Author Topic: High voltage safety and design considerations  (Read 5453 times)

0 Members and 1 Guest are viewing this topic.

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
High voltage safety and design considerations
« on: June 26, 2019, 07:13:41 am »
Hello
I have to build a current source that drives +-25mA through a coil with a resistance of 1.5k\$\Omega\$ and inductance of 2H. Since the resistance and inductance are huge, I have to use a high voltage op amp to get a quick response. The best choice was the ADHV4702-1 which can do +-110V (basically rectified mains). I haven't designed anything with high voltages before, so I'm not sure what safety considerations I should use. I'm putting 40mA fuses on both rails, directly at the input but I'm not sure if that is enough.

Also since it's used in a robot, I'm not sure what power source to use. It can't be mains since it has to move and can't be plugged in, but I couldn't find a battery with a high voltage. Maybe a boost converter would be the best? I don't have experiance with designing SMPS so it would be nice to use something else if possible.

I attached a simulation from LTspice showing the overall design. The only change is going with 5v control logic, rather than 3.3v but that's not relevant here.
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7453
  • Country: pl
Re: High voltage safety and design considerations
« Reply #1 on: June 26, 2019, 08:31:58 am »
Rectified mains is 350V peak or more if you account for some mild overvoltage. 700V rail-to-rail for bipolar rectification :scared:
But it doesn't matter if you don't want mains anyway.

You could avoid bipolar supply by using H-bridge. Then you only need one HV rail.

Mind Safe Operating Area of the output transistors. Not every transistor can conduct 25mA while having a relatively high voltage on its collector/drain.

Do you really need serious precision? Perhaps a single resistor cascoded by a BJT/FET would suffice? AKA the "one transistor constant current source".

The opamp doesn't need to be HV if you take an LV opamp and add some circuitry to shift levels and drive HV transistors.
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: High voltage safety and design considerations
« Reply #2 on: June 26, 2019, 09:05:25 am »
It's an actuator like a voice coil, so using an H bridge slows down the transition, I would like to use the bipolar approach as is. The precision is also important, I'll be using a 16 bit dac.

The last idea seems nice, but I'm not sure how to level shift with transistors  :-// the buffer I used is the only one I have seen and it requires the op amp voltage to be the same as the driving transistors in order to have a fast response.

Edit: the transistors I plan on using are:
BUL216
MJE5852G
They have enough DC SOA and high enough CE and BC voltage tolerance
« Last Edit: June 26, 2019, 09:08:04 am by OM222O »
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: High voltage safety and design considerations
« Reply #3 on: June 26, 2019, 09:14:45 am »
The class B output stage you have there, will have lots of crossover distortion, so no good for precision.

You're mistaken, an h-bridge will not slow down the transition. Why do think it will? A bridged configuration is used in audio amplifiers all the time and works very well. A bridge configuration should theoretically be twice as fast as single ended, because the voltage swing across the load is double the output of each amplifier, so the slew rate will also be doubled. For example, if each op-amp and output stage has a maximum slew rate of 1V/µs, the slew rate of the voltage across the load can reach double that at 2V/µs, because the positive side can be slewing at 1V/µs and the negative side at -1V/µs, giving 2V/µs across the load.

The supply voltage can be reduced and the power dissipated by the individual transistors can be reduced by using a bridged configuration. Here's an example.
https://www.eevblog.com/forum/projects/need-help-with-bi-directional-constant-current-source-(100ma)
« Last Edit: June 26, 2019, 09:34:33 am by Zero999 »
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: High voltage safety and design considerations
« Reply #4 on: June 26, 2019, 09:49:40 am »
I thought an H bridge similar to what is used to drive DC motors, where at 0mA, the opposing transistors turn on and connect the load in the opposite direction, that would slow things down when passing through 0 current.
I used R5 to take care of the crossover distortion by giving op amp direct control when Vbe is less than 0.7v
Sorry but I haven't designed audio amllifiers which seems to be what you are refering to.
I haven't seen that output model before. Can you please explain how it
works? For example what are R6 and R7 for? Where is the shunt resistor? Or maybe a link to a page where they detail this. Also does the op amp need to be high voltage with this config?
« Last Edit: June 26, 2019, 10:03:32 am by OM222O »
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: High voltage safety and design considerations
« Reply #5 on: June 26, 2019, 10:01:59 am »
His country flag shows USA.
No, the original poster's country flag is clearly the UK.

I thought an H bridge similar to what is used to drive DC motors, where at 0mA, the opposing transistors turn on and connect the load in the opposite direction, that would slow things down when passing through 0 current. Sorry but I haven't designed audio amllifiers which seems to be what you are refering to.
I haven't seen that output model before. Can you please explain how it
works? For example what are R6 and R7 for? Where is the shunt resistor? Or maybe a link to a page where they detail this. Also does the op amp need to be high voltage with this config?
The issue you're talking about with the driver working near 0mA, applies to both half and full bridge configurations: look up crossover distortion. A motor driver H-bridge uses PWM so is class D, rather than class B or AB in this application.

The circuit I've posted is a Howland current pump. A fairly good analysis can be found in application note linked below.
http://www.ti.com/lit/an/snoa474a/snoa474a.pdf

The thread linked to below describes how the transistors boost the op-amp's output current.
https://www.eevblog.com/forum/beginners/non-standard-op-amp-configuration

Yes, the op-amp does need to be a high voltage type to work in the circuit I've posted. There are ways to do this using a low voltage op-amp and a booster stage, but they're more tricky. See links below.
http://www.ti.com/lit/an/snoa600b/snoa600b.pdf
https://www.analog.com/media/en/technical-documentation/application-notes/an18f.pdf
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7453
  • Country: pl
Re: High voltage safety and design considerations
« Reply #6 on: June 26, 2019, 10:07:28 am »
I think an important question is whether you need only +25mA and -25mA or the possibility to set anything in between at any time.

His country flag shows USA.
I see UK, which runs on 240V IIRC.
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: High voltage safety and design considerations
« Reply #7 on: June 26, 2019, 10:09:54 am »
So even though R5 reduces the noise significantly, you think still there will be enough noise to cause issues? The simulation doesn't use ideal models, so I assumed when there isn't much noise on the current trace, the real world circuit also acts fine.

I think an important question is whether you need only +25mA and -25mA or the possibility to set anything in between at any time.

His country flag shows USA.
I see UK, which runs on 240V IIRC.

As I mentioned it's used to drive a "voice coil" (a big actuator) in a robot, so I do need values in between. UK mains is also 240V rms  :-DD I meant US mains as someone else pointed out, but that was just a comparison, not a real power supply solution.

I also had a look at voltage booster stages for op amps ... just wow! So many parts are required. Usually 3 to 4 complementary stages are used with a ton of resistors and diodes! There also doesn't seem to be a chip that does that. I think I'll stick with the ADHV for now.

Any suggestions regarding obtaining +-110V and what are safety issues that I should be concerned with high voltage DC applications?
« Last Edit: June 26, 2019, 11:49:20 am by OM222O »
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: High voltage safety and design considerations
« Reply #8 on: June 26, 2019, 04:32:34 pm »
Apex Microtechnology makes a number of high voltage drivers for applications like this and have a number of app notes well worth reading: https://www.apexanalog.com/support/app_note_library.html

I built some HV drivers for an electro-optical test bench using some Apex hybrids and some Vicor DC-DC converters to provide the +- DC supply.  I decided to run the test jig subassemblies off of 24 VDC rather than mains voltage to simplify power distribution and switching.
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: High voltage safety and design considerations
« Reply #9 on: June 26, 2019, 07:25:05 pm »
One thing that I found was using a forward converter with a center tap transformer. It seems like a nice and simple idea for boosting to high voltage with low current and they have high effeciency too. Are there any articles that help with designing a dc-dc forward converter?
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: High voltage safety and design considerations
« Reply #10 on: June 27, 2019, 09:01:13 pm »
LTSpice doesn't add noise to plots, using the transient response mode. A separate noise simulation is required, which needs the models to reflect reality.

What slew rate, in terms of Amps per second, do you need? Does it need to work down to DC? Can you drive it via a transformer?
 

Offline Someone

  • Super Contributor
  • ***
  • Posts: 5155
  • Country: au
    • send complaints here
Re: High voltage safety and design considerations
« Reply #11 on: June 28, 2019, 04:33:38 am »
Any suggestions regarding obtaining +-110V and what are safety issues that I should be concerned with high voltage DC applications?
As above, seriously consider bridged drive so you only need a single sided supply. Motor drives are somewhat similar and show examples of how to put current measurement in the feedback path all at happy low voltages. For a one off its usually cheaper to just buy off the shelf products, by the time you make a few mistakes and buy extra parts its easy to run through a budget.

100-150V at 25mA is just 5W, there are many suppliers of DC-DC supplies that will solve that, only consider building your own power supply if you plan to make 10,000 of them.  Safety is mostly about preventing people coming into contact with the 100V by adding appropriate insulation (all the way up to and possibly including the actuator).
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: High voltage safety and design considerations
« Reply #12 on: June 28, 2019, 08:55:15 pm »
We finally settled on 3 7s and one 4s LiPo packs for a voltage range of 105-80V. This will be doubled for the + and - rails. The problem with the birdge approach is the high part numbers, especially needing 2 op amps. The ADHV costs about 15$ a pop and I need 12 copies of that circuit. I think using batteries will be safe since they will be floating with respect to ground and low current fuses will prevent shorts or other issues from causing any major damage.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: High voltage safety and design considerations
« Reply #13 on: June 28, 2019, 10:35:33 pm »
We finally settled on 3 7s and one 4s LiPo packs for a voltage range of 105-80V. This will be doubled for the + and - rails. The problem with the birdge approach is the high part numbers, especially needing 2 op amps. The ADHV costs about 15$ a pop and I need 12 copies of that circuit. I think using batteries will be safe since they will be floating with respect to ground and low current fuses will prevent shorts or other issues from causing any major damage.
Then don't use the ADHV. Use ordinary, cheap op-amps, run off <30V and some discrete devices to amplify the output voltage.

Also note that it's very hard to protect semiconductors with fuses, normally the semiconductor blows first. Check that the fuse is rated for use with DC and that it has sufficient breaking capacity.
 

Offline Someone

  • Super Contributor
  • ***
  • Posts: 5155
  • Country: au
    • send complaints here
Re: High voltage safety and design considerations
« Reply #14 on: June 29, 2019, 12:59:32 am »
I think using batteries will be safe since they will be floating with respect to ground and low current fuses will prevent shorts or other issues from causing any major damage.
Until you get a short on their unfused side and then have 1kW or more fault to deal with. Batteries are much less safe than a current limited power supply here, you haven't discussed isolating the control signals or the actuator so the 100V supply will not be floating.

The problem with the birdge approach is the high part numbers, especially needing 2 op amps.
You're in the range of operation where a H-bridge will likely work (I've never seen H-bridge used to describe a bridged linear amplifier). No linear mode devices, simple hard switching and a control loop.
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7453
  • Country: pl
Re: High voltage safety and design considerations
« Reply #15 on: June 29, 2019, 06:35:57 am »
We finally settled on 3 7s and one 4s LiPo packs for a voltage range of 105-80V.
If such voltage is acceptable then I have seen ready to go, 12~30V in, up to 90V out, step-up converters on auction sites.
Not sure if inverting ones are available, though.
The problem with the birdge approach is the high part numbers, especially needing 2 op amps.
Yeah, that circuit as drawn was quite bad cost-wise.
It could be made much more elegantly with one differential output opamp (like OPA1632), but I doubt such parts exist for 100V operation. I wonder how difficult would those things be to build from discretes, probably not worth it as a first analog project.
So perhaps PWM'd H-bridge as others suggest? People surely use them to drive DC motors at variable speed.
Also, power efficiency of any linear solution will be quite crap.
 

Offline max_torque

  • Super Contributor
  • ***
  • Posts: 1327
  • Country: gb
    • bitdynamics
Re: High voltage safety and design considerations
« Reply #16 on: June 29, 2019, 09:48:10 am »
How quickly do you need to be able to modulate the current through the voice coil?  Given it's large inductance, the overall control bandwidth is going to be limited by that value and the supply voltage.  Why not actually use a small OTS audio amplifier?  Shouldn't be hard to find one with rails at around 100Vdc, and they are generally pretty robust and reliable. Easy to drive too, with a typical +-1V input with relatively high impedance?

 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: High voltage safety and design considerations
« Reply #17 on: June 29, 2019, 12:13:46 pm »
Then don't use the ADHV. Use ordinary, cheap op-amps, run off <30V and some discrete devices to amplify the output voltage.

Also note that it's very hard to protect semiconductors with fuses, normally the semiconductor blows first. Check that the fuse is rated for use with DC and that it has sufficient breaking capacity.

I had a look at voltage boost circuits for op amps, to be honest they are way above my level of understanding and each manufacturer recommends a different circuit (I had a look at TI, AD and ST micro app notes) so the ADHV with dual supplies is the simplest choice here. I will use HRC fuses which should be able to deal with this without any problems.

Until you get a short on their unfused side and then have 1kW or more fault to deal with. Batteries are much less safe than a current limited power supply here, you haven't discussed isolating the control signals or the actuator so the 100V supply will not be floating.
You're in the range of operation where a H-bridge will likely work (I've never seen H-bridge used to describe a bridged linear amplifier). No linear mode devices, simple hard switching and a control loop.

batteries will always be floating with respect to ground (actual ground, not batteries mid point) :-// so if someone accidentally touches the robot during operation, it shouldn't be a problem. I was actually considering a current limiting circuit for the batteries, but couldn't come up with a decent idea. I thought about using a difference amplifier and some delay to open and close a mosfet. this way the circuit trips for a set amount of time and then tries again, but the robot would fall if the delay is too long which is dangerous, and it would be ineffective if the delay is too short. please give more feedback if you know about a suitable solution for current limiting rather than tripping.

How quickly do you need to be able to modulate the current through the voice coil?  Given it's large inductance, the overall control bandwidth is going to be limited by that value and the supply voltage.  Why not actually use a small OTS audio amplifier?  Shouldn't be hard to find one with rails at around 100Vdc, and they are generally pretty robust and reliable. Easy to drive too, with a typical +-1V input with relatively high impedance?


do you have any recommendations for a small one which also ends up being cheaper than the DIY approach? the full range response should be about 1 or 2ms (excluding oscillations). as I mentioned before, we have to drive 12 actuators. I think off the shelf amps would be way too expensive to use.
 

Offline max_torque

  • Super Contributor
  • ***
  • Posts: 1327
  • Country: gb
    • bitdynamics
Re: High voltage safety and design considerations
« Reply #18 on: June 29, 2019, 01:13:47 pm »
Any audio amp is going to manage 500Hz!  (your 2m/s requirement)  and given the fact i guess you are not making many of these devices (ie it's not a volume production run) then the cost of the integrated IC's is going to be far outweighed by the time and complexity of any DIY approach:

Something like this:

https://uk.farnell.com/stmicroelectronics/tda7293v/ic-amp-audio-120v-100w/dp/1366582

is about £4.50 in low volume from a reputable supplier, some time with google would probably find a chinese alternative for half the price or less.  It has a 10V/us slew rate, so you can work out if that's good enough for your load......

You could buy one of those, fit a few support components around it (like some power rail caps for example) and be up and running ASAP with little risk of it "not working" or going BANG too many times, unlike any home-grown version.  That just leaves you needed to sort out the +-50V DC supply, which if you are battery powered is a trivial issue because you can make the middle of the cell string your "zero" volts reference.....
« Last Edit: June 29, 2019, 01:16:15 pm by max_torque »
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: High voltage safety and design considerations
« Reply #19 on: June 29, 2019, 02:15:26 pm »
don't forget that speaker drivers mostly care about resistance, since the inductance is usually very low. to achieve the 500Hz that you mentioned, the 100v supply is a requirement, like you mentioned before
Given it's large inductance, the overall control bandwidth is going to be limited by that value and the supply voltage.

I had a look on digikey, mouser and farnell. mouser and digikey didn't have anything with +-100v and farnell only has outdated parts  :-DD
https://uk.farnell.com/w/c/semiconductors-ics/amplifiers-comparators/audio-amplifiers?amplifier-class=ab&supply-voltage-range=posneg-20v-to-posneg-100v

unless there is something inherently wrong with my schematic and simulation, I think I'll stick to it for now.
 

Offline max_torque

  • Super Contributor
  • ***
  • Posts: 1327
  • Country: gb
    • bitdynamics
Re: High voltage safety and design considerations
« Reply #20 on: June 29, 2019, 02:47:36 pm »
Not really, speaker drivers care about impedance, which for a speaker at high frequency is mostly about inductance...  100vdc into a 4 ohm speaker is 25amps, or 2.5 kW......      :bullshit:




 

Offline max_torque

  • Super Contributor
  • ***
  • Posts: 1327
  • Country: gb
    • bitdynamics
Re: High voltage safety and design considerations
« Reply #21 on: June 29, 2019, 03:05:07 pm »
At DC, 25mA through 1.5K requires 37Vdc.


At AC if you want a full scale current reversal, ie peak positive(+25mA) current to peak negative (-25mA) current, in 2ms, that's 25 A/Second

So as V = L di/dt  with your 2H inductance,you'll need 50V to do that   (50 = 2 * 25)

Add the two together and we get 87V which makes me think that with an OTS  +-50V amplifier, you'll just about be able to meet your control bandwidth requirement.  So the question becomes, how important is a full scale drive swing in 2ms?  Is that a number that's estimated, picked from thin air /guessed, or established via validated simulation and calculation etc??

Assuming that the actual "project" is to make the robot do what you want to achieve, getting distracted down a rabbit hole of DIY amplifiers seems, well, like a massive distraction to me. Sure, it'll be fun and interesting but this is why many projects never get finished ime......  (you start a project to build a robot, you end up with a custom power amplifier for an application no one else needs......     :-DD  )  Been there, bought that Tshirt.......   :o
« Last Edit: June 29, 2019, 03:09:45 pm by max_torque »
 
The following users thanked this post: Someone

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: High voltage safety and design considerations
« Reply #22 on: June 29, 2019, 03:51:13 pm »
This is a university project, not a personal one  ;D the actuators are combined with pressure sensors for force driven joints, the 2ms is worst case scenario where stability can be achieved, so I like to stay closer to the 1mS figure to have some headroom. I'm not building everything, just the driver for the actuators and adc to read the pressure sensors (they're analog, 0 to 5v proportional output).
 

Offline max_torque

  • Super Contributor
  • ***
  • Posts: 1327
  • Country: gb
    • bitdynamics
Re: High voltage safety and design considerations
« Reply #23 on: June 29, 2019, 06:44:25 pm »
In which case, a good option, one that allows the rest of the program to proceed is to build a basic prototype with off the shelf parts, which can be done quickly and cheaply (even if you only make a single channel test setup to start with) establish how that performs, and based on those results roll-your-own final solution. Chances are, what you'll learn using the prototype will steer your design optimisation and specification into areas that you probably haven't even thought of, or that are currently considered un-important.  These days, i meet lots of young engineers, full of enthusiasm and great ideas, wanting to make complex systems, and yet they don't have the necessary experience to get a feel for "what works". In these cases, i send them away to "make a first prototype" and report back. In pretty much every case, the small delay in putting together than prototype is massively outweighed by the eventual time saving to the overall project from the early experience gained!

In this case, for probably £50 to £100, you could be up and running on the bench, with a basic actuator drive circuit, that enables to you test your control system code and servo loops etc, and one that allows you to establish the bandwidth of that basic solution as a bench mark to any eventual solution!  Given the average power consumption of the system is reasonably low (25mA @ 100V is just 2.5W) a couple of cheap, low power isolated (floating) 48V switch mode supplies would see you up and running on the bench where you can begin to explore  :-+
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: High voltage safety and design considerations
« Reply #24 on: June 29, 2019, 09:03:13 pm »
Any audio amp is going to manage 500Hz!  (your 2m/s requirement)  and given the fact i guess you are not making many of these devices (ie it's not a volume production run) then the cost of the integrated IC's is going to be far outweighed by the time and complexity of any DIY approach:

Something like this:

https://uk.farnell.com/stmicroelectronics/tda7293v/ic-amp-audio-120v-100w/dp/1366582

is about £4.50 in low volume from a reputable supplier, some time with google would probably find a chinese alternative for half the price or less.  It has a 10V/us slew rate, so you can work out if that's good enough for your load......

You could buy one of those, fit a few support components around it (like some power rail caps for example) and be up and running ASAP with little risk of it "not working" or going BANG too many times, unlike any home-grown version.  That just leaves you needed to sort out the +-50V DC supply, which if you are battery powered is a trivial issue because you can make the middle of the cell string your "zero" volts reference.....
The problem with that part is it uses bootstrapping capacitors and presumably he wants it to be work down to DC.

If the original poster is hell-bent on using dual power supply rails, then they should just use an ordinary op-amp, with a bootstrapped power supply. The op-amp just needs to be capable of outputting at least 25mA or pass transistors could be added.

Further reading:
https://www.proaudiodesignforum.com/images/pdf/Bootstrapping_Your_Op_Amp_Yields_Wide_Voltage_Swings_King_Watkins_EDN_May_13_1999.pdf
https://www.analog.com/media/en/technical-documentation/application-notes/AN-1593.pdf

 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf