Author Topic: Voltage boosting an opamp  (Read 20415 times)

0 Members and 1 Guest are viewing this topic.

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Voltage boosting an opamp
« Reply #25 on: January 02, 2017, 02:44:25 am »
f5r5e5d, that is like an oscilloscope z-axis or horizontal amplifier.  The major difference is that they use a common emitter output configuration;  the inversion allows a small amount of AC coupling between the output of the error amplifier and the current source to produce a symmetrical transient response.

But an error amplifier using an operational amplifier is too slow to require or output cascodes or coupling to the current source making for a less complex circuit.
 

Online YansiTopic starter

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Voltage boosting an opamp
« Reply #26 on: January 02, 2017, 02:52:31 am »
Just a quick reply, the rest later:

Simple SE is a powerwaster. If I need to deliver up to 10mA, it has to run with 10 or more mA of the CCS. I like the previous solution more, due to not wasting power continuously. 1W is not a significant problem, but well... why? If it can be solved by adding a trannie or two. Even if they will be unbiased, the resulting crossover distortion is not a problem for my application, as I need mostly DC operation. I'd like to try lowering the idle consumption of the whole device, as it already is wasting a lot power in other power supplies.

About the load Z spec: As I mentioned earlier. See how I did the feedback loop for it to be stable at any C load. (or I think it will be stable). Couldn't find a better image: http://i.cmpnet.com/eetimes/news/08/09/1540chart_pg48.gif
But like that. I think it should make it stable into any sensible capacitive loading, isn't it? But I might be wrong.

Also, as it will be used as a general purpose control voltage supply, one can connect anything to it, so specifying the load is not that easy here.

BTW, what purpose does the R6 C2 serve in your schematic? Such small capacitance value there.
 

Offline f5r5e5d

  • Frequent Contributor
  • **
  • Posts: 349
Re: Voltage boosting an opamp
« Reply #27 on: January 02, 2017, 03:01:52 am »
moving the goal post?

you said you needed 1 ma, I give a circuit showing how to do that - with to-92 Q no heat sink


I wanted to show simple, don't think you can get much simpler than just 2 output Q without wasting even more power in bigger negative V with a pull down R

the folded cascode drive removes the sign inversion of the output stage allowing both loops negative feedback to sum at the op amp inverting input
« Last Edit: January 02, 2017, 03:05:28 am by f5r5e5d »
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Voltage boosting an opamp
« Reply #28 on: January 02, 2017, 12:50:49 pm »
I redesigned it, disregarding all the other schematics posted here. I believe this is the most basic approach.

It should be stable, without any additional compensation capacitor because op-amp has a gain of 10 and the booster gain stage has a gain of just under 3.2, thanks R4 & R5.

It can sink 10mA but only source around 1mA. If it doesn't need to sink so much current, then Q2 & D1 can be removed. If it needs to source more current, then an additional transistor is required.

R6 could be replaced with a current sink which would improve both the linearity and reduce the current consumption when the output voltage is higher.

 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Voltage boosting an opamp
« Reply #29 on: January 02, 2017, 01:50:10 pm »
Just a quick reply, the rest later:

Simple SE is a powerwaster. If I need to deliver up to 10mA, it has to run with 10 or more mA of the CCS. I like the previous solution more, due to not wasting power continuously. 1W is not a significant problem, but well... why? If it can be solved by adding a trannie or two. Even if they will be unbiased, the resulting crossover distortion is not a problem for my application, as I need mostly DC operation. I'd like to try lowering the idle consumption of the whole device, as it already is wasting a lot power in other power supplies.

About the load Z spec: As I mentioned earlier. See how I did the feedback loop for it to be stable at any C load. (or I think it will be stable). Couldn't find a better image: http://i.cmpnet.com/eetimes/news/08/09/1540chart_pg48.gif
But like that. I think it should make it stable into any sensible capacitive loading, isn't it? But I might be wrong.

Also, as it will be used as a general purpose control voltage supply, one can connect anything to it, so specifying the load is not that easy here.

BTW, what purpose does the R6 C2 serve in your schematic? Such small capacitance value there.


This will source more than 10mA and sink 10mA.

I've implemented the current sink to reduce the power consumption.

It will have crossover distortion though, which can be removed, at the expense of more components and power consumption.
« Last Edit: January 02, 2017, 01:52:00 pm by Hero999 »
 

Offline John Heath

  • Frequent Contributor
  • **
  • Posts: 444
  • Country: ca
  • 2B or not 2B
Re: Voltage boosting an opamp
« Reply #30 on: January 04, 2017, 02:00:01 pm »
Hello!

Just got me thinking... For some project I might need a precision DC voltage source controlled via a DAC. The problem is I need 4 quadrant operation - so opamp is the obvious solution. The voltage we are talking about might be say few +5V downto -50V, output current 1mA is more than enough, BW not critical, 1kHz very well sufficient.

The problem is, for +5-50V output I'd need an opamp designed for at least 60V supply operation. I know that some high voltage opamps do exist, but mostly expensive. I'd like to try to stick with jellybean parts for this, not afraid the component count will be a little higher. Only if that fails, I'll buy a high voltage opamp for this :-)

Do you have any suggestions, how to voltage boost the output of an opamp?

One of my first ideas was to ground the output of the opamp through a rather small impedance and then to use a pair of current mirrors in the opamp supply rails, to control the output voltage.
I have drawn my idea, please see below. Does it stand a chance to work?  I think it just might, but I may well be mistaken. 

Thank you for ideas as always!
Yansi

//EDIT: Just found this voltage booster (the second image below). Almost exact schematic! Thats pleasing I have reinvented the wheel and it might work just fine. Also interesting to note the local feedback through R1 C1.
Still I do not get the purpose of matching/pairing the trannies in here. What will the assymetry cause? I can't figure out.

I built a sub woofer for my computer based on your same design. However in my case the op-amp rails went to two darlington NPN / PNP outputs through a resister to base then resister to emitter with both collectors joined for output to speaker sub woofer. It ended up with + and - 30 volts and an average of 10 volts across the op-amp. For thermal protection you can place the op-amp close to the output heat-sink. This way bias temperature corrected current built into the op-amp will temperature correct the output darlingtons. It worker for a few years. I think I still have it around somewhere. In summary I had good luck with your design and found it could be simplified by resistively going from op-amp rails to darlington bases. An added benefit is it looks like it can not possibly work on paper with op-amp output going to ground and + input as negative feedback due to output stage inversion. Lots of fun showing off the diagram to onlookers scratching their heads , ha. 
 
The following users thanked this post: Someone

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Voltage boosting an opamp
« Reply #31 on: January 04, 2017, 04:51:26 pm »
I built a sub woofer for my computer based on your same design. However in my case the op-amp rails went to two darlington NPN / PNP outputs through a resister to base then resister to emitter with both collectors joined for output to speaker sub woofer. It ended up with + and - 30 volts and an average of 10 volts across the op-amp. For thermal protection you can place the op-amp close to the output heat-sink. This way bias temperature corrected current built into the op-amp will temperature correct the output darlingtons. It worker for a few years. I think I still have it around somewhere. In summary I had good luck with your design and found it could be simplified by resistively going from op-amp rails to darlington bases. An added benefit is it looks like it can not possibly work on paper with op-amp output going to ground and + input as negative feedback due to output stage inversion. Lots of fun showing off the diagram to onlookers scratching their heads , ha.
It becomes fairly obvious how it works, once it's pictured, showing the internal schematic of the op-amp.


It's been discussed before:
https://www.eevblog.com/forum/beginners/non-standard-op-amp-configuration/
 

Offline basinstreetdesign

  • Frequent Contributor
  • **
  • Posts: 466
  • Country: ca
Re: Voltage boosting an opamp
« Reply #32 on: January 04, 2017, 05:16:23 pm »
I have a suggestion:
You can boost the output range of a jelly-bean op-amp like the 741 by sliding the operating voltage of the amp up or down within a higher voltage than it is rated for.  See the first circuit below.  The output of the 741 determines what voltage is applied to its own power pins.  The resistor dividers connected to the transistors bases set the fraction of the total supply range is applied.  It can be shown mathematically that the voltage between the + and - pins of the op-amp never changes regardless of what the output voltage is.  The operating power for the amp just slides up and down between the rails as needed by the signal.

I have used this a couple of times and it works well for not-so-large bandwidths and reasonably low loads.  Don't push the drive specs of the op-amp.  The only worry is that the op-amp becomes slightly less stable at higher frequencies than normal and may oscillate if driving capacitive loads or long lines.  In that case stick a few hundred ohms at the output pin of the 741 and all is well.

The first attachment below shows this arrangement.  But if the combined voltage tolerance of the op-amp and two transistor is not enough then the transistors can be stacked to give even more range to the amp.  The second attachment shows this with a 741 operating with close to 115V output range.

One thing you cannot do with this arrangement is use it to get large gain at the same time as that might require the inputs to be a large voltage away from the output.
« Last Edit: January 04, 2017, 05:25:22 pm by basinstreetdesign »
STAND BACK!  I'm going to try SCIENCE!
 

Online YansiTopic starter

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Voltage boosting an opamp
« Reply #33 on: January 04, 2017, 08:06:53 pm »
I think I have already mentioned that and this supply boosting topology has some serious disadvantages, against any other topologies posted here. Please see this paper.

http://joebrown.org.uk/images/DualPSU/BootstrappingOpAmps.pdf

(Also note you have only a buffer, I need voltage amplifier.)
« Last Edit: January 04, 2017, 08:09:17 pm by Yansi »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Voltage boosting an opamp
« Reply #34 on: January 05, 2017, 02:18:06 am »
It becomes fairly obvious how it works, once it's pictured, showing the internal schematic of the op-amp.

It works even better if the output of the operational amplifier is used as a current feedback input to stabilize the voltage gain of the output stage.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Voltage boosting an opamp
« Reply #35 on: January 05, 2017, 07:44:36 pm »
It becomes fairly obvious how it works, once it's pictured, showing the internal schematic of the op-amp.

It works even better if the output of the operational amplifier is used as a current feedback input to stabilize the voltage gain of the output stage.
Like this?

 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Voltage boosting an opamp
« Reply #36 on: January 05, 2017, 11:35:03 pm »
It works even better if the output of the operational amplifier is used as a current feedback input to stabilize the voltage gain of the output stage.
Like this?

Yes, exactly like that.  Some implementations just use AC coupling (capacitor in series with R12) back into the output.  Even if you do not use feedback, the series resistor on the output (R13) should be used to trim the output stage frequency response.

The output stage feedback is especially important when the load impedance is not well defined making the voltage gain at the collectors of Q11 and Q12 vary.

A high voltage implementation adds either zener diodes or cascodes in series with the operational amplifier's supply pins.
 

Online YansiTopic starter

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Voltage boosting an opamp
« Reply #37 on: November 09, 2019, 03:38:21 pm »
After loooong loong time, I finally got back to work a bit on this project.

So this is what I came up with for a prototype to test:



The idea was to have at least a bit of bias flowing through the power stage - in this case 1mA.  That equals to about 0.7V across R3 (R4), requiring about 850uA of current through T1, T2.   Diodes D1 and D2 shall do just that, as about 1.2V is left across R14+R15 (resulting in about 850uA. Of course, it all depends on the exact voltage drops across all junctions, but I think as a ballpark figure it should work.

Resistor R7 was calculated such that under any circumstances it can produce a maximum of 2mA through T1 (T2), resulting in about 10mA maximum output current.

T1 and T2 are purposely biased from 0 and 5V, which is the same voltage as the opamp has, so that with a zero output voltage, the opamp output should sit just in the middle around 2.5V.

What do you think about that? Will it work or do you see there any major flaws? Thank you for any hints. If all goes well, I will test the circuit tomorrow.
 

Offline dom0

  • Super Contributor
  • ***
  • Posts: 1483
  • Country: 00
Re: Voltage boosting an opamp
« Reply #38 on: November 09, 2019, 04:47:20 pm »
Looks fine on first glance. Since the MCP6001 is easily outpaced by a melting glacier, you might even get away without C2. Circuits that put gain or extra poles into the feedback loop of an op amp will often need at least some tweaking in practice, outside the simulator. The faster the OP, the more tweaking.
« Last Edit: November 09, 2019, 04:52:56 pm by dom0 »
,
 

Online YansiTopic starter

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Voltage boosting an opamp
« Reply #39 on: November 09, 2019, 04:57:55 pm »
Haha, melting glacier... funny.   ;D

Fortunately in this case, I don't need more bandwidth than say 100Hz, so 6001 opamp will be I think fine in this application.

So let's have some fun doing a test PCB for it.

 

Offline dom0

  • Super Contributor
  • ***
  • Posts: 1483
  • Country: 00
Re: Voltage boosting an opamp
« Reply #40 on: November 09, 2019, 05:11:47 pm »
I would like to see the application note where it originated.  The oldest reference I have seen is in the Tektronix 7A29 vertical amplifier from 1979 where it was used with 741s.

I looked at the 7A29 and I believe you are referring to the current sources feeding the input limiting bridge, A2, the circuitry surrounding U100?

The idea of that circuit is imho a bit different from the power/voltage extenders, because it seems to be mostly used to work around the limited output swing of the 741, which wouldn't allow for the ~10-11 V or so at the node of U300 pin 3 if you add the output follower that it would need to achieve that. So you can't use a follower directly, so we either have to shift level from the OP's output to get to a follower or we use a common emitter amplifier. But that we can't control anyway with the OP, because it never reaches its rails. So shifting that through the power rail makes sense. RT109/R109/R107/R113 configure U111/Q114 for a gain of -0.75 or so for ~11 V at U100 pin 3, and U121/Q124 is just a straight inverter working off that node for -11 V at pin 13 of U100; diode bridge current is then set by the resistors internal to U100.

Hm, guess I do agree after all, same idea, just a bit of a different application.
,
 

Online YansiTopic starter

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Voltage boosting an opamp
« Reply #41 on: November 09, 2019, 05:51:01 pm »
Could you please post the part of Tek schematic you're talking about?

...meanwhile, I made the PCB. So let's clear the mess on my bench so I can solder it.

 

Offline dom0

  • Super Contributor
  • ***
  • Posts: 1483
  • Country: 00
Re: Voltage boosting an opamp
« Reply #42 on: November 09, 2019, 06:00:02 pm »
[ Specified attachment is not available ]
,
 

Online YansiTopic starter

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Voltage boosting an opamp
« Reply #43 on: November 09, 2019, 06:52:41 pm »
Now if it would just work at least a bit how it was intended...

Time got spared, so I am working like no tomorrow.  Testing within a moment.

//Unfortunately, I could not find any brand name BCP53. Of course, I found those from ONsemi right after I soldered this chinese junk on the pcb.

 

Online YansiTopic starter

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Voltage boosting an opamp
« Reply #44 on: November 09, 2019, 08:39:29 pm »
So, taking exactly the schematic in post #38 and testing it on a prototype pcb, yields  very promising results:

Bias current through the final stage settled to 1.25mA (I'd say close enough to my design goal of 1mA).
Even though a glacier melts faster than MCP6001T moves its ass, I get 140kHz (-3dB) bandwidth, test from full swing. (test pcb supplied with +-20V and +5V for the opamp). Which I think is quite impressive, considering no tweaking was done whatsoever.

So I think it will be good enough for 100Hz :)

//EDIT: I will test the output short circuit current later. I will also change the supply voltages to the desired +12 and -55V to see what it'll do.

//EDIT: Great! current limits to about 12.5mA. I have used a smaller R7, so das stimmt.
« Last Edit: November 09, 2019, 08:45:29 pm by Yansi »
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Voltage boosting an opamp
« Reply #45 on: November 09, 2019, 11:14:15 pm »
The buffer will improve the slew rate because it's much faster than the op-amp. The oo-amp's output only has to change a little for the buffer's output to change much more. Look at the op-amp's output voltage.
 

Online YansiTopic starter

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Voltage boosting an opamp
« Reply #46 on: November 10, 2019, 01:38:49 pm »
^ That of course makes sense.

Currently thinking about those 4 diodes in series, biasing the whole amplifier.

What would happen, if I'd replace those 4 diodes with a NPN Vbe multiplier (NPN with Rbe, Rcb resistors)? That would also give a bit more flexibility in selecting the bias voltage.

Just wondering, how could I get more stable biasing. Huh.  ???  What would be the dis/advantages of using 4 diodes vs. Vbe multiplier?
 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15797
  • Country: fr
Re: Voltage boosting an opamp
« Reply #47 on: November 10, 2019, 03:39:25 pm »
I know that some high voltage opamps do exist, but mostly expensive.

Define expensive. How many units are you going to build? Are a few dollars going to really matter? A couple references: OPA452, ADA4700-1...

And do you think the added parts, time to design, and extra PCB area for a discrete version will really yield a lower cost solution? Unless you go for large quantities, I even kinda doubt it.
 

Online YansiTopic starter

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Voltage boosting an opamp
« Reply #48 on: November 10, 2019, 03:50:49 pm »
Yes, OPA454 is exactly what I mean by EXPENSIVE NONSENSE.  OPA454 is about $7 + VAT at single quantities from Mouser.

MCP6001T is below $0.5 in single counts, and those 4 jellybean PNP/NPN transistors cost next to nothing.  So say about $2 maximum per discrete amplifier unit?

Does not pay off to use OPA454 even at single quantities. Every idiot can find a ready made solution given amount of $$$. Sure, time is valuable, but knowledge even more so.  I value knowledge more, than the time spent getting the knowledge.

Next time I will need a HV amplifier, I know sure how to make one, cheap. 

Now, don't you have anything better to do, then question the obvious from over 2 years old posts?

 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15797
  • Country: fr
Re: Voltage boosting an opamp
« Reply #49 on: November 10, 2019, 04:16:47 pm »
Learning things is interesting, but the choice here is absolutely NOT obvious. Is gaining say $5 (and again you didn't count the extra PCB area needed and time to design, so it's likely less than that) worth the trouble? The added part count also means a larger BOM, which does not always pay off either. If it floats your boat, fine, but saying using HV opamp is idiotic doesn't make sense in general.

And I didn't notice it was initially 2 years old, as there are a number of recent posts discussing it. You're apparently still working on it 2 years after judging from your above posts? If you still think it was worth the trouble, fine. Just saying.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf