Author Topic: Low ESR Output Capacitor for REF5010 and REF5025  (Read 4761 times)

0 Members and 1 Guest are viewing this topic.

Offline EPAIIITopic starter

  • Super Contributor
  • ***
  • Posts: 1165
  • Country: us
Low ESR Output Capacitor for REF5010 and REF5025
« on: July 06, 2022, 09:38:17 am »
I am getting back to a project that I had to put on hold around a year ago. I am trying to build a fairly accurate (0.05%) Voltage reference to use for calibrating analog meters. And I plan to use similar circuitry to produce a current reference of similar accuracy for the same purpose. In my discussion of this project here last year I more or less was convinced that the REF50xx series of Voltage references were a good choice. Perhaps the best choice after I was shown that they could be stacked to produce a chain that would give steps of 10 Volts and smaller steps of 2.5 Volts by adding several of the 2.5 Volt version. With them I can literally have from 2.5 to around 300 Volts at 2.5 Volt steps in one unit. More than enough to check and calibrate most analog meters.

So I am trying to put together a parts list and the REF50xx spec. sheet says that a 1 to 50 uF capacitor with low ESR should be used on the output. It mentions a 1 or 1.5 Ohm value. I assume that lower would be OK. Although the supply houses have excellent search facilities, the ESR does not seem to be among the parameters that come up early in the search process. You need to be down to around 1000 remaining parts before they list it and that slows the search down a lot.

I have dived into lists of aluminum electrolytic, tantalum, and multi layer ceramic capacitors and found that very few aluminum types have ESR specs. at all. Out of hundreds of thousands of types I found only one with an ESR in that range. Cross them out, I guess. But I am finding that tantalum capacitors do have ESRs in that range and many are even a lot lower. And the multi layer ceramics just say they have a low or very low ESR but don't seem to give values. Both have suitable Voltage ratings (16 to 35 Volts in my searches) The prices on both types seem affordable.

So I ask, which should I pick, tantalum or multi layer ceramics? Or should I expand my search?

As always, thanks in advance for any help you can provide.
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 

Offline tom66

  • Super Contributor
  • ***
  • Posts: 7334
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #1 on: July 06, 2022, 10:25:26 am »
Nothing stops you putting a small series resistor inline with an otherwise inadequate capacitor choice.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #2 on: July 06, 2022, 02:35:28 pm »
The TI datasheet says that the output capacitor ESR must be lower than 1.5 ohms for stability, but lower than 1 ohm will increase noise.  Small tantalum capacitors meet this requirement but an electrolytic capacitor will need to be 10 microfarads or higher.  I suspect a ceramic or film capacitor will cause problems unless a series resistor is added.
 

Offline EPAIIITopic starter

  • Super Contributor
  • ***
  • Posts: 1165
  • Country: us
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #3 on: July 07, 2022, 04:22:17 am »
Seems like I need to read the data sheet again. So between 1 Ohm and 1.5 Ohms. That looks like a fairly narrow range. IIRC, it does seem to show a board layout with a series resistor.

I guess the biggest question in my mind at the present time is just what is the ESR of a multi layer ceramic. What do they mean by "low" or "very low"? They seem to include the low ESR in the bullet list of advantages of these capacitors, but list no actual values. I find that rather strange. Perhaps I just need to check more manufacturers' spec. sheets.
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 

Offline Someone

  • Super Contributor
  • ***
  • Posts: 5155
  • Country: au
    • send complaints here
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #4 on: July 07, 2022, 05:34:26 am »
what is the ESR of a multi layer ceramic. What do they mean by "low" or "very low"? They seem to include the low ESR in the bullet list of advantages of these capacitors, but list no actual values. I find that rather strange. Perhaps I just need to check more manufacturers' spec. sheets.
You need to go past the "family" datasheet/catalogue (which is mostly to show the production parameters) all the major ceramic capacitor manufactures have huge databases of part specific data.
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7453
  • Country: pl
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #5 on: July 07, 2022, 04:56:27 pm »
How exactly is noise supposed to increase with too low ESR? ???

Did they mean "borderline stability and gain peaking at some particular frequency"? BTW, that could still be acceptable for calibrating analog meters, perhaps.

Of course, can't go wrong using MLCC and 1Ω in series.
 

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 9334
  • Country: fi
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #6 on: July 07, 2022, 05:09:54 pm »
Note that there is "low ESR", and then there is "low ESR".

For some, 1 ohms is low ESR, for others, 0.01 ohms is low ESR.

From MLCC perspective, "low-ESR" aluminum electrolytics are high-ESR parts.

Ignoring some special series, MLCCs are close to zero ESR. Like tens of milliohms. Often this is specified in the datasheet.

If datasheets just talk about "low ESR" but do not explicitly mention "stable with MLCC" or "stable with 0.020 ohm ESR", I would be careful and not use MLCC (without series resistance).

Tantalums actually come with specified and well controlled ESR. This even appears in distributor parametric search! Tantalums also come with stable capacitance value, unlike X7R/similar MLCCs. So tantalum is often a decent choice, you can shop whether you want 0.5 or 1 or 2 or 5 ohms, and can get +/-10% capacitance tolerance!

Aluminum electrolytics are out, I'd say, for such high-tech project. They are fine for input bulk capacitance where exact value of C and ESR do not matter at all. Especially the ESR is all over the place with temperature and aging, with order of magnitude of uncertainty.

I'd probably design in a tantalum, this is what I do with small linear regulators requiring certain range of ESR or certain amount of C. Linear regulators and references have limited output current, also making the inrush problem disappear. Do not use tantalum on the input side without knowing what you are doing. Also remember bare minimum of 40% voltage derating, i.e., 10V part is good to 6V.


« Last Edit: July 07, 2022, 05:12:00 pm by Siwastaja »
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #7 on: July 07, 2022, 05:15:54 pm »
Also mind the ceramic multilayers lose their capacitance with DC bias voltage - up to 80% decrease at their max DC bias voltage..
« Last Edit: July 07, 2022, 06:09:12 pm by imo »
Readers discretion is advised..
 

Offline EPAIIITopic starter

  • Super Contributor
  • ***
  • Posts: 1165
  • Country: us
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #8 on: July 08, 2022, 09:47:50 am »
Well ain't that something! Seems to make the tantalums look better and better. Thanks for that tip.

I wonder what the physical mechanism behind that is.

And magic asked, "How exactly is noise supposed to increase with too low ESR?" I have no idea. The
TI data sheet says it, but there is no explanation. And their block diagram of how it works is really very basic, no details. "Proprietary technology", their words, not mine. Perhaps they explain it somewhere else.



Also mind the ceramic multilayers lose their capacitance with DC bias voltage - up to 80% decrease at their max DC bias voltage..
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 

Offline tom66

  • Super Contributor
  • ***
  • Posts: 7334
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #9 on: July 08, 2022, 10:38:19 am »
Don't forget tantalums have deteriorated maximum operating voltage with temperature.

At room temp full bias is OK.  At 85C typically 30-50% reduction must be observed.

This does not apply to ceramic or electrolytic, although ceramic will drift in capacitance with temperature and bias. 
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7453
  • Country: pl
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #10 on: July 08, 2022, 12:21:24 pm »
And magic asked, "How exactly is noise supposed to increase with too low ESR?" I have no idea. The
TI data sheet says it, but there is no explanation. And their block diagram of how it works is really very basic, no details. "Proprietary technology", their words, not mine. Perhaps they explain it somewhere else.
The block diagram is typical for that sort of IC and a few other types that have been torn down showed good match with their datasheet block diagrams. Hence, it is likely correct.

I suspect gain peaking, i.e. the scaling opamp is barely stable when driving ultra low ESR and amplifies reference noise more than normally at some high frequency (many kHz). In such case, low frequency noise wouldn't be affected. You would probably see ringing after load transients too ::)
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #11 on: July 08, 2022, 07:26:10 pm »
I guess the biggest question in my mind at the present time is just what is the ESR of a multi layer ceramic.

All ceramic and plastic film capacitors have an ESR of essentially zero.

Quote
What do they mean by "low" or "very low"? They seem to include the low ESR in the bullet list of advantages of these capacitors, but list no actual values. I find that rather strange. Perhaps I just need to check more manufacturers' spec. sheets.

Aluminum electrolytic and solid tantalum capacitors are the standard and come in standard and low ESR versions, although many old parts would be considered low ESR by modern standards.  Ceramic and film have always had essentially zero ESR.

Where this has changed is that polymer aluminum electrolytic and polymer capacitors have become available which fit between "low" ESR parts and "zero" ESR parts.
« Last Edit: July 08, 2022, 07:28:04 pm by David Hess »
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 15154
  • Country: de
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #12 on: July 09, 2022, 09:04:39 am »
Film type capacitors usually have very low ESR. For the X7R and similar MLCCs it depends on how the ESR is actually defiened.  The model with ideal capacitor and series resistor is not very accurate. So it depends on how it is measrured / relevant or the circuit. X7R is low ESR, but still relatively high loss, while the film capacitors may have slightly higher ESR, but usually lower loss. So the capacitance and ESR alone is not suficient to describe the capacitor.

An amplfier or votlage regulator usually does not like a large very low ESR capacitor at the output, as the amplifier's own output impedance is often inductive, at least for the higher frequencies. With a low loss capacitors this gives weakly damped ringing. A combination of some low ESR capacitance (e.g. 100 nF X7R) and some lossy capacitance (Al-electrolytic) may be more suitable.
 

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 9334
  • Country: fi
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #13 on: July 09, 2022, 09:06:54 am »
So it depends on how it is measrured / relevant or the circuit. X7R is low ESR, but still relatively high loss, while the film capacitors may have slightly higher ESR, but usually lower loss. So the capacitance and ESR alone is not suficient to describe the capacitor.

What is this "loss" you are talking about, and how do you define it?
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 15154
  • Country: de
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #14 on: July 09, 2022, 09:28:42 am »
A major loss mechinsm in capacitors is dielectric relaxation: that is the dielectric takes some time to react to the electric flield. For the X7R capacitors the mechnism is likely to a large part from obstacles in the domain movement of ferroelectrc parts. Another part is local intermediate conductivity that can bypass parts of the dielectric. This results in a frequency dependent capacity and a phase shift between voltage and current that is slightly smaller than the ideal 90 degrees.
The simple ESR model gives loss proportional to the frequency, but this is not a good description for all capacitors.
The more detailed model circuit for a real capacitor is including a sereies of RC elements in parallel to the main capacitor. As the losses are frequency dependent it is not just 1 element, but ideally many, like 1 for every decade in the frequency. X7R adds additional complications with a nonlinear capacitance.

Capacitor data-sheets may give the dissipation factor for some frequencies (depends on the use and capacitor size) to describe the loss - though in quite some cases this is only and upper limit, possibly even just the limit of the test instrument. Despite the reltively low ERS the X7R capacitors have quite high dissipation factor over a large frequency range.
 
The following users thanked this post: tom66

Offline magic

  • Super Contributor
  • ***
  • Posts: 7453
  • Country: pl
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #15 on: July 09, 2022, 10:29:17 am »
By the way, it seems that you intend to stack 120 of those chips.

Are you sure you don't want to use a single 2.5V reference chip and 120 identical 0.02% resistors to scale its output up to 300V in the same discrete steps?
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 15154
  • Country: de
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #16 on: July 09, 2022, 01:17:09 pm »
By the way, it seems that you intend to stack 120 of those chips.

Are you sure you don't want to use a single 2.5V reference chip and 120 identical 0.02% resistors to scale its output up to 300V in the same discrete steps?
I abosolutely support the idea of single reference and than stack resistors. Though I would more like start with 10 V and than use only 30 Resistors for the coarse part an a sperate fine part at one end.
The idea with stacking refrence is mainly useful if one wants to measure them individually and this way get a rather accurate step from the sum.
 

Offline EPAIIITopic starter

  • Super Contributor
  • ***
  • Posts: 1165
  • Country: us
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #17 on: July 12, 2022, 09:51:14 am »
120 of them stacked? Er, NO! That would cost around $1K and for that much I would rather try to find a used standard that I could repair and get calibrated. NO!

Thirty (30) of the 10 Volt reference chips would get to 300 V. And an additional four of the 2.5 V ones would allow selections in 2.5 V steps all the way up. And, with around 30 of the REF chips in series, that would cost a few hundred. Actually I will probably need to top it off at a bit under 300 V due to the available transformers at reasonable prices.

But I do not understand how one REF5010 plus a bunch of resistors will get me up to several hundred Volts. A Voltage divider could divide the 10 Volts down to, perhaps 0.1 V steps, but I would want to go the other way and multiplying the 10 Volts up, not divide it down. Are you talking about some kind of amplifier circuit with some other IC? An op amp, perhaps? One that allows 300 V supply rail? I do not know of such a chip. Discrete output stage added to an op amp? I can smell smoke now.



By the way, it seems that you intend to stack 120 of those chips.

Are you sure you don't want to use a single 2.5V reference chip and 120 identical 0.02% resistors to scale its output up to 300V in the same discrete steps?
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 

Offline tom66

  • Super Contributor
  • ***
  • Posts: 7334
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #18 on: July 12, 2022, 10:00:16 am »
Why not just use precision resistors and an op-amp and power electronics circuit (or just a high-voltage opamp)?  That way you only need one reference.
 

Offline EPAIIITopic starter

  • Super Contributor
  • ***
  • Posts: 1165
  • Country: us
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #19 on: July 12, 2022, 10:04:21 am »
The idea is to have a moderately accurate Voltage reference that has steps which would be useful for checking and calibrating analog meters. I would like to be able to check at least one scale, perhaps a 100 V or a 250 V one at several points (10%, 20%, or 25% steps) and also check the other scales at least at one point, as high as possible in the needle's deflection. I would probably add a divider to get down to a 1 V level and perhaps to a 0.1V level.

I hope to accomplish this without access to a reference meter that would allow the circuit to be calibrated. I presently do not have such a meter.

Perhaps I dream.



By the way, it seems that you intend to stack 120 of those chips.

Are you sure you don't want to use a single 2.5V reference chip and 120 identical 0.02% resistors to scale its output up to 300V in the same discrete steps?
I abosolutely support the idea of single reference and than stack resistors. Though I would more like start with 10 V and than use only 30 Resistors for the coarse part an a sperate fine part at one end.
The idea with stacking refrence is mainly useful if one wants to measure them individually and this way get a rather accurate step from the sum.
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 

Offline EPAIIITopic starter

  • Super Contributor
  • ***
  • Posts: 1165
  • Country: us
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #20 on: July 12, 2022, 10:18:03 am »
A quick search did not yield any 300 Volt op amps. At least not such chips. Perhaps a discrete circuit? Or an op amp with a Voltage boost output circuit?

Any suggestions where I should look?



Why not just use precision resistors and an op-amp and power electronics circuit (or just a high-voltage opamp)?  That way you only need one reference.
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 

Offline tom66

  • Super Contributor
  • ***
  • Posts: 7334
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #21 on: July 12, 2022, 12:48:00 pm »
This is available:

https://www.top-electronics.com/en/power-op-amp-900v-500v-us-100ma

Though its parameters are probably somewhat compromised given it is monolithic, and it is not cheap.

An alternative approach would be to build a low voltage opamp circuit with a feedback resistor network (which can be precisely set using e.g. 0.1% tolerance parts matched to each other) and either a DAC/PWM/potentiometer to set the feedback voltage from your precision reference.

The opamp would drive a form of high voltage amplifier.  A common approach is two common emitter circuits in series, with high voltage NPNs.  This adds phase delay to the circuit, so you need to compensate the opamp appropriately.

However complex this sounds, I doubt it is more complex than tying 30 odd Vref's in series with each other!
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #22 on: July 12, 2022, 07:57:08 pm »
Why not just use precision resistors and an op-amp and power electronics circuit (or just a high-voltage opamp)?  That way you only need one reference.

That is possible but presents calibration issues compared to stacked references.

Stacked references can be checked individually at low voltage.  A precision high voltage amplifier depends on its high voltage feedback divider for an accurate output, and checking the accuracy of a high voltage divider is not easy.  A low voltage test of the divider is not sufficient because at high voltages, the voltage coefficient of resistance becomes a factor and that can only be tested at high voltage.

The issues with a high voltage feedback divider can be overcome but require specific attention.  What I might do is grade a bunch of the same resistor for precision, and then stack them to make the divider.  Since all of the resistors will have the same voltage across them, the voltage coefficient of resistance will be minimized, and cancel out if they are graded properly.  There are of course also ready made high voltage dividers, but they may lack the needed precision.

A quick search did not yield any 300 Volt op amps. At least not such chips. Perhaps a discrete circuit? Or an op amp with a Voltage boost output circuit?

Any suggestions where I should look?

Apex Microtechnology and Burr-Brown were good sources for high voltage operational amplifiers.  I have a PA42 on my desk which is good to a supply voltage of 350 volts but a high voltage operational amplifier is not a panacea because they are not usually designed with the kind of precision that you want, so trimming or offset correction would be required anyway.

A discrete voltage boosted operational amplifier would allow the use of a precision operational amplifier so no trimming would be required, but requires some design work for either the bootstrapping, level shifting, or high voltage gain stage.

If the high voltage output is over a very limited range, like only 300 volts, then a fixed level shifter can be used.  Tektronix did this in their 50 and 100 volt regulators.  The operational amplifier operated at +/-15 volts or whatever, and had a 50 or 100 volt zener diode on its output to "add" 50 or 100 volts to its low voltage output.  This has the virtue of providing a gain of "1" inside of the loop despite the increase in voltage so frequency compensation is simplified.

Otherwise either the operational amplifier is bootstrapped so that its supply voltage follows the output, or a high voltage gain stage is added within the feedback loop.  Either way works.

Note that all of the methods described above require care to avoid exceeding the common mode and differential input voltage range of the operational amplifier.  This is particularly the case when a bootstrap configuration is used.

I would try the (poorly drawn) circuit shown below, extended to 300 volt operation of course.  R6 and R7 are the precision divider.
« Last Edit: July 12, 2022, 07:59:35 pm by David Hess »
 
The following users thanked this post: EPAIII

Online mawyatt

  • Super Contributor
  • ***
  • Posts: 4119
  • Country: us
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #23 on: July 12, 2022, 08:07:21 pm »
A quick search did not yield any 300 Volt op amps. At least not such chips. Perhaps a discrete circuit? Or an op amp with a Voltage boost output circuit?

Any suggestions where I should look?



Why not just use precision resistors and an op-amp and power electronics circuit (or just a high-voltage opamp)?  That way you only need one reference.

APEX PA441 or 443, 350V op-amps might fit your needs. We evaluated the PA443 awhile back, pretty good HV op-amp, however ended up selecting the TI OPA-462 instead for various other reasons.

Here's a evaluation HV AWG Buffer Amp with PA443 we developed awhile back.

Best,
« Last Edit: July 13, 2022, 12:23:06 am by mawyatt »
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 

Offline EPAIIITopic starter

  • Super Contributor
  • ***
  • Posts: 1165
  • Country: us
Re: Low ESR Output Capacitor for REF5010 and REF5025
« Reply #24 on: July 13, 2022, 12:15:30 am »
Thanks for that link. As I said, my search in the early morning hours was brief. I am sure there are others, perhaps not many, but others.

The latest posts have me seriously thinking about this approach. The REF chips are +/-0.05%. The 0.1% resistors suggested are above that level so that alone gives an accuracy hit. I would prefer 0.01% resistors or better, but then the price goes up. And I guess I can sort through a large batch of them with a meter of lesser accuracy because the goal would be matching values, not absolute ones.

Then there is the question of switching. If i keep the numerous Voltage steps that I presently envision, then a switching arrangement with around 30 positions would be needed. That is going to be a limitation if a single switch is to be employed. Multiple switches? I presently envision multiple, miniature, gold plated banana jacks, one for each output value. And yes, I know that's a bunch of banana jacks and $$$$. Perhaps something similar would work with a jumper cable. But, in usage, care would be needed so that a high Voltage is not produced while switching as that could damage a meter connected to the output which is set to a low Voltage scale.

And speaking of cost, the hi Voltage op amp in your link is expensive, but not so much so as 30+ REF chips. I need to see how it's cost compares with the needed resistors. And then, do I buy just one or do I get a spare? Will it still be available next year, in two years, in five years, etc. I can afford a few extra REF chips for maintenance spares in the event they are discontinued.

Back to switching those resistors. Some years ago I came up with a circuit for a decade resistance circuit that allowed it to be used as a decade pot. I believe I used twelve matched resistors for each decade and two wafers on a rotary switch which effectively jumped across two of those twelve with the next, lower decade. The problem was that each successive decade needed resistors that were 0.2X those of the previous one so only one decade could use 10xxx resistor values. Then 20xx and 40x ones would be needed (example 12 each: 10.00K, 0.01%; 2.0K, 0.1%; and then 400, 1%). But will that work with an op amp? I think yes, but it should be tested.



Why not just use precision resistors and an op-amp and power electronics circuit (or just a high-voltage opamp)?  That way you only need one reference.

That is possible but presents calibration issues compared to stacked references.

Stacked references can be checked individually at low voltage.  A precision high voltage amplifier depends on its high voltage feedback divider for an accurate output, and checking the accuracy of a high voltage divider is not easy.  A low voltage test of the divider is not sufficient because at high voltages, the voltage coefficient of resistance becomes a factor and that can only be tested at high voltage.

The issues with a high voltage feedback divider can be overcome but require specific attention.  What I might do is grade a bunch of the same resistor for precision, and then stack them to make the divider.  Since all of the resistors will have the same voltage across them, the voltage coefficient of resistance will be minimized, and cancel out if they are graded properly.  There are of course also ready made high voltage dividers, but they may lack the needed precision.

A quick search did not yield any 300 Volt op amps. At least not such chips. Perhaps a discrete circuit? Or an op amp with a Voltage boost output circuit?

Any suggestions where I should look?

Apex Microtechnology and Burr-Brown were good sources for high voltage operational amplifiers.  I have a PA42 on my desk which is good to a supply voltage of 350 volts but a high voltage operational amplifier is not a panacea because they are not usually designed with the kind of precision that you want, so trimming or offset correction would be required anyway.

A discrete voltage boosted operational amplifier would allow the use of a precision operational amplifier so no trimming would be required, but requires some design work for either the bootstrapping, level shifting, or high voltage gain stage.

If the high voltage output is over a very limited range, like only 300 volts, then a fixed level shifter can be used.  Tektronix did this in their 50 and 100 volt regulators.  The operational amplifier operated at +/-15 volts or whatever, and had a 50 or 100 volt zener diode on its output to "add" 50 or 100 volts to its low voltage output.  This has the virtue of providing a gain of "1" inside of the loop despite the increase in voltage so frequency compensation is simplified.

Otherwise either the operational amplifier is bootstrapped so that its supply voltage follows the output, or a high voltage gain stage is added within the feedback loop.  Either way works.

Note that all of the methods described above require care to avoid exceeding the common mode and differential input voltage range of the operational amplifier.  This is particularly the case when a bootstrap configuration is used.

I would try the (poorly drawn) circuit shown below, extended to 300 volt operation of course.  R6 and R7 are the precision divider.
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf