Forums

opamps coming off the rail

Started by John Larkin May 16, 2013

I need to drive a mosfet in current-sink mode. It's similar to this
circuit

https://dl.dropboxusercontent.com/u/53724080/Circuits/Current_Sources/Mirror2.JPG

but V- will be ground and I'll run the opamp off +8 or +10 or
something, whatever drive the fet needs to sink 150 amps.

I'd prefer to not have a negative power supply. At zero requested
current, I want the current to be really zero, namely the fet gate at
ground. I can futz in a little DC offset to guarantee that the opamp
rails low.

Opamps are often not spec'd for what happens when they rail, namely
how slow and how ugly they come out of saturation. An internal
compensation node can wind up. I'd like to get back in business in
ballpark 1 us.

Does anybody have favorite R-R opamps that come out of saturation fast
and clean?

I recall one Intersil chopamp that took several seconds to recover.


-- 

John Larkin         Highland Technology, Inc

jlarkin at highlandtechnology dot com
http://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser drivers and controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro   acquisition and simulation
On Thursday, May 16, 2013 9:20:37 PM UTC+2, John Larkin wrote:
> I need to drive a mosfet in current-sink mode. It's similar to this > > circuit > > > > https://dl.dropboxusercontent.com/u/53724080/Circuits/Current_Sources/Mirror2.JPG > > > > but V- will be ground and I'll run the opamp off +8 or +10 or > > something, whatever drive the fet needs to sink 150 amps. > > > > I'd prefer to not have a negative power supply. At zero requested > > current, I want the current to be really zero, namely the fet gate at > > ground. I can futz in a little DC offset to guarantee that the opamp > > rails low. > > > > Opamps are often not spec'd for what happens when they rail, namely > > how slow and how ugly they come out of saturation. An internal > > compensation node can wind up. I'd like to get back in business in > > ballpark 1 us. > > > > Does anybody have favorite R-R opamps that come out of saturation fast > > and clean? > > > > I recall one Intersil chopamp that took several seconds to recover. >
LM324.... just kidding :-) Can't you add a clamping circuit to make sure it never saturates, like a Baker clamp, just wound around the opamp and with very small clamp voltage? (maybe adding a small bias) Regards Klaus
On Thursday, May 16, 2013 3:20:37 PM UTC-4, John Larkin wrote:

http://www.intersil.com/content/intersil/en/products/amplifiers-and-buffers=
/all-amplifiers/amplifiers/EL2157.html

"Connecting the load resistor to ground and operating from a single supply,=
 the outputs swing completely to ground without saturating. The outputs can=
 also drive to within 1.2V of the top rail. The EL2150 and EL2157 will outp=
ut =B1100mA and will operate with single supply voltages as low as 2.7V, ma=
king them ideal for portable, low power applications."
On Thu, 16 May 2013 14:09:02 -0700 (PDT), Klaus Kragelund
<klauskvik@hotmail.com> wrote:

>On Thursday, May 16, 2013 9:20:37 PM UTC+2, John Larkin wrote: >> I need to drive a mosfet in current-sink mode. It's similar to this >> >> circuit >> >> >> >> https://dl.dropboxusercontent.com/u/53724080/Circuits/Current_Sources/Mirror2.JPG >> >> >> >> but V- will be ground and I'll run the opamp off +8 or +10 or >> >> something, whatever drive the fet needs to sink 150 amps. >> >> >> >> I'd prefer to not have a negative power supply. At zero requested >> >> current, I want the current to be really zero, namely the fet gate at >> >> ground. I can futz in a little DC offset to guarantee that the opamp >> >> rails low. >> >> >> >> Opamps are often not spec'd for what happens when they rail, namely >> >> how slow and how ugly they come out of saturation. An internal >> >> compensation node can wind up. I'd like to get back in business in >> >> ballpark 1 us. >> >> >> >> Does anybody have favorite R-R opamps that come out of saturation fast >> >> and clean? >> >> >> >> I recall one Intersil chopamp that took several seconds to recover. >> > >LM324.... just kidding :-) > >Can't you add a clamping circuit to make sure it never saturates, like a Baker clamp, just wound around the opamp and with very small clamp voltage? (maybe adding a small bias) > >Regards > >Klaus
CMOS doesn't "saturate", i.e. there's no storage. Although compensation caps can "wind up", even in CMOS OpAmps. Solutions would depend on how fast John needs the current to change, and/or whether the object is to switch on and off a fixed current. ...Jim Thompson -- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.
On Thursday, May 16, 2013 3:20:37 PM UTC-4, John Larkin wrote:

Ummm... "However, special circuitry built into the
EL2150 and EL2157 allows the output to follow the input
signal to ground without recovery delays."

"Special circuitry" performance should be verified by the user.

They do spec the recovery from upper limit clamp at 7ns.
On Thu, 16 May 2013 14:09:02 -0700 (PDT), Klaus Kragelund
<klauskvik@hotmail.com> wrote:

>On Thursday, May 16, 2013 9:20:37 PM UTC+2, John Larkin wrote: >> I need to drive a mosfet in current-sink mode. It's similar to this >> >> circuit >> >> >> >> https://dl.dropboxusercontent.com/u/53724080/Circuits/Current_Sources/Mirror2.JPG >> >> >> >> but V- will be ground and I'll run the opamp off +8 or +10 or >> >> something, whatever drive the fet needs to sink 150 amps. >> >> >> >> I'd prefer to not have a negative power supply. At zero requested >> >> current, I want the current to be really zero, namely the fet gate at >> >> ground. I can futz in a little DC offset to guarantee that the opamp >> >> rails low. >> >> >> >> Opamps are often not spec'd for what happens when they rail, namely >> >> how slow and how ugly they come out of saturation. An internal >> >> compensation node can wind up. I'd like to get back in business in >> >> ballpark 1 us. >> >> >> >> Does anybody have favorite R-R opamps that come out of saturation fast >> >> and clean? >> >> >> >> I recall one Intersil chopamp that took several seconds to recover. >> > >LM324.... just kidding :-) > >Can't you add a clamping circuit to make sure it never saturates, like a Baker clamp, just wound around the opamp and with very small clamp voltage? (maybe adding a small bias) > >Regards > >Klaus
With a negative supply, I'd just hang a diode across (or instead of) the cap in my circuit. The opamp output would hang at -0.6 or whatever, still closed-loop. It takes more parts to do that without a negative supply, to hang at +0.3 or something just below the mosfet threshold. It's a parts count tradeoff against just making a negative supply voltage with a charge pump or something. -- John Larkin Highland Technology, Inc jlarkin at highlandtechnology dot com http://www.highlandtechnology.com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom laser drivers and controllers Photonics and fiberoptic TTL data links VME thermocouple, LVDT, synchro acquisition and simulation
On Thu, 16 May 2013 14:20:58 -0700 (PDT),
bloggs.fredbloggs.fred@gmail.com wrote:

>On Thursday, May 16, 2013 3:20:37 PM UTC-4, John Larkin wrote: > >http://www.intersil.com/content/intersil/en/products/amplifiers-and-buffers/all-amplifiers/amplifiers/EL2157.html > >"Connecting the load resistor to ground and operating from a single supply, the outputs swing completely to ground without saturating. The outputs can also drive to within 1.2V of the top rail. The EL2150 and EL2157 will output &#2013266097;100mA and will operate with single supply voltages as low as 2.7V, making them ideal for portable, low power applications."
Nice amp, but it seems to be "retired." -- John Larkin Highland Technology, Inc jlarkin at highlandtechnology dot com http://www.highlandtechnology.com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom laser drivers and controllers Photonics and fiberoptic TTL data links VME thermocouple, LVDT, synchro acquisition and simulation
On Thu, 16 May 2013 14:44:41 -0700, John Larkin
<jlarkin@highlandtechnology.com> wrote:

>On Thu, 16 May 2013 14:09:02 -0700 (PDT), Klaus Kragelund ><klauskvik@hotmail.com> wrote: > >>On Thursday, May 16, 2013 9:20:37 PM UTC+2, John Larkin wrote: >>> I need to drive a mosfet in current-sink mode. It's similar to this >>> >>> circuit >>> >>> >>> >>> https://dl.dropboxusercontent.com/u/53724080/Circuits/Current_Sources/Mirror2.JPG >>> >>> >>> >>> but V- will be ground and I'll run the opamp off +8 or +10 or >>> >>> something, whatever drive the fet needs to sink 150 amps. >>> >>> >>> >>> I'd prefer to not have a negative power supply. At zero requested >>> >>> current, I want the current to be really zero, namely the fet gate at >>> >>> ground. I can futz in a little DC offset to guarantee that the opamp >>> >>> rails low. >>> >>> >>> >>> Opamps are often not spec'd for what happens when they rail, namely >>> >>> how slow and how ugly they come out of saturation. An internal >>> >>> compensation node can wind up. I'd like to get back in business in >>> >>> ballpark 1 us. >>> >>> >>> >>> Does anybody have favorite R-R opamps that come out of saturation fast >>> >>> and clean? >>> >>> >>> >>> I recall one Intersil chopamp that took several seconds to recover. >>> >> >>LM324.... just kidding :-) >> >>Can't you add a clamping circuit to make sure it never saturates, like a Baker clamp, just wound around the opamp and with very small clamp voltage? (maybe adding a small bias) >> >>Regards >> >>Klaus > >With a negative supply, I'd just hang a diode across (or instead of) >the cap in my circuit. The opamp output would hang at -0.6 or >whatever, still closed-loop. It takes more parts to do that without a >negative supply, to hang at +0.3 or something just below the mosfet >threshold. It's a parts count tradeoff against just making a negative >supply voltage with a charge pump or something.
Doesn't the feed-forward compensation (for stability) guarantee that there will be a recovery time from off to on? ...Jim Thompson -- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.
John - take a look at the CMOS 4000 series "buffers". They looked ok on a 
test jig for driving 250 volt mosfets rated around 20 amps. If I remember 
correctly, the results weren't blitzkreig speedy but ok. A single mosfet
could be driven by all the "buffers" in the package, which might match the 
150 amp device you mentioned. Price is nice, too.

Hul


John Larkin <jlarkin@highlandtechnology.com> wrote:


> I need to drive a mosfet in current-sink mode. It's similar to this > circuit
> https://dl.dropboxusercontent.com/u/53724080/Circuits/Current_Sources/Mirror2.JPG
> but V- will be ground and I'll run the opamp off +8 or +10 or > something, whatever drive the fet needs to sink 150 amps.
> I'd prefer to not have a negative power supply. At zero requested > current, I want the current to be really zero, namely the fet gate at > ground. I can futz in a little DC offset to guarantee that the opamp > rails low.
> Opamps are often not spec'd for what happens when they rail, namely > how slow and how ugly they come out of saturation. An internal > compensation node can wind up. I'd like to get back in business in > ballpark 1 us.
> Does anybody have favorite R-R opamps that come out of saturation fast > and clean?
> I recall one Intersil chopamp that took several seconds to recover.
> --
> John Larkin Highland Technology, Inc
> jlarkin at highlandtechnology dot com > http://www.highlandtechnology.com
> Precision electronic instrumentation > Picosecond-resolution Digital Delay and Pulse generators > Custom laser drivers and controllers > Photonics and fiberoptic TTL data links > VME thermocouple, LVDT, synchro acquisition and simulation
On Thu, 16 May 2013 14:49:43 -0700, Jim Thompson
<To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:

>On Thu, 16 May 2013 14:44:41 -0700, John Larkin ><jlarkin@highlandtechnology.com> wrote: > >>On Thu, 16 May 2013 14:09:02 -0700 (PDT), Klaus Kragelund >><klauskvik@hotmail.com> wrote: >> >>>On Thursday, May 16, 2013 9:20:37 PM UTC+2, John Larkin wrote: >>>> I need to drive a mosfet in current-sink mode. It's similar to this >>>> >>>> circuit >>>> >>>> >>>> >>>> https://dl.dropboxusercontent.com/u/53724080/Circuits/Current_Sources/Mirror2.JPG >>>> >>>> >>>> >>>> but V- will be ground and I'll run the opamp off +8 or +10 or >>>> >>>> something, whatever drive the fet needs to sink 150 amps. >>>> >>>> >>>> >>>> I'd prefer to not have a negative power supply. At zero requested >>>> >>>> current, I want the current to be really zero, namely the fet gate at >>>> >>>> ground. I can futz in a little DC offset to guarantee that the opamp >>>> >>>> rails low. >>>> >>>> >>>> >>>> Opamps are often not spec'd for what happens when they rail, namely >>>> >>>> how slow and how ugly they come out of saturation. An internal >>>> >>>> compensation node can wind up. I'd like to get back in business in >>>> >>>> ballpark 1 us. >>>> >>>> >>>> >>>> Does anybody have favorite R-R opamps that come out of saturation fast >>>> >>>> and clean? >>>> >>>> >>>> >>>> I recall one Intersil chopamp that took several seconds to recover. >>>> >>> >>>LM324.... just kidding :-) >>> >>>Can't you add a clamping circuit to make sure it never saturates, like a Baker clamp, just wound around the opamp and with very small clamp voltage? (maybe adding a small bias) >>> >>>Regards >>> >>>Klaus >> >>With a negative supply, I'd just hang a diode across (or instead of) >>the cap in my circuit. The opamp output would hang at -0.6 or >>whatever, still closed-loop. It takes more parts to do that without a >>negative supply, to hang at +0.3 or something just below the mosfet >>threshold. It's a parts count tradeoff against just making a negative >>supply voltage with a charge pump or something. > >Doesn't the feed-forward compensation (for stability) guarantee that >there will be a recovery time from off to on? > > ...Jim Thompson
Then add the gate charge plateau. Interesting problem with so many interleaved issues. ...Jim Thompson -- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.