[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Shaping time Setting (fwd)
FYI
( `,_' )+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=
) `\ -
/ '. | +
| `, Elke-Caroline Aschenauer =
\,_ `-/ -
,&&&&&V Jefferson Lab +
,&&&&&&&&: HALL-D 12C / F381 121-A Atlantic Avenue =
,&&&&&&&&&&; Mailstop: 12H5 Hampton, VA 23664 -
| |&&&&&&&;\ 12000 Jefferson Ave +
| | :_) _ Newport News, VA 23606 Tel.: 001-757-224-1216 =
| | ;--' | Mail: elke@jlab.org Mobil: 001-757-256-5224 -
'--' `-.--. | +
\_ | |---' Tel.: 001-757-269-5352 =
`-._\__/ Fax.: 001-757-269-6248 -
+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+=-+
---------- Forwarded message ----------
Date: Sun, 02 Dec 2007 01:42:46 -0500
From: Mitch Newcomer <mitch@hep.upenn.edu>
To: "Visser, Gerard" <gvisser@indiana.edu>
Cc: Fernando J. Barbosa <barbosa@jlab.org>,
Yves Van Haarlem <yvhaarle@ernest.phys.cmu.edu>,
Daniel S. Carman <carman@jlab.org>, Simon Taylor <staylor@jlab.org>,
Curtis Meyer <cmeyer@ernest.phys.cmu.edu>,
Elke-Caroline Aschenauer <elke@jlab.org>,
Armen Stepanyan <armenak@jlab.org>, Chris Cuevas <cuevas@jlab.org>,
Mark Taylor <taylorw@jlab.org>
Subject: Re: Shaping time Setting
Yes 18 Meters is a significant distance, but the track signals are far
from pristine. It sound like a good time to make some in situ (track)
signal observations. They are sure to make Fe55 signals look like the
Taj Mahal. of pulses. My real motivation for responding however, is to
remind you all that there is 1K to Vdd and 1K to Gnd on chip so you
are terminated with 500 ohms into Vdd/2. This allow termination
across the differential lines at the far end. It should hold the
common mode at the far end to a reasonable value without requiring
external resistors. No I didnt try it yet, yes I would keep the back
termination option at least on the board. -- Mitch
Visser, Gerard wrote:
> Hi Mitch,
> Our cables will be about 18m long, I think that counts as "long" and source termination is probably going to be useful. Besides, it is only these resistors which set the common-mode termination of the outputs, i.e., the load end of the cable has only a floating differential termination. So they cannot be omitted entirely. But you are right, we could possibly save some power by increasing the value of the source termination resistors to at least the point where reflections on the cable are barely within the error budget (probably 1%, I say that on the basis of the likely precision of the cable frequency equalization i.e. tail cancellation - but this, I mean practical achievable "error budget" for the pulse tail, is being studied right now, I don't have a final determination yet). We should keep it in mind, to use higher values of these resistors possibly.
> Thanks,
>
> Gerard
>
> p.s. I forgot about the internal 500 Ohm resistor to ground on each output... That means the preamp boards should have been stuffed with 56.2 Ohm resistors not 51.1 Ohm to get a matched termination here. Oops. Sorry Fernando!
>
>
> ________________________________________
> From: Mitch Newcomer [mitch@hep.upenn.edu]
> Sent: Friday, November 30, 2007 8:26 PM
> To: Fernando J. Barbosa
> Cc: Yves Van Haarlem; Visser, Gerard; Daniel S. Carman; Simon Taylor; Curtis Meyer; Elke-Caroline Aschenauer; Armen Stepanyan; Chris Cuevas; Mark Taylor
> Subject: Shaping time Setting
>
> I would suggest leaving out the resistor that switches in additional
> shaping time (GT on the silk screen bottom side near output
> connector). You could also install it so that it goes to 0V having the
> same effect. Using a pulser shaped to look like an Ar gas signal we see
> a peaking time of 14ns with the resistor removed ( defaults to 0V) and
> 17.5ns with it in. Our board is stuffed with back termination which
> we resistors have removed. Since our load is 25 ohns per side when we
> use a mini-circuits RF transformer to go from the differential outputs
> to a single ended 50 ohm resistor going into the scope. An easy way to
> increase the observed gain is to increase the back termination resistor
> values. I'm not convinced they will be needed unless you are using
> long cables. In this case the cable is discharged more quickly with
> back termination. Increasing the size of the back termination has the
> added value of removing
> heat from the detector area. If you remove them completely you will
> still measure an impedance of ~1K between outputs or 500Ohms to Gnd
> due to the internal
> resistance on the outputs. -- Mitch
>
> PS We have measured an equivalent input noise of 3300e with a ~15pF
> input capacitance. The uncertainty is on the amout of capacitance on
> the preamp card and various connectors. I should say 3300e is measured
> when adding 10pF to the existing capacitance of board and connectors....
>
>
>
> Fernando J. Barbosa wrote:
>
>> Hello everyone,
>>
>> The preamp cards have now been assembled and functionally tested. I
>> will be shipping/delivering the preamp cards, power cables, signal
>> cables and interposer boards, as appropriate, on Monday so that you
>> will have them early next week. I will also email you some documentation.
>>
>> Gerard, please send me your address for delivery by Fedex. I have
>> Mitch's and Yves but please confirm.
>>
>> I should mention that our techs, Armen and Mark, did a superb job
>> in assembling and ordering all these parts in such short notice and
>> with limited resources. Thanks so much!
>>
>> Regards,
>> Fernando
>>