[Halld-tracking-hw] New preamp

[Fernando J. Barbosa]

Hi Naomi,

My comments follow:

Naomi Jarvis wrote:
> Hi Fernando,
>
> I started testing  the new preamp which arrived today and have some 
> questions about it.
>
> There is a scope trace on 
> http://www.jlab.org/Hall-D/software/wiki/index.php/Preamp_3
> This is from cosmic ray data, from the shaper unit output, and it 
> shows little noise.

The signals from the new preamp look great. The noise has been improved 
on the new ASIC.
> The baseline voltage was originally 600mV before adjusting the shaper 
> unit to bring the baseline down to that shown in the oscilloscope 
> trace (upper left picture on the Wiki).  With the new preamp and old 
> cable the baseline was around 200mV (lower left).  I understand that 
> the cable inverts the polarity but I expected it to invert the 
> baseline voltage too.  Maybe the shaper unit is treating the positive 
> and negative signals differently, should I be concerned about this?  
> We can invert the ADC values in the data acquisition program if 
> necessary, but I just wanted to confirm that the inversion is deliberate?

The old preamp had only one gain setting and it was deliberately set 
about mid-way to where we estimated the dynamic ranges for the FDC 
cathodes and the anodes (CDC and FDC). After testing with the FDC and 
the CDC, two gain settings were implemented (this was always the 
intention to have one ASIC design cover both of the drift chambers). The 
CDC anodes use the low gain setting and the scope pictures shows that 
the ratio (old and new versions) to be about right and therefore, you 
should see a lower offset. The actual offset will be different from 
channel-to-channel and from chip-to-chip but not by much. So, what you 
observe is correct. Also, the shaper was optimized for the old preamp 
and the 72-ch fADC125 will be properly set for the new preamp (I sent 
Gerard a new preamp and a new cable). Inverting the ADC values is the 
right approach with the new cable.
> My other concern is over the mapping of the preamp channels through 
> the cable.  We are now using the new preamp and new cable with the old 
> HV board.  The correspondence between the numbering on the HV board 
> and the preamp channels has changed.  Is this now the final mapping?
>
> Previously we were using the cable with channels 1-8 and 17-24 feeding 
> into the shaper unit, which has 16 channels total, and then into two 
> 8-channel fADCs.  The chamber prototype is configured so that the 
> missing 8 channels correspond to the straws along 2 sides of the 
> chamber; the two groups of 8 channels in use correspond to two blocks 
> of 8 straws, one above the other.  With the new cable it seems that we 
> now have access to the 8 channels which were missing before (along the 
> edge), but since we only have 16 channels from the shaper, we have 
> lost access to one block of 8 straws.  This makes cosmic ray tracking 
> much slower, since we are looking for tracks passing through at least 
> 4 straws.  This is hard to explain without a diagram but the effect is 
> that the configuration of our available channels is much less as 
> efficient than before.  This will be solved when we have a 72-channel 
> fADC; alternatively we could just use the old cable until the new fADC 
> is available.

The old preamps (also designed for 24-channels) had only two chips 
installed (on top) because the ASIC package was not reliable for 
assembly on the bottom (it had a removable lid in case we needed to 
probe the die). The preamp card has channels 1-8 from one chip on top, 
9-16 from bottom chip and 17-24 from the second chip on top. Because of 
the 16-channel readout on the shaper, the cable (old cable) was modified 
at the ERNI connector (connects to preamp) to read channels 1-8 and 
17-24. This is the reason for what you observe with the new cable. This 
won't be the case with the 72-ch fADC125 because it reads from all 24 
channels on a preamp card. Yes, please you use the old cable with the 
new preamp and shaper for your immediate tests and then use the new 
cable when the 72-ch fADC125 becomes available (late April).

Great job and thanks for your feedback.

Best regards,
Fernando





>
> Thank you for your help,
>
> Naomi.
>
>
>
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: barbosa.vcf
Type: text/x-vcard
Size: 188 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/halld-tracking-hw/attachments/20100408/76d73c1c/attachment.vcf