Forums

Isolated high voltage opamp

Started by bitrex May 24, 2015
So I have a need for an analog isolation amplifier thing into a HV
 section at low audio frequencies.  I remembered the "HV opamp"
 thing that John Larkin posted a while back and that seems to work
 well, but I'd like to isolate the feedback via pwming it and
 sending it back across the barrier via an optocoupler and
 integrating.  I dont really want to chop it first and have more
 LV electronics on the HV side than necessary. 

Instead of having to pick a resistor divider for the largest
 possible output signal and lose resolution on the PWM I thought
 I'd bootstrap the PWM IC from the output like so:

http://pastebin.com/1Ud9R9nM

The components are all just representative and probably not what
 I'd actually use.  It seems like this test circuit should work
 (the full loop hasn't been closed) but it doesn't appear to sim
 properly.

Any ideas?
-- 


----Android NewsGroup Reader----
http://usenet.sinaapp.com/
On Sun, 24 May 2015 15:36:43 -0400 (EDT), bitrex
<bitrex@de.lete.earthlink.net> wrote:

>So I have a need for an analog isolation amplifier thing into a HV > section at low audio frequencies.
What is HV ? 1000 Vac/1500 Vdc according to IEC or something else ? What is low audio frequency ? 20 Hz, 100 Hz or 300 hz (telephony) ?
On Sun, 24 May 2015 15:36:43 -0400 (EDT), bitrex
<bitrex@de.lete.earthlink.net> wrote:

> >So I have a need for an analog isolation amplifier thing into a HV > section at low audio frequencies. I remembered the "HV opamp" > thing that John Larkin posted a while back and that seems to work > well, but I'd like to isolate the feedback via pwming it and > sending it back across the barrier via an optocoupler and > integrating. I dont really want to chop it first and have more > LV electronics on the HV side than necessary. > >Instead of having to pick a resistor divider for the largest > possible output signal and lose resolution on the PWM I thought > I'd bootstrap the PWM IC from the output like so: > >http://pastebin.com/1Ud9R9nM > >The components are all just representative and probably not what > I'd actually use. It seems like this test circuit should work > (the full loop hasn't been closed) but it doesn't appear to sim > properly. > >Any ideas?
I tweaked the basic HV opamp circuit a little... Version 4 SHEET 1 1920 1056 WIRE 624 -288 624 -320 WIRE 768 -288 768 -320 WIRE 912 -288 912 -320 WIRE 1072 -288 1072 -320 WIRE 176 -208 176 -304 WIRE 224 -208 176 -208 WIRE 496 -208 496 -304 WIRE 496 -208 416 -208 WIRE 624 -160 624 -208 WIRE 768 -160 768 -208 WIRE 912 -160 912 -208 WIRE 1072 -160 1072 -208 WIRE 496 -144 416 -144 WIRE 64 -112 48 -112 WIRE 176 -112 128 -112 WIRE 224 -112 176 -112 WIRE 432 -112 416 -112 WIRE 432 -64 432 -112 WIRE 240 16 16 16 WIRE 336 16 304 16 WIRE 432 48 432 16 WIRE 496 48 496 -144 WIRE 496 48 432 48 WIRE 176 128 176 -112 WIRE 496 128 496 48 WIRE 832 128 496 128 WIRE 896 128 832 128 WIRE -128 144 -224 144 WIRE 16 144 16 16 WIRE 16 144 -48 144 WIRE 144 144 16 144 WIRE 336 160 336 16 WIRE 336 160 208 160 WIRE 144 176 96 176 WIRE 336 176 336 160 WIRE 96 240 96 176 WIRE 336 272 336 256 WIRE -224 304 -224 224 WIRE 48 368 48 -112 WIRE 176 368 176 192 WIRE 176 368 48 368 WIRE 224 368 176 368 WIRE 496 368 496 128 WIRE 496 368 416 368 WIRE 496 432 416 432 WIRE 224 464 176 464 WIRE 432 464 416 464 WIRE 432 512 432 464 WIRE 176 576 176 464 WIRE 432 624 432 592 WIRE 496 624 496 432 WIRE 496 624 432 624 WIRE 544 624 496 624 WIRE 576 624 544 624 WIRE 16 704 16 144 WIRE 384 704 16 704 WIRE 832 704 832 128 WIRE 832 704 464 704 FLAG 176 -304 Vcc FLAG 336 272 0 FLAG 176 576 Vee FLAG 496 -304 HV+ FLAG 544 624 HV- FLAG 96 240 0 FLAG 624 -160 0 FLAG 624 -320 Vcc FLAG 768 -160 0 FLAG 768 -320 Vee FLAG 912 -160 0 FLAG 912 -320 HV+ FLAG 1072 -160 0 FLAG 1072 -320 HV- FLAG -224 304 0 FLAG 896 128 Out IOPIN 896 128 Out SYMBOL Opamps\\LT1024 176 96 R0 WINDOW 0 56 27 Left 2 WINDOW 3 35 96 Left 2 SYMATTR InstName U1 SYMBOL Optos\\MOC207 320 -144 R0 SYMATTR InstName U2 SYMBOL res 320 160 R0 WINDOW 0 63 42 Left 2 WINDOW 3 57 73 Left 2 SYMATTR InstName R1 SYMATTR Value 200 SYMBOL Optos\\MOC207 320 432 R0 SYMATTR InstName U3 SYMBOL res -32 128 R90 WINDOW 0 -52 59 VBottom 2 WINDOW 3 -42 59 VTop 2 SYMATTR InstName R2 SYMATTR Value 10k SYMBOL voltage 624 -304 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 WINDOW 0 44 40 Left 2 WINDOW 3 50 78 Left 2 SYMATTR InstName V1 SYMATTR Value 5 SYMBOL voltage 768 -304 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 WINDOW 0 45 41 Left 2 WINDOW 3 47 80 Left 2 SYMATTR InstName V2 SYMATTR Value -5 SYMBOL voltage 912 -304 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 WINDOW 0 47 34 Left 2 WINDOW 3 40 72 Left 2 SYMATTR InstName V3 SYMATTR Value 150 SYMBOL voltage 1072 -304 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 WINDOW 0 58 36 Left 2 WINDOW 3 45 71 Left 2 SYMATTR InstName V4 SYMATTR Value -150 SYMBOL res 480 688 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R4 SYMATTR Value 1Meg SYMBOL voltage -224 128 R0 WINDOW 123 46 154 Left 2 WINDOW 39 0 0 Left 2 WINDOW 0 60 80 Left 2 WINDOW 3 28 117 Left 2 SYMATTR Value2 AC 1 SYMATTR InstName V5 SYMATTR Value SINE(0 1 1000) SYMBOL cap 304 0 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C1 SYMATTR Value 1n SYMBOL res 416 -80 R0 WINDOW 0 -28 119 Left 2 WINDOW 3 -36 156 Left 2 SYMATTR InstName R3 SYMATTR Value 1Meg SYMBOL res 416 496 R0 WINDOW 0 -47 59 Left 2 WINDOW 3 -71 94 Left 2 SYMATTR InstName R5 SYMATTR Value 1Meg SYMBOL cap 128 -128 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C2 SYMATTR Value 330n TEXT -192 -72 Left 2 !.tran 20m TEXT -248 -224 Left 2 ;JL/Bitrex HV Opamp TEXT -240 -168 Left 2 ;JL May 24, 2015 which seems to make the output waveform a little cleaner. That LT1024 is a little wussy for this application, in my opinion. I didn't understand the PWM thing, and it really slows the sim, so I ripped it out.
On 5/24/2015 4:08 PM, John Larkin wrote:
> On Sun, 24 May 2015 15:36:43 -0400 (EDT), bitrex > <bitrex@de.lete.earthlink.net> wrote: > >> >> So I have a need for an analog isolation amplifier thing into a HV >> section at low audio frequencies. I remembered the "HV opamp" >> thing that John Larkin posted a while back and that seems to work >> well, but I'd like to isolate the feedback via pwming it and >> sending it back across the barrier via an optocoupler and >> integrating. I dont really want to chop it first and have more >> LV electronics on the HV side than necessary. >> >> Instead of having to pick a resistor divider for the largest >> possible output signal and lose resolution on the PWM I thought >> I'd bootstrap the PWM IC from the output like so: >> >> http://pastebin.com/1Ud9R9nM >> >> The components are all just representative and probably not what >> I'd actually use. It seems like this test circuit should work >> (the full loop hasn't been closed) but it doesn't appear to sim >> properly. >> >> Any ideas? > > > I tweaked the basic HV opamp circuit a little...
Looks good but I guess we need to know how much isolation he needs. Yours is 1Meg out-to-in, yes? Also, the 1Meg has a voltage limit as discussed in another thread. Have I overlooked something?
On Mon, 25 May 2015 08:14:15 -0500, John S <Sophi.2@invalid.org> Gave
us:

>Have I overlooked something?
Yes. You are missing about 40 trillion functional neurons... 2 just isn't enough. Try researching what the highest isolation voltage available in an opto-isolator is.
On Monday, 25 May 2015 23:18:48 UTC+10, DecadentLinuxUserNumeroUno  wrote:
> On Mon, 25 May 2015 08:14:15 -0500, John S <Sophi.2@invalid.org> Gave > us: > > >Have I overlooked something? > > Yes. You are missing about 40 trillion functional neurons... > 2 just isn't enough.
AlwaysWrong reports his own experience, more or less correctly by his own standards - he's actually got a few more neurones than two, but not nearly enough. -- Bill Sloman, Sydney
On Mon, 25 May 2015 06:50:53 -0700 (PDT), Bill Sloman
<bill.sloman@gmail.com> Gave us:

>On Monday, 25 May 2015 23:18:48 UTC+10, DecadentLinuxUserNumeroUno wrote: >> On Mon, 25 May 2015 08:14:15 -0500, John S <Sophi.2@invalid.org> Gave >> us: >> >> >Have I overlooked something? >> >> Yes. You are missing about 40 trillion functional neurons... >> 2 just isn't enough. > >AlwaysWrong reports his own experience, more or less correctly > by his own standards - he's actually got a few more neurones > than two, but not nearly enough.
Says the 'maroon' who cannot even spell the word 'neurons' correctly. I'll bet you say "idear" as well.
John S <Sophi.2@invalid.org> Wrote in message:
> On 5/24/2015 4:08 PM, John Larkin wrote: >> On Sun, 24 May 2015 15:36:43 -0400 (EDT), bitrex >> <bitrex@de.lete.earthlink.net> wrote: >> >>> >>> So I have a need for an analog isolation amplifier thing into a HV >>> section at low audio frequencies. I remembered the "HV opamp" >>> thing that John Larkin posted a while back and that seems to work >>> well, but I'd like to isolate the feedback via pwming it and >>> sending it back across the barrier via an optocoupler and >>> integrating. I dont really want to chop it first and have more >>> LV electronics on the HV side than necessary. >>> >>> Instead of having to pick a resistor divider for the largest >>> possible output signal and lose resolution on the PWM I thought >>> I'd bootstrap the PWM IC from the output like so: >>> >>> http://pastebin.com/1Ud9R9nM >>> >>> The components are all just representative and probably not what >>> I'd actually use. It seems like this test circuit should work >>> (the full loop hasn't been closed) but it doesn't appear to sim >>> properly. >>> >>> Any ideas? >> >> >> I tweaked the basic HV opamp circuit a little... > > Looks good but I guess we need to know how much isolation he needs. > Yours is 1Meg out-to-in, yes? Also, the 1Meg has a voltage limit as > discussed in another thread. > > Have I overlooked something? > >
At the moment the +150 -150 volts is sufficient, but I was also interested in developing a topology that would work for significantly higher voltages...like a thousand or more plus to minus. I don't think a resistor from output to input would be appropriate in that case. Hence the PWM idea. -- ----Android NewsGroup Reader---- http://usenet.sinaapp.com/
On 5/25/2015 9:09 AM, bitrex wrote:
> John S <Sophi.2@invalid.org> Wrote in message: >> On 5/24/2015 4:08 PM, John Larkin wrote: >>> On Sun, 24 May 2015 15:36:43 -0400 (EDT), bitrex >>> <bitrex@de.lete.earthlink.net> wrote: >>> >>>> >>>> So I have a need for an analog isolation amplifier thing into a HV >>>> section at low audio frequencies. I remembered the "HV opamp" >>>> thing that John Larkin posted a while back and that seems to work >>>> well, but I'd like to isolate the feedback via pwming it and >>>> sending it back across the barrier via an optocoupler and >>>> integrating. I dont really want to chop it first and have more >>>> LV electronics on the HV side than necessary. >>>> >>>> Instead of having to pick a resistor divider for the largest >>>> possible output signal and lose resolution on the PWM I thought >>>> I'd bootstrap the PWM IC from the output like so: >>>> >>>> http://pastebin.com/1Ud9R9nM >>>> >>>> The components are all just representative and probably not what >>>> I'd actually use. It seems like this test circuit should work >>>> (the full loop hasn't been closed) but it doesn't appear to sim >>>> properly. >>>> >>>> Any ideas? >>> >>> >>> I tweaked the basic HV opamp circuit a little... >> >> Looks good but I guess we need to know how much isolation he needs. >> Yours is 1Meg out-to-in, yes? Also, the 1Meg has a voltage limit as >> discussed in another thread. >> >> Have I overlooked something? >> >> > > At the moment the +150 -150 volts is sufficient, but I was also > interested in developing a topology that would work for > significantly higher voltages...like a thousand or more plus to > minus. I don't think > a resistor from output to input would be > appropriate in that case. Hence the PWM idea. >
Can you tell us how much isolation you need? A 1T ohm resistor can provide 1T ohm of isolation, of course. And JL has some, so they are available. Need to know more.
On Tuesday, 26 May 2015 00:03:13 UTC+10, DecadentLinuxUserNumeroUno  wrote:
> On Mon, 25 May 2015 06:50:53 -0700 (PDT), Bill Sloman > <bill.sloman@gmail.com> Gave us: > > >On Monday, 25 May 2015 23:18:48 UTC+10, DecadentLinuxUserNumeroUno wrote: > >> On Mon, 25 May 2015 08:14:15 -0500, John S <Sophi.2@invalid.org> Gave > >> us: > >> > >> >Have I overlooked something? > >> > >> Yes. You are missing about 40 trillion functional neurons... > >> 2 just isn't enough. > > > > AlwaysWrong reports his own experience, more or less correctly > > by his own standards - he's actually got a few more neurones > > than two, but not nearly enough. > > Says the 'maroon' who cannot even spell the word 'neurons' correctly. > I'll bet you say "idear" as well.
You'd be wrong. If you weren't AlwaysWrong, you'd be aware that "neuron" is one of the mistakes Noah Webster wished on the gullible American public in his scheme to make Americans spell English in a way that wasn't documented in the English language dictionaries published in the UK. He foisted his dictionary on the American public, and you are still suffering from his commercial acumen and total lack of linguistic principle. I speak and write English English - which isn't wildly different from the Australian English that I learnt as child - and spent 22 years working in the UK, so I write "neurone". I understand what Americans mean when they write "neuron" but I don't bother spelling the word the way they do. -- Bill Sloman, Sydney