Maker Pro
Maker Pro

Transimpedance op amps?

J

jack

Jan 1, 1970
0
As part of an ongoing learning project I've come up on a bit of
confusion I hope someone can help me with. I will just list the
issues...
1. Is a voltage feed back op amp wired to be a transimpedance
op amp the same as a "charge sensitive" op amp?
2. Can a current feed back op amp be wired to be a transimpedance
device in the same way one would do it with a voltage feedback op
amp,
in order to take advantage of its high slew rates and thus be a easy
way to make a short current pulse amplifier?
AofE ( thanks to Winfred Hill's suggestion) has a nice high speed
pulse amp but I am still looking for a easier way to implement it with
op amps, hopefully with current feedback , is it possible? thanks
again for any help with this. jack
 
J

John Larkin

Jan 1, 1970
0
As part of an ongoing learning project I've come up on a bit of
confusion I hope someone can help me with. I will just list the
issues...
1. Is a voltage feed back op amp wired to be a transimpedance
op amp the same as a "charge sensitive" op amp?

A charge amp has a capacitor as the feedback impedance, whereas a
transimpedance amp uses a resistor.
2. Can a current feed back op amp be wired to be a transimpedance
device in the same way one would do it with a voltage feedback op
amp,
in order to take advantage of its high slew rates and thus be a easy
way to make a short current pulse amplifier?

Yup; but eyeball the current noise specs first. There are also some
specific transimpedance amps, designed for amplifying the output of
PIN diodes in fiberoptic systems; they're very good at this, but are
generally intended to be used in AC-coupled systems. Maxim has a few
nice ones.

John
 
R

Rene Tschaggelar

Jan 1, 1970
0
jack said:
As part of an ongoing learning project I've come up on a bit of
confusion I hope someone can help me with. I will just list the
issues...
1. Is a voltage feed back op amp wired to be a transimpedance
op amp the same as a "charge sensitive" op amp?
2. Can a current feed back op amp be wired to be a transimpedance
device in the same way one would do it with a voltage feedback op
amp,
in order to take advantage of its high slew rates and thus be a easy
way to make a short current pulse amplifier?
AofE ( thanks to Winfred Hill's suggestion) has a nice high speed
pulse amp but I am still looking for a easier way to implement it with
op amps, hopefully with current feedback , is it possible? thanks
again for any help with this. jack

Could you specify the pulses you're intending to measure ?
Length, current and from what source ?

Rene
 
J

jack

Jan 1, 1970
0
Yes,thankyou. I posted alot of the parameters above in a thread called
"advice needed...PMT" . Also apologies to winfield hill for the
misspelling of his name.
 
J

John Larkin

Jan 1, 1970
0
Yes,thankyou. I posted alot of the parameters above in a thread called
"advice needed...PMT" . Also apologies to winfield hill for the
misspelling of his name.


Why use a TIA? Just run a coax from the PMT, terminate in 50 ohms
(plus some overload clamps!) and use a voltage amp. PMTs generally
deliver lots of signal. If s/n is critical, run the coax into a 50-ohm
input SiGe MMIC.

Oh, yeah, I think he spells his name Winfield Hill.

John
 
B

Bill Sloman

Jan 1, 1970
0
John Larkin said:
Why use a TIA? Just run a coax from the PMT, terminate in 50 ohms
(plus some overload clamps!) and use a voltage amp. PMTs generally
deliver lots of signal. If s/n is critical, run the coax into a 50-ohm
input SiGe MMIC.

Over on his PMT thread, we can see that he is looking at 0.8mV pulses,
with a FWHM of 6.5nsec, so he doesn't more than about 100MHz of
bandwidth. SiGe would be something of an overkill - most of the
silicon-based op amps that would give him enough gain-bandwidth have
got about the same sort of voltage noise as the 50R resistor, so a
quieter amplifier would be similarly pointless.
 
J

jack

Jan 1, 1970
0
Why use a TIA? Just run a coax from the PMT, terminate in 50 ohms
(plus some overload clamps!) and use a voltage amp. PMTs generally
deliver lots of signal. If s/n is critical, run the coax into a 50-ohm
input SiGe MMIC.

Oh, yeah, I think he spells his name Winfield Hill.

John
Yep,I'm gona start there. The Hamamatsu Handbook throws a lot of
issues at you (noise,linearity ect) so I'll try all three approaches
just for learning,i.e. resistor terminated,TIA,and charge sensitve op
amp. Thanks all for the contributions......jack
 
I

Ian Buckner

Jan 1, 1970
0
John Larkin said:
Contrary to what some people will tell you, a 50 ohm resistor followed
by an amp will be noisier than an amp that is itself a 50 ohm load.
Given a 6 dB noise figure overall (achievable with a termination
followed by a very quiet amp) and 50 MHz bandwidth, the amp's noise
contribution will be about 14 microvolts RMS. If that's OK for your
needs, go for it. An uncooled, matched amp can cut that noise about in
half. Sirenza has some nice parts, below 3 dB NF, dirt cheap and easy
to use.

John

Jure pointed the OP (jack) at the OPA657 datasheet. John's post above
reminded me of an application cct in the OPA687 datasheet for a low
noise 50 ohm amp (see Fig 5, p11) where the Zin is set by feedback.
That one is a fair bit quieter than the OPA657. (Unfortunate part
number similarity - note it is OPA6_8_7).

Regards
Ian
 
W

Winfield Hill

Jan 1, 1970
0
Ian Buckner wrote...
Jure pointed the OP (jack) at the OPA657 datasheet. John's post
above reminded me of an application cct in the OPA687 datasheet
for a low noise 50 ohm amp (see Fig 5, p11) where the Zin is set
by feedback. That one is a fair bit quieter than the OPA657.
(Unfortunate part number similarity - note it is OPA6_8_7).

Don't forget the OPA687 has a 20uA or more input bias current.

Thanks,
- Win
 
Top