On Wednesday, September 12, 2018 at 11:45:38 AM UTC-4, John Larkin wrote:> ADA4522 is an amazing amp. 55 volt supplies, low noise, 5 uV max > offset, EMI hardened, pA bias current. But it has a hangup mode as a > follower: if the output goes to V+ the input back-to-back diodes keep > both inputs above the legal common-mode range. Pity it's not RRIO. > > Similar amps from TI and Maxim seem to have the same issue. > > Here's one fix: > > https://www.dropbox.com/s/tdzs1ydi7quzbs7/ADA4522_hangup.JPG?raw=1 > > We had the depletion fets there already, to protect the amp, so all > we'll do is add the dual zener. Above 8 volts or so, small zeners [1] > have tiny pre-breakdown currents, so this should be fine for +-10 volt > operating inputs. We'll test some dual zeners to be sure. > > [1] some PITA will naturally open the zener/avalanche debate again. > Everybody calls them zeners.There is no such hang-up mode according to fig 73 of the datasheet, and that's a pretty low frequency signal at 400 seconds/div: http://www.analog.com/media/en/technical-documentation/data-sheets/ada4522-1_4522-2_4522-4.pdf You have something else going on.> > > > > -- > > John Larkin Highland Technology, Inc > > lunatic fringe electronics
ADA4522 opamp hangup
Started by ●September 12, 2018
Reply by ●September 15, 20182018-09-15
Reply by ●September 15, 20182018-09-15
On Saturday, 15 September 2018 18:49:22 UTC+2, bloggs.fred...@gmail.com wrote:> On Wednesday, September 12, 2018 at 11:45:38 AM UTC-4, John Larkin wrote: > > ADA4522 is an amazing amp. 55 volt supplies, low noise, 5 uV max > > offset, EMI hardened, pA bias current. But it has a hangup mode as a > > follower: if the output goes to V+ the input back-to-back diodes keep > > both inputs above the legal common-mode range. Pity it's not RRIO. > > > > Similar amps from TI and Maxim seem to have the same issue. > > > > Here's one fix: > > > > https://www.dropbox.com/s/tdzs1ydi7quzbs7/ADA4522_hangup.JPG?raw=1 > > > > We had the depletion fets there already, to protect the amp, so all > > we'll do is add the dual zener. Above 8 volts or so, small zeners [1] > > have tiny pre-breakdown currents, so this should be fine for +-10 volt > > operating inputs. We'll test some dual zeners to be sure. > > > > [1] some PITA will naturally open the zener/avalanche debate again. > > Everybody calls them zeners. > > There is no such hang-up mode according to fig 73 of the datasheet, and that's a pretty low frequency signal at 400 seconds/div: > http://www.analog.com/media/en/technical-documentation/data-sheets/ada4522-1_4522-2_4522-4.pdf > > You have something else going on. > > > > > > > > > > > > -- > > > > John Larkin Highland Technology, Inc > > > > lunatic fringe electronicsLook at page 30 of the datasheet. It's a documented bug (a.k.a. "a feature"). It really does this trick, I can confirm. In other configurations it works just fine. I've probed around with an FFT analyser - the spikes both in the voltage and current noise are not huge, and they start in the tens of kHz. Below that it's well-behaved. Among the chopping amps with low voltage rails, OPA388 is worth mentioning. Cheers, Nikolai
Reply by ●September 15, 20182018-09-15
On Sat, 15 Sep 2018 10:15:14 -0700 (PDT), Castorp <nikolaibeev@gmail.com> wrote:>On Saturday, 15 September 2018 18:49:22 UTC+2, bloggs.fred...@gmail.com wrote: >> On Wednesday, September 12, 2018 at 11:45:38 AM UTC-4, John Larkin wrote: >> > ADA4522 is an amazing amp. 55 volt supplies, low noise, 5 uV max >> > offset, EMI hardened, pA bias current. But it has a hangup mode as a >> > follower: if the output goes to V+ the input back-to-back diodes keep >> > both inputs above the legal common-mode range. Pity it's not RRIO. >> > >> > Similar amps from TI and Maxim seem to have the same issue. >> > >> > Here's one fix: >> > >> > https://www.dropbox.com/s/tdzs1ydi7quzbs7/ADA4522_hangup.JPG?raw=1 >> > >> > We had the depletion fets there already, to protect the amp, so all >> > we'll do is add the dual zener. Above 8 volts or so, small zeners [1] >> > have tiny pre-breakdown currents, so this should be fine for +-10 volt >> > operating inputs. We'll test some dual zeners to be sure. >> > >> > [1] some PITA will naturally open the zener/avalanche debate again. >> > Everybody calls them zeners. >> >> There is no such hang-up mode according to fig 73 of the datasheet, and that's a pretty low frequency signal at 400 seconds/div: >> http://www.analog.com/media/en/technical-documentation/data-sheets/ada4522-1_4522-2_4522-4.pdf >> >> You have something else going on. >> >> >> > >> > >> > >> > >> > -- >> > >> > John Larkin Highland Technology, Inc >> > >> > lunatic fringe electronics > > >Look at page 30 of the datasheet. It's a documented bug (a.k.a. "a feature").Right.> >It really does this trick, I can confirm. In other configurations it works just fine. I've probed around with an FFT analyser - the spikes both in the voltage and current noise are not huge, and they start in the tens of kHz. Below that it's well-behaved."May rail to the positive supply" is wussydatasheetspeak for "hangs up." All opamps rail around their positive supplies. It will come out of the latched rail condition if the input signal yanks the input down pretty hard... about a mA of sink maybe. The cited fig 73 does that. Our product is a resistor simulator, and we can't predict what the customer might do that could latch it up. A resistor is not expected to sit there and generate 11 volts. We can latch it up by connecting it to some auto-ranging DVMs that can output a bunch of voltage on some of their ohms ranges. Except for that little glitch, ADA4522 is fabulous.> >Among the chopping amps with low voltage rails, OPA388 is worth mentioning. > >Cheers, >NikolaiChoppers are getting way better lately, but too many are low voltage things, we need a high voltage follower. We have bootstrapped the power supplies of low voltage amps to do this, but that is sort of a nightmare. Or really a nightmare. -- John Larkin Highland Technology, Inc lunatic fringe electronics
Reply by ●September 15, 20182018-09-15
On Saturday, September 15, 2018 at 1:53:58 PM UTC-4, John Larkin wrote:> On Sat, 15 Sep 2018 10:15:14 -0700 (PDT), Castorp > <nikolaibeev@gmail.com> wrote: > > >On Saturday, 15 September 2018 18:49:22 UTC+2, bloggs.fred...@gmail.com wrote: > >> On Wednesday, September 12, 2018 at 11:45:38 AM UTC-4, John Larkin wrote: > >> > ADA4522 is an amazing amp. 55 volt supplies, low noise, 5 uV max > >> > offset, EMI hardened, pA bias current. But it has a hangup mode as a > >> > follower: if the output goes to V+ the input back-to-back diodes keep > >> > both inputs above the legal common-mode range. Pity it's not RRIO. > >> > > >> > Similar amps from TI and Maxim seem to have the same issue. > >> > > >> > Here's one fix: > >> > > >> > https://www.dropbox.com/s/tdzs1ydi7quzbs7/ADA4522_hangup.JPG?raw=1 > >> > > >> > We had the depletion fets there already, to protect the amp, so all > >> > we'll do is add the dual zener. Above 8 volts or so, small zeners [1] > >> > have tiny pre-breakdown currents, so this should be fine for +-10 volt > >> > operating inputs. We'll test some dual zeners to be sure. > >> > > >> > [1] some PITA will naturally open the zener/avalanche debate again. > >> > Everybody calls them zeners. > >> > >> There is no such hang-up mode according to fig 73 of the datasheet, and that's a pretty low frequency signal at 400 seconds/div: > >> http://www.analog.com/media/en/technical-documentation/data-sheets/ada4522-1_4522-2_4522-4.pdf > >> > >> You have something else going on. > >> > >> > >> > > >> > > >> > > >> > > >> > -- > >> > > >> > John Larkin Highland Technology, Inc > >> > > >> > lunatic fringe electronics > > > > > >Look at page 30 of the datasheet. It's a documented bug (a.k.a. "a feature"). > > Right.Ummm, no. The caveat on pg 30 refers to a power supply ramp up issue, and the fixes in Table 11 have Vin within the input CMR for each case. Your observation is about the amp input being overvoltaged, and fig 73 clearly shows the *output* clipped so as to keep IN(-) within the ICMR. Your working with a broken chip. Someone zapped it with ESD or something. If you can't reproduce fig. 73, your chip is broken. And this sound like the wrong part to trust to the outside world anyway. You either protect the input better or get another part.> > > > >It really does this trick, I can confirm. In other configurations it works just fine. I've probed around with an FFT analyser - the spikes both in the voltage and current noise are not huge, and they start in the tens of kHz. Below that it's well-behaved. > > "May rail to the positive supply" is wussydatasheetspeak for "hangs > up." All opamps rail around their positive supplies. > > It will come out of the latched rail condition if the input signal > yanks the input down pretty hard... about a mA of sink maybe. The > cited fig 73 does that. > > Our product is a resistor simulator, and we can't predict what the > customer might do that could latch it up. A resistor is not expected > to sit there and generate 11 volts. We can latch it up by connecting > it to some auto-ranging DVMs that can output a bunch of voltage on > some of their ohms ranges. > > Except for that little glitch, ADA4522 is fabulous. > > > > >Among the chopping amps with low voltage rails, OPA388 is worth mentioning. > > > >Cheers, > >Nikolai > > Choppers are getting way better lately, but too many are low voltage > things, we need a high voltage follower. We have bootstrapped the > power supplies of low voltage amps to do this, but that is sort of a > nightmare. Or really a nightmare. > > > -- > > John Larkin Highland Technology, Inc > > lunatic fringe electronics
Reply by ●September 15, 20182018-09-15
On Sat, 15 Sep 2018 16:45:20 -0700 (PDT), bloggs.fredbloggs.fred@gmail.com wrote:>On Saturday, September 15, 2018 at 1:53:58 PM UTC-4, John Larkin wrote: >> On Sat, 15 Sep 2018 10:15:14 -0700 (PDT), Castorp >> <nikolaibeev@gmail.com> wrote: >> >> >On Saturday, 15 September 2018 18:49:22 UTC+2, bloggs.fred...@gmail.com wrote: >> >> On Wednesday, September 12, 2018 at 11:45:38 AM UTC-4, John Larkin wrote: >> >> > ADA4522 is an amazing amp. 55 volt supplies, low noise, 5 uV max >> >> > offset, EMI hardened, pA bias current. But it has a hangup mode as a >> >> > follower: if the output goes to V+ the input back-to-back diodes keep >> >> > both inputs above the legal common-mode range. Pity it's not RRIO. >> >> > >> >> > Similar amps from TI and Maxim seem to have the same issue. >> >> > >> >> > Here's one fix: >> >> > >> >> > https://www.dropbox.com/s/tdzs1ydi7quzbs7/ADA4522_hangup.JPG?raw=1 >> >> > >> >> > We had the depletion fets there already, to protect the amp, so all >> >> > we'll do is add the dual zener. Above 8 volts or so, small zeners [1] >> >> > have tiny pre-breakdown currents, so this should be fine for +-10 volt >> >> > operating inputs. We'll test some dual zeners to be sure. >> >> > >> >> > [1] some PITA will naturally open the zener/avalanche debate again. >> >> > Everybody calls them zeners. >> >> >> >> There is no such hang-up mode according to fig 73 of the datasheet, and that's a pretty low frequency signal at 400 seconds/div: >> >> http://www.analog.com/media/en/technical-documentation/data-sheets/ada4522-1_4522-2_4522-4.pdf >> >> >> >> You have something else going on. >> >> >> >> >> >> > >> >> > >> >> > >> >> > >> >> > -- >> >> > >> >> > John Larkin Highland Technology, Inc >> >> > >> >> > lunatic fringe electronics >> > >> > >> >Look at page 30 of the datasheet. It's a documented bug (a.k.a. "a feature"). >> >> Right. > >Ummm, no. The caveat on pg 30 refers to a power supply ramp up issue, and the fixes in Table 11 have Vin within the input CMR for each case. Your observation is about the amp input being overvoltaged, and fig 73 clearly shows the *output* clipped so as to keep IN(-) within the ICMR. Your working with a broken chip. Someone zapped it with ESD or something. If you can't reproduce fig. 73, your chip is broken. And this sound like the wrong part to trust to the outside world anyway. You either protect the input better or get another part. >What can I say, after I say, "You don't understand this" ? -- John Larkin Highland Technology, Inc lunatic fringe electronics
Reply by ●September 16, 20182018-09-16
On Saturday, September 15, 2018 at 9:10:35 PM UTC-4, John Larkin wrote:> On Sat, 15 Sep 2018 16:45:20 -0700 (PDT), > bloggs.fredbloggs.fred@gmail.com wrote: > > >On Saturday, September 15, 2018 at 1:53:58 PM UTC-4, John Larkin wrote: > >> On Sat, 15 Sep 2018 10:15:14 -0700 (PDT), Castorp > >> <nikolaibeev@gmail.com> wrote: > >> > >> >On Saturday, 15 September 2018 18:49:22 UTC+2, bloggs.fred...@gmail.com wrote: > >> >> On Wednesday, September 12, 2018 at 11:45:38 AM UTC-4, John Larkin wrote: > >> >> > ADA4522 is an amazing amp. 55 volt supplies, low noise, 5 uV max > >> >> > offset, EMI hardened, pA bias current. But it has a hangup mode as a > >> >> > follower: if the output goes to V+ the input back-to-back diodes keep > >> >> > both inputs above the legal common-mode range. Pity it's not RRIO. > >> >> > > >> >> > Similar amps from TI and Maxim seem to have the same issue. > >> >> > > >> >> > Here's one fix: > >> >> > > >> >> > https://www.dropbox.com/s/tdzs1ydi7quzbs7/ADA4522_hangup.JPG?raw=1 > >> >> > > >> >> > We had the depletion fets there already, to protect the amp, so all > >> >> > we'll do is add the dual zener. Above 8 volts or so, small zeners [1] > >> >> > have tiny pre-breakdown currents, so this should be fine for +-10 volt > >> >> > operating inputs. We'll test some dual zeners to be sure. > >> >> > > >> >> > [1] some PITA will naturally open the zener/avalanche debate again. > >> >> > Everybody calls them zeners. > >> >> > >> >> There is no such hang-up mode according to fig 73 of the datasheet, and that's a pretty low frequency signal at 400 seconds/div: > >> >> http://www.analog.com/media/en/technical-documentation/data-sheets/ada4522-1_4522-2_4522-4.pdf > >> >> > >> >> You have something else going on. > >> >> > >> >> > >> >> > > >> >> > > >> >> > > >> >> > > >> >> > -- > >> >> > > >> >> > John Larkin Highland Technology, Inc > >> >> > > >> >> > lunatic fringe electronics > >> > > >> > > >> >Look at page 30 of the datasheet. It's a documented bug (a.k.a. "a feature"). > >> > >> Right. > > > >Ummm, no. The caveat on pg 30 refers to a power supply ramp up issue, and the fixes in Table 11 have Vin within the input CMR for each case. Your observation is about the amp input being overvoltaged, and fig 73 clearly shows the *output* clipped so as to keep IN(-) within the ICMR. Your working with a broken chip. Someone zapped it with ESD or something. If you can't reproduce fig. 73, your chip is broken. And this sound like the wrong part to trust to the outside world anyway. You either protect the input better or get another part. > > > > What can I say, after I say, "You don't understand this" ? > > > -- > > John Larkin Highland Technology, Inc > > lunatic fringe electronicsWhat could "power supply slow ramp time" possibly mean other than a turn on latch-up type of fault?? There's something wrong with your chip if it doesn't self-clip as shown in fig 73.
Reply by ●September 17, 20182018-09-17
> Choppers are getting way better lately, but too many are low voltage > things, we need a high voltage follower. We have bootstrapped the > power supplies of low voltage amps to do this, but that is sort of a > nightmare. Or really a nightmare. > >Indeed, such bootstraps are tough. I've been drooling over the datasheet of OPAx189: * CMRR: 168 dB (!) * Open-loop gain: 170 dB (!!) * en = 5.2 nV/sqrtHz, in = 165 fA/sqrtHz * drift: 5 nV/deg C typ. * huge input Z * Vsupply up to 36V Has anyone used these yet? Cheers, Nikolai
Reply by ●September 17, 20182018-09-17
On Monday, September 17, 2018 at 12:47:00 PM UTC-4, Castorp wrote:> > Choppers are getting way better lately, but too many are low voltage > > things, we need a high voltage follower. We have bootstrapped the > > power supplies of low voltage amps to do this, but that is sort of a > > nightmare. Or really a nightmare. > > > > > > Indeed, such bootstraps are tough. > > I've been drooling over the datasheet of OPAx189:Well for a few bucks you should buy a few and stop drooling. :^)> > * CMRR: 168 dB (!) > * Open-loop gain: 170 dB (!!) > * en = 5.2 nV/sqrtHz, in = 165 fA/sqrtHz > * drift: 5 nV/deg C typ. > * huge input Z > * Vsupply up to 36V > > Has anyone used these yet?No, but I like the opa192. Say the opa189 has the same type noise bump above ~100kHz as the ADA4522. (though not as much) http://www.ti.com/lit/ds/symlink/opa189.pdf See fig 17. A weird looking noise spectra. George H.> > Cheers, > Nikolai
Reply by ●September 17, 20182018-09-17
On Mon, 17 Sep 2018 09:46:55 -0700 (PDT), Castorp <nikolaibeev@gmail.com> wrote:> >> Choppers are getting way better lately, but too many are low voltage >> things, we need a high voltage follower. We have bootstrapped the >> power supplies of low voltage amps to do this, but that is sort of a >> nightmare. Or really a nightmare. >> >> > >Indeed, such bootstraps are tough. > >I've been drooling over the datasheet of OPAx189: > >* CMRR: 168 dB (!) >* Open-loop gain: 170 dB (!!) >* en = 5.2 nV/sqrtHz, in = 165 fA/sqrtHz >* drift: 5 nV/deg C typ. >* huge input Z >* Vsupply up to 36V > >Has anyone used these yet? > >Cheers, >NikolaiThat's a great amp, but its input common-mode range, near V+, is even worse than the ADA4522. All the super chopamps seem to have that problem. This one doesn't have input back-to-back diodes, so it won't have my hangup problem. DC psrr is great, but AC psrr is terrible. That could be a real-life noise limit. One big problem we had with bootstrapping a low-voltage chopamp was instability caused by psrr. -- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Reply by ●September 17, 20182018-09-17
On 9/17/18 3:31 PM, John Larkin wrote:> On Mon, 17 Sep 2018 09:46:55 -0700 (PDT), Castorp > <nikolaibeev@gmail.com> wrote: > >> >>> Choppers are getting way better lately, but too many are low voltage >>> things, we need a high voltage follower. We have bootstrapped the >>> power supplies of low voltage amps to do this, but that is sort of a >>> nightmare. Or really a nightmare. >>> >>> >> >> Indeed, such bootstraps are tough. >> >> I've been drooling over the datasheet of OPAx189: >> >> * CMRR: 168 dB (!) >> * Open-loop gain: 170 dB (!!) >> * en = 5.2 nV/sqrtHz, in = 165 fA/sqrtHz >> * drift: 5 nV/deg C typ. >> * huge input Z >> * Vsupply up to 36V >> >> Has anyone used these yet? >> >> Cheers, >> Nikolai > > That's a great amp, but its input common-mode range, near V+, is even > worse than the ADA4522. All the super chopamps seem to have that > problem. This one doesn't have input back-to-back diodes, so it won't > have my hangup problem. > > DC psrr is great, but AC psrr is terrible. That could be a real-life > noise limit. > > One big problem we had with bootstrapping a low-voltage chopamp was > instability caused by psrr. > >Yup. CMRR and PSRR are input-referred, i.e. the error gets multiplied by the gain of the stage, which can be a bit of a rude shock. 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
