On 02/10/2018 08:53 PM, Frank Miles wrote:> On Sat, 10 Feb 2018 17:18:55 -0500, Phil Hobbs wrote: > >> Is horrible. >> >> I'm just debugging a nice diode laser controller for one customer and >> getting ready to do another one for another (much better) customer. It >> uses a ZXTP25020 PNP with a biggish degeneration resistor and two-pole >> bypassing of the base to get low noise at high frequency. >> >> Turns out the 1/f noise is atrocious. The total noise from 2 Hz on up >> is only a couple of PPM, but it should be a factor of 10 better than >> that. With a total base resistance of 1k, the 1/f corner at 40 mA I_C >> is over 10 kHz. >> >> The bias network looks like this:+9V 0----------*--------*-------* | | | | | | | | R | 4.7 | 3.3 R 39 ohm TF | nF | nF R CCC CCC | CCC CCC | | | | / | | |< ZXTP25020 From 0--RRRR--*--RRRR--*---- | Op Amp |\ 499 499 | \ | | --- \ / --> V --> ----- | | GND>> >> The emitter resistor can drop up to about 5V. >> >> I've got some higher-beta transistors on order (FZT788B), and will >> reduce the impedance of the bias network, but I was fairly shocked that >> the 1/f noise was so bad. >> >> Cheers >> >> Phil Hobbs > > Are you sure your 9V supply is quiet, or that the op amp is quiet > relative to that supply?Both are a lot quieter than the collector current! > Even if you already have one > of your capacitance multipliers on the supply, lowering the filter > cutoff frequency might give a quick verification that it really is > 1/f. The acid test was putting a 1-uF film cap across the 3.3 nF to make the output stage bandwidth drop to about 160 Hz--the 1/f noise didn't change at all. That's pretty diagnostic for a base current problem, as you say.> > Thanks for the on-topic question!Well, some of us still work for a living. ;) Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
BJT base current 1/f noise
Started by ●February 10, 2018
Reply by ●February 11, 20182018-02-11
Reply by ●February 11, 20182018-02-11
"Phil Hobbs" <pcdhSpamMeSenseless@electrooptical.net> wrote in message news:OdadnTsVovEUoh3HnZ2dnUU7-dnNnZ2d@supernews.com...> Building a new-to-me kind of high dynamic range system always seems to > produce a bit of buried treasure at first. The ZXTP25020 has super-low > Rbb' and Ree', so it would be an excellent candidate except for the 1/f > noise. If the higher-beta device and lower drive impedance doesn't > improve it enough, I'll have to try Darlingtonizing it with one of those > nice SiGe:C devices to return the base current to the collector circuit.Would a hypothetical P-channel BF862 (rare as hens' teeth, I know) do as well? Tim -- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: https://www.seventransistorlabs.com/
Reply by ●February 11, 20182018-02-11
On 02/11/2018 04:47 AM, Kevin Aylward wrote:>> "Phil Hobbs" wrote in message >> news:11b6d95b-4867-8132-efc3-0926b68497eb@electrooptical.net... > >> Is horrible. > >> I'm just debugging a nice diode laser controller for one customer and >> getting ready to do another one for another (much better) customer. It >> uses a ZXTP25020 PNP with a biggish degeneration resistor and two-pole >> bypassing of the base to get low noise at high frequency. > > >> I've got some higher-beta transistors on order (FZT788B), and will >> reduce the impedance of the bias network, but I was fairly shocked that >> the 1/f noise was so bad. > > Why? > > I had a deko at the data sheet from Diodes Inc. It don't have any spec > for 1/f. So, again, what should one expect it to be....?They never have decent specs for 1/f, and their models never seem to include it either, so you just have to suck it and see. The FTZ788B has the highest guaranteed beta of any suitable transistor I could find, so it seemed like a good thing to try. The SiGe:C Darlington thing will be next on the list if reducing the base-circuit impedance doesn't solve it. Nice discretes seem to be disappearing fast though. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
Reply by ●February 11, 20182018-02-11
On 02/11/2018 08:13 AM, Tim Williams wrote:> "Phil Hobbs" <pcdhSpamMeSenseless@electrooptical.net> wrote in message > news:OdadnTsVovEUoh3HnZ2dnUU7-dnNnZ2d@supernews.com... >> Building a new-to-me kind of high dynamic range system always seems to >> produce a bit of buried treasure at first.� The ZXTP25020 has super-low >> Rbb' and Ree', so it would be an excellent candidate except for the >> 1/f noise.� If the higher-beta device and lower drive impedance >> doesn't improve it enough, I'll have to try Darlingtonizing it with >> one of those >> nice SiGe:C devices to return the base current to the collector circuit. > > Would a hypothetical P-channel BF862 (rare as hens' teeth, I know) do as > well?The late lamented BF862 had a 1/f corner of about a kilohertz, but its voltage noise never got below about 0.8 nV in 1 Hz. A BJT with super low Rbb' and Ree' and running at highish I_C can do a lot better than that. Within the loop bandwidth, I could suppress the noise by putting the sense resistor in the collector circuit, but then I'd have full shot noise at high frequency, which is what I'm trying to avoid. I'm dropping as much as 5V across the emitter resistor to try to get way below the shot noise. There are more complicated things I can do, but that takes actual work, and I'm working Saturdays as it is. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
Reply by ●February 11, 20182018-02-11
Reply by ●February 11, 20182018-02-11
Den lørdag den 10. februar 2018 kl. 23.19.02 UTC+1 skrev Phil Hobbs:> Is horrible. > > I'm just debugging a nice diode laser controller for one customer and > getting ready to do another one for another (much better) customer. It > uses a ZXTP25020 PNP with a biggish degeneration resistor and two-pole > bypassing of the base to get low noise at high frequency. > > Turns out the 1/f noise is atrocious. The total noise from 2 Hz on up > is only a couple of PPM, but it should be a factor of 10 better than > that. With a total base resistance of 1k, the 1/f corner at 40 mA I_C > is over 10 kHz. > > The bias network looks like this: > > +9V 0----------*--------*-------* > | | | > | | | > | | R > | 4.7 | 3.3 R 39 ohm TF > | nF | nF R > CCC CCC | > CCC CCC | > | | | / > | | |< ZXTP25020 > From 0--RRRR--*--RRRR--*---- | > Op Amp |\ > 499 499 | \ > | > | > --- > \ / --> > V --> > ----- > | > | > GND > > The emitter resistor can drop up to about 5V. > > I've got some higher-beta transistors on order (FZT788B), and will > reduce the impedance of the bias network, but I was fairly shocked that > the 1/f noise was so bad. >https://people.phys.ethz.ch/~pmaerki/public/resistor_flicker_noise/20171219a_white_paper_resistor_flicker_noise.pdf ?
Reply by ●February 11, 20182018-02-11
On 02/11/2018 08:59 AM, Long Hair wrote:> Phil Hobbs wrote: > > snip > > What is the resistor in your bias circuit made of? (type) >They're all thin (metal) film, precisely to reduce 1/f noise. Good thought though. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
Reply by ●February 11, 20182018-02-11
On 02/11/2018 09:11 AM, Lasse Langwadt Christensen wrote:> Den lørdag den 10. februar 2018 kl. 23.19.02 UTC+1 skrev Phil Hobbs: >> Is horrible. >> >> I'm just debugging a nice diode laser controller for one customer and >> getting ready to do another one for another (much better) customer. It >> uses a ZXTP25020 PNP with a biggish degeneration resistor and two-pole >> bypassing of the base to get low noise at high frequency. >> >> Turns out the 1/f noise is atrocious. The total noise from 2 Hz on up >> is only a couple of PPM, but it should be a factor of 10 better than >> that. With a total base resistance of 1k, the 1/f corner at 40 mA I_C >> is over 10 kHz. >> >> The bias network looks like this: >> >> +9V 0----------*--------*-------* >> | | | >> | | | >> | | R >> | 4.7 | 3.3 R 39 ohm TF >> | nF | nF R >> CCC CCC | >> CCC CCC | >> | | | / >> | | |< ZXTP25020 >> From 0--RRRR--*--RRRR--*---- | >> Op Amp |\ >> 499 499 | \ >> | >> | >> --- >> \ / --> >> V --> >> ----- >> | >> | >> GND >> >> The emitter resistor can drop up to about 5V. >> >> I've got some higher-beta transistors on order (FZT788B), and will >> reduce the impedance of the bias network, but I was fairly shocked that >> the 1/f noise was so bad. >> > > https://people.phys.ethz.ch/~pmaerki/public/resistor_flicker_noise/20171219a_white_paper_resistor_flicker_noise.pdf > > ? >The resistors are all thin film. Thick film ones are different-looking, for one thing. Maerki's data are interesting though, thanks. I expect that it would be interesting to measure the 1/f noise and drift of just the batteries, using the cross-power-spectrum method. Down in the <1 Hz range their temperature drift isn't insignificant. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
Reply by ●February 11, 20182018-02-11
Phil Hobbs wrote:> On 02/11/2018 08:59 AM, Long Hair wrote: >> Phil Hobbs wrote: >> >> snip >> >> What is the resistor in your bias circuit made of? (type) >> > > They're all thin (metal) film, precisely to reduce 1/f noise. Good > thought though. > > Cheers > > Phil Hobbs >I wonder what the noise is on these... https://www.techexplorist.com/first-evidence-light-can-stop-electrons/11501/
Reply by ●February 11, 20182018-02-11
Am 11.02.2018 um 15:25 schrieb Phil Hobbs: though, thanks. I expect that it would be> interesting to measure the 1/f noise and drift of just the batteries, > using the cross-power-spectrum method. Down in the <1 Hz range their > temperature drift isn't insignificant.< https://www.researchgate.net/publication/3622215_Measurement_of_voltage_noise_in_chemical_batteries > < http://www.hoffmann-hochfrequenz.de/downloads/NoiseMeasurementsOnChemicalBatteries.pdf > My 89441A that I use as a FFT analyzer is severely 1/f challenged. I'm looking for something better. I'm working on a chopper amplifier that seems to break the 100pV/rt HZ, but my FFT analyzer hides just that where it starts to be interesting. But it could do cross spectrum. Since the 1/f seems to come from the sample clock and not from the inputs, there still is hope. cheers, Gerhard
