>Isn't a classic differential amp configuration a composite amp?
No, because there's no overall feedback from output to input.
Cheers
Phil Hobbs
Reply by Flyguy●May 14, 20202020-05-14
On Sunday, May 27, 2018 at 12:09:42 PM UTC-7, Phil Hobbs wrote:
> Hi, all,
>
> I'm generally prejudiced against composite amplifiers (two op amps
> inside one feedback loop) because they're generally squirrelly, with
> poor settling performance and weird transient response.
>
> On the other hand, my aversion to them means that I don't have as much
> experience with them as do composite-amp fans. So what do you folks say
> about them?
>
> Orchids? Onions? Actual expertise?
>
> Cheers
>
> Phil Hobbs
>
> --
> Dr Philip C D Hobbs
> Principal Consultant
> ElectroOptical Innovations LLC / Hobbs ElectroOptics
> Optics, Electro-optics, Photonics, Analog Electronics
> Briarcliff Manor NY 10510
>
> http://electrooptical.net
> http://hobbs-eo.com
Isn't a classic differential amp configuration a composite amp?
Reply by Phil Hobbs●May 13, 20202020-05-13
On 2020-05-13 16:51, plastcontrol.ru@gmail.com wrote:
> Another dinosaur with capacitors between the base and the emitter:
>
> A Low-Noise Precision Op Amp
> National Semiconductor Linear Brief 52
> Robert A. Pease 1980
> http://www.ti.com/lit/an/snoa702/snoa702.pdf
>
> I don’t understand why HP shunted the transition at the second transistor with a collector current 5 times the input stage...
>
It's sort of fun--it's a Darlington at DC and not at AC.
The Darlington topology is primarily intended to reduce the input bias
current. The 10 uF caps get rid of the gross AC voltage noise of the
driver pair Q1A / Q1B, and as a bonus they eliminate the transfer
function problems of the Darlington.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510
http://electrooptical.nethttp://hobbs-eo.com
Reply by ●May 13, 20202020-05-13
Another dinosaur with capacitors between the base and the emitter:
A Low-Noise Precision Op Amp
National Semiconductor Linear Brief 52
Robert A. Pease 1980
http://www.ti.com/lit/an/snoa702/snoa702.pdf
I don’t understand why HP shunted the transition at the second transistor with a collector current 5 times the input stage...
Reply by Phil Hobbs●May 4, 20202020-05-04
On 2020-05-03 15:00, plastcontrol.ru@gmail.com wrote:
>>>
>>> my composite voltage follower -
>>> http://ixbt.photo/?id=photo:1332241
>>>
>>> output signal from ADA4860 with photocurrent 1.23渙
>>> http://ixbt.photo/?id=photo:1332242
>>>
>>> output signal from OPA656 with photocurrent 1.23渙
>>> http://ixbt.photo/?id=photo:1332243
>>>
>>> rise time 70ns, noise 150mVp-p Channel 2 = output ADA4860
>>>
>>> I wish you all good health.
>>
>>
>> I have a new design, a little thing to keep busy with during the
>> shutdown. It's a GHz o/e converter. The trick is to use both ends
>> of the photodiode, use the current twice. One end goes into a fast
>> AC-coupled amp, and one end into a slow, precise DC gain path, and
>> combine them later.
Using the photocurrent twice is fun, I agree. I sometimes use a cascode
with two collector resistors in series, with a diode clamp on the
higher-value one. You digitize both, and use the lower-gain one when
the high-gain one rails.
That's handy when you have a relatively slow pulsed application where
the signal varies over a wideish range and you don't want to have to do
range switching. I last used it a couple of years ago, in the follow-on
version of my blood spot detector for hen's eggs.
That feedbeside thing is disappointing IME--it settles OK to 1% or so,
but then it's all whoopdedoos. Good for scopes, which is where it was
originally used.
> It’s interesting to try using still floating power supply op amp.
Yup. You do have to watch out for the CMR though--TIAs with
bootstrapped power supplies can lead to oscillations. CMR is
input-referred, and in a TIA the noise gain gets pretty high at high
frequency, just where the CMR is tanking.
It's usually the negative CMR that's the problem.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510
http://electrooptical.nethttp://hobbs-eo.com
> On Sun, 22 Mar 2020 02:17:10 -0700 (PDT), plastcontrol.ru@gmail.com
> wrote:
>
> >"Shunt bootstrapping technique to improve bandwidth of transimpedance amplifiers"
> >https://www.researchgate.net/publication/3380589_Shunt_bootstrapping_technique_to_improve_bandwidth_of_transimpedance_amplifiers
> >
> >fig.3 Output of bootstrap amplifier
> >
> >my composite voltage follower -
> >http://ixbt.photo/?id=photo:1332241
> >
> >output signal from ADA4860 with photocurrent 1.23渙
> >http://ixbt.photo/?id=photo:1332242
> >
> >output signal from OPA656 with photocurrent 1.23渙
> >http://ixbt.photo/?id=photo:1332243
> >
> >rise time 70ns, noise 150mVp-p
> >Channel 2 = output ADA4860
> >
> >I wish you all good health.
>
>
> I have a new design, a little thing to keep busy with during the
> shutdown. It's a GHz o/e converter. The trick is to use both ends of
> the photodiode, use the current twice. One end goes into a fast
> AC-coupled amp, and one end into a slow, precise DC gain path, and
> combine them later.
>
>
>
> --
>
> John Larkin Highland Technology, Inc
>
> Science teaches us to doubt.
>
> Claude Bernard
It’s interesting to try using still floating power supply op amp.
Dmitriy P.
I have a new design, a little thing to keep busy with during the
shutdown. It's a GHz o/e converter. The trick is to use both ends of
the photodiode, use the current twice. One end goes into a fast
AC-coupled amp, and one end into a slow, precise DC gain path, and
combine them later.
--
John Larkin Highland Technology, Inc
Science teaches us to doubt.
Claude Bernard
Reply by Cursitor Doom●May 3, 20202020-05-03
On Sun, 03 May 2020 09:15:17 -0700, Michael Terrell wrote:
On Sunday, May 3, 2020 at 8:09:24 AM UTC-4, Winfield Hill wrote:
> plastcontrol wrote...
> >
> > Retrospective "composite amplifier" challenge))
> > Question for experts and authors of the AoE3.
> > What is the function of capacitors 10nF between base and emitter MMBT6429 ?
> > https://electronix.ru/forum/uploads/monthly_06_2016/post-21169-1466350930.png
> >Why is there only one capacitor in the service manual between base and emitter ?
> > https://electronix.ru/forum/uploads/monthly_06_2016/post-21169-1466350949.png
> > Dmitriy P.
>
> You're right, my copy of the service manual
> only shows one capacitor. The symmetrical
> extra one came in, either when I made the
> pencil drawing for the draftsman, or they
> made a drafting error. As for the purpose
> of the capacitors, we can only guess. It'd
> be nice to be able to ask the HP engineer(s).
>
> Did we point out in the text that those are
> cascode transistors, whose only purpose is to
> isolate the op-amp active nodes from the very
> high capacitance of the huge IF3602 JFETs?
> The Darlington is to insure all of the emitter
> current ends up in the collector, and almost
> none in the base.
>
> The performance specs for their G=10,000
> amplifier are pretty amazing. We have one of
> those HP 34420A multimeters in my lab, but
> it's rarely used, it's far too good! And the
> special nanovolt input connector is awkward.