we will test with cosmics.
> 
> Thanks,
> Fernando