Hi,
The lowest noise dual JFETop-amp that I can find is the OPA1642 with 5.1 nV/root-Hz of voltage noise. I need the low current noise too (it's 0.6 fA/root-Hz for the aforementioned op amp).
Does anyone know of a lower voltage noise dual JFET op amp? I know there are lower noise dual op amps, but all that I've seen have much higher current noise than my application can tolerate.
Oh, I'd like one that runs with +/- 15 volt supplies too.
Other spec I need is at least 10 MHz GBP.
Regards,
Ken
The lowest noise dual JFETop-amp that I can find is the OPA1642 with 5.1 nV/root-Hz of voltage noise. I need the low current noise too (it's 0.6 fA/root-Hz for the aforementioned op amp).
Does anyone know of a lower voltage noise dual JFET op amp? I know there are lower noise dual op amps, but all that I've seen have much higher current noise than my application can tolerate.
Oh, I'd like one that runs with +/- 15 volt supplies too.
Other spec I need is at least 10 MHz GBP.
Regards,
Ken
where does the noise need to be low - A weighting?, what source/feedback Z?
The lowest noise jfet input is the single AD743/5
if lowest possible 1/f corner frequency isn't important then some CMOS parts may be useful
plot from thread http://www.diyaudio.com/forums/solid-state/68793-new-interesting-cmos-op-amps.html
The lowest noise jfet input is the single AD743/5
if lowest possible 1/f corner frequency isn't important then some CMOS parts may be useful
plot from thread http://www.diyaudio.com/forums/solid-state/68793-new-interesting-cmos-op-amps.html
Attachments
Thanks,
The AD743 has too high input C for my application, otherwise the 2.9 nV/root-Hz is attractive. The high C increases the noise gain of the first stage.
My application isn't audio, it's a transimpedance amplifier that's used to amplify a photo diode. I need ultra-low noise for this application. I attached a schematic to this reply. It has a unique topology in that I use two capacitors to multiply the photodiode AC current by a factor of 30, noise-free. I get a total effective transimpedance of 41 MOhm, but only get noise that's equivalent to a 1.34 MOhm resistor.
Gain peaking occurs around 1 kHz and the Vn starts to become a problem because the noise gain of the op-amps starts to get large. Current noise needs to be low too, so a JFET (or CMOS) op-amp is the only choice.
Ken
The AD743 has too high input C for my application, otherwise the 2.9 nV/root-Hz is attractive. The high C increases the noise gain of the first stage.
My application isn't audio, it's a transimpedance amplifier that's used to amplify a photo diode. I need ultra-low noise for this application. I attached a schematic to this reply. It has a unique topology in that I use two capacitors to multiply the photodiode AC current by a factor of 30, noise-free. I get a total effective transimpedance of 41 MOhm, but only get noise that's equivalent to a 1.34 MOhm resistor.
Gain peaking occurs around 1 kHz and the Vn starts to become a problem because the noise gain of the op-amps starts to get large. Current noise needs to be low too, so a JFET (or CMOS) op-amp is the only choice.
Ken
Hi,
I looked at a few CMOS op-amps from the thread you sent me, however they operate from low supply voltages. I need a dual that operates from +/-15V.
Also, my circuit's bandwidth is 100 Hz to 10 KHz.
Regards,
Ken
I looked at a few CMOS op-amps from the thread you sent me, however they operate from low supply voltages. I need a dual that operates from +/-15V.
Also, my circuit's bandwidth is 100 Hz to 10 KHz.
Regards,
Ken
you might seach photodiode transimpedance bootstrapping
I remember seeing descrete jfet Q in front of op amp photodiode transimpedance amps
I need to spend some time to see if this paper's bjt cascode/bootstrap is a good idea http://www.electrooptical.net/www/frontends/frontends.pdf
also you shouldn't let Vsupply prevent you from using the best op amps for the applicaiton - higher speed processes are lower V, you can make a compostie op amp circuit with a 2nd hgier V output chip if you need the Vout
I remember seeing descrete jfet Q in front of op amp photodiode transimpedance amps
I need to spend some time to see if this paper's bjt cascode/bootstrap is a good idea http://www.electrooptical.net/www/frontends/frontends.pdf
also you shouldn't let Vsupply prevent you from using the best op amps for the applicaiton - higher speed processes are lower V, you can make a compostie op amp circuit with a 2nd hgier V output chip if you need the Vout
Last edited:
JFET bootstrap is a good idea, I thought of that too, may have to go that way if I can't find a dual with Vn < 3 nV/root-Hz. Real estate is kind of a problem since the PCB is has to be about the size of a two pieces of Dentyne gum.
Hi,
Sounds like you need to use a NE5532 with a pair of discrete dual J-Fet's type 2SK389 (Toshiba) or LSK389 (Linear Systems) and make a hybrid.
This should manage < 1nV |/Hz J-fet Dual. Only needs the dual J-Fet and 3 SMD resistors added per op-amp, survivable even in a confined space, SMD versions of Op-Amp and J-Fets are available.
Ciao T
Sounds like you need to use a NE5532 with a pair of discrete dual J-Fet's type 2SK389 (Toshiba) or LSK389 (Linear Systems) and make a hybrid.
This should manage < 1nV |/Hz J-fet Dual. Only needs the dual J-Fet and 3 SMD resistors added per op-amp, survivable even in a confined space, SMD versions of Op-Amp and J-Fets are available.
Ciao T
Last edited:
could one assemble a pair of lsk389 and a pair of 5534, all smd size, onto a converter board that fits an 8 pin dil - dual opamp socket?
The 389 would feed into pins 5&8 of the 5534, thus removing the BJT input pair. Check which pins go to the BJT collectors !
The 389 would feed into pins 5&8 of the 5534, thus removing the BJT input pair. Check which pins go to the BJT collectors !
Thanks Andrew,
Collectors of the NE5534 come out to pins 1 and 8.
Thanks everyone for the suggestions.
Ken
Collectors of the NE5534 come out to pins 1 and 8.
Thanks everyone for the suggestions.
Ken
- Status
- Not open for further replies.