Can changing amplifier in Sullen-Key filter affect the frequency response much?

[loop123]

Consider this Sullen-Key low pass filter.



If you change the amplifier to LF411CP (with the same pins). How much would the frequency response change?

When I increase the value of the capacitor from 1000pF to 6800pF, the frequency get lowered. Why is that?

I found a Ltspice file using the third order filter (see attached). I changed it to the 2nd order Sullen-key by deleting some parts and changing the values. Please check if the entry is correct. Also the voltage source is reverse. This is ok for AC, isn't it? Second. What amplifier is in the Ltspice attached Ltspice file?. How do you change it to the LF411CP? I want to see if the frequency response would change. If it's hard to change. Please change it to LF411CP and attached the edited Ltspice.

 sullen-key.asc (1.21 kB - downloaded 53 times.)



What does ".ac oct 20 10 10000" mean? If I removed it, I can't run the simulator anymore.

Thank you.

[Benta]

Perhaps you should add power supplies and an output to your simulation schematic?

[loop123]

Perhaps you should add power supplies and an output to your simulation schematic?

It ran. I just want to know the frequency response of the output. You mean it didn't have any power supplies? But why did the plot run? I don't know how to add power supply since I just edit the existing file with third order filter. I don't know how to use Ltspice except to delete components, change values and run the simulator.

[magic]

Yes, it ran. But do you see the vertical scale?

At -160dB you are simulating parasitic coupling through the feedback network or things like that...

[loop123]

Yes, it ran. But do you see the vertical scale?

At -160dB you are simulating parasitic coupling through the feedback network or things like that...

There is a V1, AC1 in the left side. What is it? I thought it was the power supply. Where and how do you add the power supply? I don't know how to use LTspice Pls download the file in original thread and add it and upload it back so I can run the frequency response plot. Thanks.

[Solder_Junkie]

It ran. I just want to know the frequency response of the output.

While I have both LTSpice and Tina, I prefer the latter for simple circuits using generic components...

See attached using the original circuit (1n), and also another run using 6n8. Note the "blip" in the response curve. The "knee" is around 38 KHz with 1n and 18 KHz with 6n8. It makes no difference which op amp you use.

SJ

[loop123]

It ran. I just want to know the frequency response of the output.

While I have both LTSpice and Tina, I prefer the latter for simple circuits using generic components...

See attached using the original circuit (1n), and also another run using 6n8. Note the "blip" in the response curve. The "knee" is around 38 KHz with 1n and 18 KHz with 6n8. It makes no difference which op amp you use.

SJ

Please change the capacitor C2 to 6800pF (instead of 1000pF) and share the frequency response. Thanks.

[Solder_Junkie]

Please change the capacitor C2 to 6800pF (instead of 1000pF) and share the frequency response. Thanks.

Check my earlier post, that is what the 6n8 (6800pF = 6n8F) image shows...

SJ

[loop123]

Please change the capacitor C2 to 6800pF (instead of 1000pF) and share the frequency response. Thanks.

Check my earlier post, that is what the 6n8 (6800pF = 6n8F) image shows...

SJ

In a 2 pole Sulley-key low pass filter. What components must you change to adjust the frequency response? You mean both 1000pF and 6800pF give both about 50,000 Hz frequency response? What must you change to make it only 4000 Hz response?

[loop123]

Please change the capacitor C2 to 6800pF (instead of 1000pF) and share the frequency response. Thanks.

Check my earlier post, that is what the 6n8 (6800pF = 6n8F) image shows...

SJ

Anyway, here is my complete question. How does putting a 2 pole sulley-key filters exactly remove the ripples in the following circuit?  If someone used an LF411CP amp instead and with values of C2=6800pF and C1=33pF with similar resistors. How would it affect filtering the ripples? This is the specific information I seek. Please simulate them and see the result as it's very important to me. Thanks.

[Solder_Junkie]

Texas Instruments have a design tool: https://webench.ti.com/filter-design-tool/

You can add filters in series, such as this one: https://www.qsl.net/g4aon/pdfs/aonrx_afpre_smd_v3.pdf
Which gives the attached response curve.

Just play around with modelling software and see what results you can obtain.

SJ

[jonpaul]

SALLEN-KEY!

Since 1970s much better active fil topologies are avail and Lin Tech, Maxim made IC specific for AF.

Easite to build a breadboard and MAKE THE CIRCUIT than to bother with simulators, GIGO.

Just place a 8 pin DIM socket, collect a bunch of opamps and play

Suggest NE5534 wideband low noise audio opamp.

Enjoy,

j

[loop123]

Texas Instruments have a design tool: https://webench.ti.com/filter-design-tool/

You can add filters in series, such as this one: https://www.qsl.net/g4aon/pdfs/aonrx_afpre_smd_v3.pdf
Which gives the attached response curve.

Just play around with modelling software and see what results you can obtain.

SJ

My situation is. I got a device that has the ISO122 isolator and salley key but the manufacturer somehow used LF411CP, 6800pF, 33pF instead of using the OPA602, 1000pF, 220pF and I'm getting ripples like in the following:



If I'd order the original OPA602, 1000pF, 220pF and replace these. Would the ripples go away?  Do merely more poles can remove the ripples? how are they removed exactly?  Is it the values of the capacitors or resistors or combined? Please simulate them in any way you can so I can decide whether to order the OPA602.

[Terry Bites]

fc=1/sqrt(R1*R2*C1*C2) to make life simple set C1=C2, R1=R2.

The opamp type does impact performance. If the opamp bandwidth rolls off below or close to fc the response will not be as calculated.
See "GBWP"
The rule of thumb is to use an ampfier with more than 10x bandwidth of your low pass cut off. For a 10kHz lpf, the opamp must provide the required filter gain to at least 100kHz.

Here's a very thourough treatment of this filter type. You love maths right?
Its Sallen, not Sulley from Monsters Inc.

[loop123]

Please tell me the detail theory why adding more poles that makes it closer to the vertical ideal response can remove the ripples of the carrier frequency.

What unity gain finish module is available with the highest pole or order? 

Thanks.

[loop123]

It ran. I just want to know the frequency response of the output.

While I have both LTSpice and Tina, I prefer the latter for simple circuits using generic components...

See attached using the original circuit (1n), and also another run using 6n8. Note the "blip" in the response curve. The "knee" is around 38 KHz with 1n and 18 KHz with 6n8. It makes no difference which op amp you use.

SJ

I'm reflecting on your graphs and comments. Although the "knee" is around 38 KHz with 1n and 18 KHz with 6n8. How can you conclude it made no difference if the OPA602 or LF114CP is used?

[sharow]

My situation is. I got a device that has the ISO122 isolator and salley key but the manufacturer somehow used LF411CP, 6800pF, 33pF instead of using the OPA602, 1000pF, 220pF and I'm getting ripples like in the following:

You can see about 5 cycle between 19:01.00 and 19:01.10 .
100ms/5 = 20ms  or  50Hz.

[vk6zgo]

Yes, it ran. But do you see the vertical scale?

At -160dB you are simulating parasitic coupling through the feedback network or things like that...

There is a V1, AC1 in the left side. What is it? I thought it was the power supply. Where and how do you add the power supply? I don't know how to use LTspice Pls download the file in original thread and add it and upload it back so I can run the frequency response plot. Thanks.

V1 is the input signal.

LT Spice, in their wisdom, always use a DC generator symbol with + & - polarity markings on it, unlike the ac generator symbol which I am familiar with from decades of seeing it used.
That is why LTSpice annoys me so much.

If the generator is so marked, it is easy for someone to think it is a DC power supply source.

[RFDx]

My situation is. I got a device that has the ISO122 isolator and salley key but the manufacturer somehow used LF411CP, 6800pF, 33pF instead of using the OPA602, 1000pF, 220pF and I'm getting ripple ...

6.8nF and 33pF are not a good choice, this values give you alot of peaking (gain) at the corner frequency of the filter.

If I'd order the original OPA602, 1000pF, 220pF and replace these. Would the ripples go away?  Do merely more poles can remove the ripples?

The ripples can only be attenuated and not removed completely, depending on the order of the chosen filter. The single pole 50kHz filter inside of the ISO122 amplifier achieves an attenuation of 20dB per decade. This means the 500kHz modulation frequency used by the ISO122 will already be attenuated by 20dB at the output. The Sallen & Key filter following the output provides two additional poles and another 40dB attenuation. With a total of 60dB of attenuation, the 500kHz modulation frequency should be, according to the datasheet, below the noise floor.

As for the choice between the OPA602 or the LF411, the only difference is in the attenuation far above the passband (a known disadvantage of Sallen & Key filters) where the OPA602 has lower output impedance and gives a somewhat better result. See attached frequency and phase responses for the 2nd oder filter alone and that of the maximally flat filter including the pole inside the ISO122 amplifier.

You never mentioned what you expect of your "device". Do you really need 50kHz bandwidth? What kind of input signals do you have?

[loop123]

My situation is. I got a device that has the ISO122 isolator and salley key but the manufacturer somehow used LF411CP, 6800pF, 33pF instead of using the OPA602, 1000pF, 220pF and I'm getting ripple ...

6.8nF and 33pF are not a good choice, this values give you alot of peaking (gain) at the corner frequency of the filter.

I was mistaken. 332 capacitor means 3300pF. so it's really Sullen and key of 6800pF and 3300pF with the LF411CP. This means low pass of less than 10kHz. Compare this to the original datasheet Sullen and Key 1 000pF and 220pF with cutoff of 50,000kHz. What does this do to the ripples? Does it mean the ripples at all frequencies will be less when the 50kHz cutoff is used versus the say 7kHz cutoff? Or does it mean only ripples below 7kHz would be cutoff in the latter?

Also more poles means more abrupt cutoff. What has this got to do with suppressing the ripples? Kindly explain because I want to understand the mechanism.

It's just an amplifier. I don't need 50kHz but 7kHz may be too low for say audio use.

If I'd order the original OPA602, 1000pF, 220pF and replace these. Would the ripples go away?  Do merely more poles can remove the ripples?

The ripples can only be attenuated and not removed completely, depending on the order of the chosen filter. The single pole 50kHz filter inside of the ISO122 amplifier achieves an attenuation of 20dB per decade. This means the 500kHz modulation frequency used by the ISO122 will already be attenuated by 20dB at the output. The Sallen & Key filter following the output provides two additional poles and another 40dB attenuation. With a total of 60dB of attenuation, the 500kHz modulation frequency should be, according to the datasheet, below the noise floor.

As for the choice between the OPA602 or the LF411, the only difference is in the attenuation far above the passband (a known disadvantage of Sallen & Key filters) where the OPA602 has lower output impedance and gives a somewhat better result. See attached frequency and phase responses for the 2nd oder filter alone and that of the maximally flat filter including the pole inside the ISO122 amplifier.

You never mentioned what you expect of your "device". Do you really need 50kHz bandwidth? What kind of input signals do you have?

[loop123]

The following is what I meant above. Also the input is 1V even in the file at original message. It's just that when I drew the feedback from - Vin to output. A pixel must have been missed so it's not continuous and the voltage is only in nV instead of 1V. Why do you get only nV when output is not connected to -Vin?

[magic]

You are still simulating without power supplies to the chip, so don't expect miracles.
You need to create two DC voltage sources, one pin of each to ground and the other pin to the supply pins of the opamp. Press F2 and search for "voltage".

If you still don't know how to do it, press F1 to open built-in help or look up some LTspice tutorial, because you are going nowhere without it.

[loop123]

You are still simulating without power supplies to the chip, so don't expect miracles.
You need to create two DC voltage sources, one pin of each to ground and the other pin to the supply pins of the opamp. Press F2 and search for "voltage".

If you still don't know how to do it, press F1 to open built-in help or look up some LTspice tutorial, because you are going nowhere without it.

I've been running this for 2 days already can't do it. Can you or someone please add the power supply to the file attached. It has 6800pF and 3300pF. Why can it show low pass filtering below 10kHz if there is no power?

[magic]

I don't know how to use Ltspice except to delete components, change values and run the simulator.

This is your fundamental problem.
Learning how to add voltage sources, draw connections and maybe use net labels would take not two days but less than two hours.

Tip: F2, F3, F4, respectively.

[RFDx]

I was mistaken. 332 capacitor means 3300pF. so it's really Sullen and key of 6800pF and 3300pF with the LF411CP. This means low pass of less than 10kHz.

With 6.8n/3.3n the corner frequency is at ~4.7kHz (see attached screenshot).

Compare this to the original datasheet Sullen and Key 1 000pF and 220pF with cutoff of 50,000kHz. What does this do to the ripples? Does it mean the ripples at all frequencies will be less when the 50kHz cutoff is used versus the say 7kHz cutoff? Or does it mean only ripples below 7kHz would be cutoff in the latter?

I don't know what frequency the ripples have. The screenshot you provided was unclear because you cut off the units respectively the values of the time and amplitude scale.

If you use the 4.7kHz filter variant, the amplitude of the ripples that are far above the corner frequency will be less compared to the 50kHz filter. Ripples inside the filter passband can obviously not be attenuated at all.

Also more poles means more abrupt cutoff. What has this got to do with suppressing the ripples? Kindly explain because I want to understand the mechanism.

As already mentioned, you can only suppress ripples that are outside of the filter passband. A steeper filter (more poles) ensures ripple voltage(s) with lesser amplitude.

It's just an amplifier. I don't need 50kHz but 7kHz may be too low for say audio use.

So 20kHz bandwidth will suffice?

[loop123]

I was mistaken. 332 capacitor means 3300pF. so it's really Sullen and key of 6800pF and 3300pF with the LF411CP. This means low pass of less than 10kHz.

With 6.8n/3.3n the corner frequency is at ~4.7kHz (see attached screenshot).

Compare this to the original datasheet Sullen and Key 1 000pF and 220pF with cutoff of 50,000kHz. What does this do to the ripples? Does it mean the ripples at all frequencies will be less when the 50kHz cutoff is used versus the say 7kHz cutoff? Or does it mean only ripples below 7kHz would be cutoff in the latter?

I don't know what frequency the ripples have. The screenshot you provided was unclear because you cut off the units respectively the values of the time and amplitude scale.

If you use the 4.7kHz filter variant, the amplitude of the ripples that are far above the corner frequency will be less compared to the 50kHz filter. Ripples inside the filter passband can obviously not be attenuated at all.

Also more poles means more abrupt cutoff. What has this got to do with suppressing the ripples? Kindly explain because I want to understand the mechanism.

As already mentioned, you can only suppress ripples that are outside of the filter passband. A steeper filter (more poles) ensures ripple voltage(s) with lesser amplitude.

It's just an amplifier. I don't need 50kHz but 7kHz may be too low for say audio use.

So 20kHz bandwidth will suffice?

Did you use R1=4.75k, R2=9.76k, C1=3300pF, C2=6800pF?  Why did online calculator shows cutoff of only 4.934kHz?? Instead yours show more than 50kHz??

20kHz will suffice but I want to change the C1=1000pF, C2=220pF, and use OPA602 just like in the original ISO122P datasheet. This is in order to hopefully remove all the ripples in the following.

https://www.youtube.com/watch?v=DO58qtAMYBg&feature=youtu.be

It started with the high pass filter at 100Hz (clean signal). This is irregardless if the input frequency is 10 Hz or 40Hz or 90Hz or higher). Then I switched it to 1000Hz, then 3000 Hz, 10000Hz, 30000Hz, 50000Hz (Yes my amplifier has dials for all those frequencies). The noises keep getting worse. You can hear the click at the video as I adjusted them. Halfway I returned from 50kHz back to 100Hz.

Going to the 6800pF, 3300pF, LF411CP which the actual circuit is using. If it filters anything above 5kHz. Why is the output showing up to 50kHz?? This greatly puzzles me now.

[loop123]

Can't you tell the frequency of the ripple even with the datasheet? It says the 20mV peak to peak came from carrier modulation of 500kHz. What is the ripple frequency?

https://www.ti.com/lit/ds/symlink/iso122.pdf?ts=1707196390069&ref_url=https%253A%252F%252Fwww.google.com%252F



Remember it mentioned in page 10 the Sullen-Key low pass filter up to 50,000Hz can remove the ripples. You were saying out the 50kHz and not within? But what is the frequency of the ripples in the first place?

[RFDx]

Did you use R1=4.75k, R2=9.76k, C1=3300pF, C2=6800pF?

Yes.

Why did online calculator shows cutoff of only 4.934kHz?? Instead yours show more than 50kHz??

My (last) screenshot shows a corner frequency of ~4.656kHz @ -3dB and that is including the 1st order lowpass inside the ISO122 amp.

20kHz will suffice but I want to change the C1=1000pF, C2=220pF, and use OPA602 just like in the original ISO122P datasheet. This is in order to hopefully remove all the ripples in the following.

More bandwidth means more noise. Increasing the bandwidth unnecessarily from 20kHz to 50kHz doesn't make much sense if you want less noise/ripples at the output.

https://www.youtube.com/watch?v=DO58qtAMYBg&feature=youtu.be

It started with the high pass filter at 100Hz (clean signal). This is irregardless if the input frequency is 10 Hz or 40Hz or 90Hz or higher). Then I switched it to 1000Hz, then 3000 Hz, 10000Hz, 30000Hz, 50000Hz (Yes my amplifier has dials for all those frequencies). The noises keep getting worse. You can hear the click at the video as I adjusted them. Halfway I returned from 50kHz back to 100Hz.

The video shows that the frequency of the input signal fed to the amplifier is always the same. The "frequency dial" your amplifier has, seems to be an option to change the bandwidth of the amplifier and not the frequency of an input signal. I don't think your amplifier has a built in signal generator. If you increase the bandwith of the amplifier more noise comes through (which btw is perfectly normal).

Going to the 6800pF, 3300pF, LF411CP which the actual circuit is using. If it filters anything above 5kHz. Why is the output showing up to 50kHz?? This greatly puzzles me now.

See above response. You never applied 50kHz to the amplifier so you can't really tell if 50kHz is showing up or not. With 50kHz at the input of the ISO122 amp and with the additional Sallen-Key filter (6.8n/3.3n), the output must show an attenuation > 40dB (that's less than 1/100 in amplitude) of the input signal.

[RFDx]

Can't you tell the frequency of the ripple even with the datasheet? It says the 20mV peak to peak came from carrier modulation of 500kHz. What is the ripple frequency?

The frequency would be 500kHz, that is the frequency the modulator is clocked with. This unwanted product shows up, according to the datasheet, with typ. 20mVpp at the output (that is after the 50kHz 1st order lowpass filter inside the ISO122).
With the additional 2nd order 50kHz (1nF/220pF, OPA602 or LF411) Sallen-Key lowpass filter, the amplitude of the 500kHz clock signal will be further attenuated by a factor of 100 (40dB) and the remaining amplitude should be ~0.2mVpp.
With the corner frequency of the filter lowered to 4.656kHz (6.8nF/3.3nF, LF411), the amplitude of the 500kHz clock signal will be attenuated by ~75dB and should be ~3.5uVpp.
Do you have test equipment (AC voltmeter, spectrum analyzer etc.) that goes up to 500kHz and can measure such low voltages?

[loop123]

By the way. The input is 1mV about 80 Hz using signal generator (I tried 2 brands) passing through the ISO122p unit. When main unit passband is set to 100Hz. No ripple. When it is set to 1000Hz. There are already ripples as you can see in the video.

I thought this occurred because the manufacturer didn't follow the datasheet of 1000pF, 220pF, OPA602 and instead using 6800pF, 3300pF, LF411CP which I thought meant the filtered noise is only 0 to 4.75kHz. I thought using the datasheet original 50kHz passband means everything inside 50kHz will be ripple free. But it's the opposite as you explained.

But how come at mere 1000Hz pass band switch, ripples at 500kHz already showed up with the input of 1mV, 80Hz via the ISO122?

Note that without the ISO122. You can input even 30kHz into main unit without any ripples. Does this mean the ISO122 is only good up to 1mV 100Hz input only?

[RFDx]

By the way. The input is 1mV about 80 Hz using signal generator (I tried 2 brands) passing through the ISO122p unit. When main unit passband is set to 100Hz. No ripple. When it is set to 1000Hz. There are already ripples as you can see in the video.

A 1mV input signal is way to small. The ISO-amp produces itself almost 1mV of noise with a bandwidth of 50kHz, which means the signal to noise ratio (SNR) will be very bad. This noise is inside the passband and can not be removed.

To increase the SNR you can either increase the input signal or decrease the bandwidth of the ISO-amp. With a power supply of +/-15V the ISO amp accepts an input signal of max. +/-12.5V.

[loop123]

By the way. The input is 1mV about 80 Hz using signal generator (I tried 2 brands) passing through the ISO122p unit. When main unit passband is set to 100Hz. No ripple. When it is set to 1000Hz. There are already ripples as you can see in the video.

A 1mV input signal is way to small. The ISO-amp produces itself almost 1mV of noise with a bandwidth of 50kHz, which means the signal to noise ratio (SNR) will be very bad. This noise is inside the passband and can not be removed.

To increase the SNR you can either increase the input signal or decrease the bandwidth of the ISO-amp. With a power supply of +/-15V the ISO amp accepts an input signal of max. +/-12.5V.

But why is the ISO122 signal clean if the filter is set to 100Hz? Also even without using the Sullen-key and tapping directly the ISO122, the 90Hz, 2mV signal is clean with 100Hz cutoff selected in the main unit.

How about an opto-isolator? Can it accept 50kHz without any ripple? What is the maximum bandwidth of opticoupler?

Btw.. i couldn't find in the ISO122P data sheet the noise at 1mV. Where did you read it?

[loop123]

I don't know how to use Ltspice except to delete components, change values and run the simulator.

This is your fundamental problem.
Learning how to add voltage sources, draw connections and maybe use net labels would take not two days but less than two hours.

Tip: F2, F3, F4, respectively.

I have spent just that 2 hours putting the power supply. I did. However the frequency response is still the same with or without a power supply (15V vs 0V/not connected).



The above is with 15V inputted. Below is with 0V (same as cutting the power supply)



This is DC check for both V+ and V- for 15V.





Without power supply. The frequency response came directly from the connected resistors and capacitors, even in real circuit, right? With this. Even if the real amp such as the LF411CP is defective. It can still show frequency response? I was wondering if my real LF411CP is defective and the frequency response came from the R1,R2, C1,C2 in the 2 pole Sullen-Key

In LTspice, how come the frequency response is same with both power supply 15V inputted and no power supply?

here is the file, pls check it out, remove the battery and see the frequency response is the same

[RFDx]

But why is the ISO122 signal clean if the filter is set to 100Hz? Also even without using the Sullen-key and tapping directly the ISO122, the 90Hz, 2mV signal is clean with 100Hz cutoff selected in the main unit.

Less bandwidth equals less noise voltage at the output riding on top of the actual signal. The overall bandwidth is determined by the 100Hz bandwidth of the main amplifier. Compared to the full 50kHz, the noise voltage will be smaller by a factor of SQR(50000/100)= 22.36 when the bandwidth is switched to 100Hz.

How about an opto-isolator? Can it accept 50kHz without any ripple? What is the maximum bandwidth of opticoupler?

You would need an analog optocoupler. The datasheet will probably give a hint to how much noise is to be expected. Depending on the mode of operation, the max bandwidth is maybe 50...100kHz.

How about putting a low noise preamp with a gain of 100 (or even 1000) in front of the ISO-amp?

Btw.. i couldn't find in the ISO122P data sheet the noise at 1mV. Where did you read it?

The noise spectral density of the ISO122 is 4uV/SQR(Hz) ("Noise" on p.5 of the datasheet). To get the noise voltage (RMS) multiply with SQR(bandwidth-in-Hz). With the bandwidth switched to the max. of 50kHz, the noise voltage at the output would be 0.894mV RMS. That is alot for an input signal of only 1mV (RMS?). With a bandwidth of only 100Hz, the noise voltage is only 40uV RMS. The main amplifier also contributes some noise to the output but it is probably very little in comparison to the ISO-amp and can be ignored.

Fig. 12 in the datasheet shows the unfiltered 20mVpp ripple from the 500kHz clock feedthrough plus the broadband noise from the ISO-amp. Fig. 13 shows only the broadband noise at the output after being filtered by the additional 50kHz Sallen-Key lowpass filter. The 500kHz ripple is gone (buried in noise). The broadband noise has at least 6...7mVpp. Divide this by 6.6 to get the RMS value.

In LTspice, how come the frequency response is same with both power supply 15V inputted and no power supply?
here is the file, pls check it out, remove the battery and see the frequency response is the same

I don't have LTspice or else I would have already done the necessary changes for you. In the simulation software I am using, there are virtual opamps that have default/built in supply voltages and hence don't need external power supplies. I guess LTspice has something similar?

[magic]

I have spent just that 2 hours putting the power supply. I did. However the frequency response is still the same with or without a power supply (15V vs 0V/not connected).

Sorry, it hasn't occurred to me that you are using an opamp model which ignores power supplies.
I downloaded this file and it appears to be working correctly now, with or without the 15V supplies.
I don't know why it was showing -160dB initially, you must have changed something that fixed it.

[loop123]

I have spent just that 2 hours putting the power supply. I did. However the frequency response is still the same with or without a power supply (15V vs 0V/not connected).

Sorry, it hasn't occurred to me that you are using an opamp model which ignores power supplies.
I downloaded this file and it appears to be working correctly now, with or without the 15V supplies.
I don't know why it was showing -160dB initially, you must have changed something that fixed it.

The circuit was originally a 3rd order Butterworth using the MCP607. I edited the drawing to delete some lines and change the resistors and capacitors. When a 2nd order has short from V- to output. I tried to use draw lines but incorrect added lines so the line is not connected. After I change the line again from -V to output, it connects and the values are correct.

In 2 Pole Sullen-Key. You can put any op-amp and the frequency response is mostly identical? Because the amp in all files is still modeled after MCP607. I didn't change anything and just assume the LF411CP has same values or functions.

[Smokey]

Just a heads up since people are still typing "Sullen"... it's "Sallen-Key" with an "a".  It's two dudes' names.
https://en.wikipedia.org/wiki/Sallen%E2%80%93Key_topology

The dudes were R.P. Sallen and E.L. Key of MIT's Lincoln Labs in 1955.

[magic]

In 2 Pole Sullen-Key. You can put any op-amp and the frequency response is mostly identical? Because the amp in all files is still modeled after MCP607. I didn't change anything and just assume the LF411CP has same values or functions.

Closed loop response is determined by the feedback network, as long as gain bandwidth product is high enough. For only a few kHz BW and unity gain anything works, with exception of some exotic ultra low power types.

[RFDx]

1000Hz is enough for me because EMG is just 1000Hz and below. So how do I remove the ripples at 1000Hz? Maybe by changing the capacitor values to higher ones? what values for C1 and C2 do you recommend for the ripples to vanish, something where corner frequency is just about 1kHz

The ripples are not at 1000Hz. The "ripple" seen in Audacity is actually broadband noise that occupies the entire effective bandwidth. You can't get rid of the noise but you can get around it.
The first step, as you also suggested, would be to change the bandwidth of the S&K LP-filter to 1kHz and thereby lowering the noise voltage contributed by the ISO-amp by a factor of ~7.07 [SQR(50000/1000)]. Two suggestions (2nd & 3rd order S&K LPF) with component values are attached.

Your input signal is very small and the noise voltage is still substantial, even after the 1kHz filter, which means the signal to noise ratio (SNR) after the ISO-amp/filter will be rather modest. To raise the SNR, run the input signal through a low noise preamp with a gain of 100 (or even 1000). The amplified signal then passes through the noisy ISO-amp, the S&K LP-filter and last but not least, if neccessary, through a 100:1 (1000:1) attenuator to recover the original amplitude of the input signal (2mV). Despite the attenuation stage, the SNR will be (mostly) retained and the signal will be free of visible noise and look clean. The attached screenshot shows the difference between the two noisy output voltages with no amplification at all and amplification by a factor of 100, including the recovery of the original 2mV signal with the help of a simple attenuation stage.

Now that you know the ripple is 500kHz.. So there are no ripples inside the filter passband (which obviously can't be attenuated at all as your described)? So the noises I'm seeing in the waveforms below (see the message) are not from the ripples but from an inherent "noise spectral density of the ISO122" being 4uV/SQR(Hz) that is not related to the ripples at all?  These are two independent thing? I can remove the ripples using bigger capacitor values to make the bandwidth smaller but I can't remove the noise spectrual density because it is within the passband. is all these what you mean?

Exactly. The sufficiently suppressed 500kHz ripple form the modulator is not the actual problem, the broadband noise, added by the ISO-amp to the signal, is.

Please confirm so I know how junk the ISO122p is. However, if I put 100 or 1000 gain before the ISO. Would it produce clean signal at all? That means the amplifier before it would be almost like the main amplifier, i don't know if the DC-DC converter can supply current to it. And the output will be 1V and the main amplifier won't be useful at all because it would just accept it as 1X?

The ISO-amp is definitely not junk, it just doesn't do well with small input signals. Before adding a low noise preamp, check out yourself what happens when you increase the input signal. Inject 202mV instead of 2mV into the amp, insert a 100:1 attenuator (1kOhm + 10Ohm resistors) after the S&K LP-filter  and take a look at the output signal in Audacity. Does it look cleaner?
A low noise amp needs only a few mA. I don't see the DC-DC converter having problems supplying the additional current.

[loop123]

But why is the ISO122 signal clean if the filter is set to 100Hz? Also even without using the Sullen-key and tapping directly the ISO122, the 90Hz, 2mV signal is clean with 100Hz cutoff selected in the main unit.

Less bandwidth equals less noise voltage at the output riding on top of the actual signal. The overall bandwidth is determined by the 100Hz bandwidth of the main amplifier. Compared to the full 50kHz, the noise voltage will be smaller by a factor of SQR(50000/100)= 22.36 when the bandwidth is switched to 100Hz.

How about an opto-isolator? Can it accept 50kHz without any ripple? What is the maximum bandwidth of opticoupler?

You would need an analog optocoupler. The datasheet will probably give a hint to how much noise is to be expected. Depending on the mode of operation, the max bandwidth is maybe 50...100kHz.

How about putting a low noise preamp with a gain of 100 (or even 1000) in front of the ISO-amp?

Btw.. i couldn't find in the ISO122P data sheet the noise at 1mV. Where did you read it?

The noise spectral density of the ISO122 is 4uV/SQR(Hz) ("Noise" on p.5 of the datasheet). To get the noise voltage (RMS) multiply with SQR(bandwidth-in-Hz). With the bandwidth switched to the max. of 50kHz, the noise voltage at the output would be 0.894mV RMS. That is alot for an input signal of only 1mV (RMS?). With a bandwidth of only 100Hz, the noise voltage is only 40uV RMS. The main amplifier also contributes some noise to the output but it is probably very little in comparison to the ISO-amp and can be ignored.

In your comments above. The application of the formulas for suppressing broadband noise is via selecting 100Hz in the switch in the main amplifier. You were not describing changing the 2 pole LF411 capacitor values. But in the following statement. You were describing same formulas but changing the 2 pole values. Im a bit confused. If suppression of the broadband noises can occur (right?) from either changing 2 pole capacitor values or selecting the 1000Hz cut off filter switch in main amplifier. why cant I just select the switch to 1000Hz and avoid changing the 2 pole capacitor values to 1000Hz cutoff? What would happen when both 2 pole values is changed and selecting the main switch to 1000Hz filter? Isnt it either will work. what is advantage by doing both? Thanks. you wrote:

"The first step, as you also suggested, would be to change the bandwidth of the S&K LP-filter to 1kHz and thereby lowering the noise voltage contributed by the ISO-amp by a factor of ~7.07 [SQR(50000/1000)]. Two suggestions (2nd & 3rd order S&K LPF) with component values are attached."

[RFDx]

Im a bit confused. If suppression of the broadband noises can occur (right?) from either changing 2 pole capacitor values or selecting the 1000Hz cut off filter switch in main amplifier. why cant I just select the switch to 1000Hz and avoid changing the 2 pole capacitor values to 1000Hz cutoff?

I had the impression you wanted to lower the bandwith to 1kHz, my bad. It is perfectly fine to keep the original components (6.8n/3.3n/LF411) in the S&K LPF and switch the biomed amp to 1kHz.

What would happen when both 2 pole values is changed and selecting the main switch to 1000Hz filter? Isnt it either will work. what is advantage by doing both? Thanks.

Not much. The overall filtering would be tighter with somewhat less noise but not enough to make a real difference in the end.

[Andy Chee]

I'll just put a resistor in series to the batteries to make sure it won't reach greater than 40mA for example, and avoid the ISO122 and DCP010512DBP  altogether.

Thanks a lot!

No.  The correct location for the resistor is in series with your skin probe wire connection.

[RFDx]

Thanks. The ISO122 is powered by an isolated DC-DC converter called  DCP010512DBP see datasheet at  https://rocelec.widen.net/view/pdf/6mnzrokr1h/sbvs012f.pdf?t.download=true&u=5oefqw

It has input of 5 Volts and output of 12 Volts 83mA. However my ISO-Z isolation head stage and the main unit is only powered by batteries. 12 pcs of 1.2v rechargeable NiMH batteries for total of 1.2V x 12 = 14.4V, but since the main unit uses V+ and V- then it got it from the following configuration where each V+ is 7.2V and V- is -7.5V. I never use any AC-DC adaptor so no problem about isolating ground of AC.

You don't need an ISO-amp or a DC-DC converter at all. The biomed amp is powered from batteries, is not earth referenced and already floating.

Most uses of Isolated DC-DC Converter is to isolate the main power ground. But my use is to protect from 90mA from batteries. A sailor who used multimeter died from 9V battery after it reach 90mA (maybe because of defective multimeter and he punctured his skin), see https://darwinawards.com/darwin/darwin1999-50.html It's in the official Navy guidelines, true?

Dying from a 9V block battery? That's BS. A fairytale probably told to gullible new navy cadets for some reason. 

Do you think the all the components can be supported by 83mA?

It has the following chips:

1 pc AMP01
1 pc 555 tiimer
1 pc G6H-2 relay
1pc LF412CP
3 pcs TLE2061AIP
1 pc LM386.
and other components like capacitors, resistors, diodes, etc.

Can they be powered by 83mA?

I have no clue. How about measuring the current?

[loop123]

Thanks. The ISO122 is powered by an isolated DC-DC converter called  DCP010512DBP see datasheet at  https://rocelec.widen.net/view/pdf/6mnzrokr1h/sbvs012f.pdf?t.download=true&u=5oefqw

It has input of 5 Volts and output of 12 Volts 83mA. However my ISO-Z isolation head stage and the main unit is only powered by batteries. 12 pcs of 1.2v rechargeable NiMH batteries for total of 1.2V x 12 = 14.4V, but since the main unit uses V+ and V- then it got it from the following configuration where each V+ is 7.2V and V- is -7.5V. I never use any AC-DC adaptor so no problem about isolating ground of AC.

You don't need an ISO-amp or a DC-DC converter at all. The biomed amp is powered from batteries, is not earth referenced and already floating.

Most uses of Isolated DC-DC Converter is to isolate the main power ground. But my use is to protect from 90mA from batteries. A sailor who used multimeter died from 9V battery after it reach 90mA (maybe because of defective multimeter and he punctured his skin), see https://darwinawards.com/darwin/darwin1999-50.html It's in the official Navy guidelines, true?

Dying from a 9V block battery? That's BS. A fairytale probably told to gullible new navy cadets for some reason. 

Do you think the all the components can be supported by 83mA?

It has the following chips:

1 pc AMP01
1 pc 555 tiimer
1 pc G6H-2 relay
1pc LF412CP
3 pcs TLE2061AIP
1 pc LM386.
and other components like capacitors, resistors, diodes, etc.

Can they be powered by 83mA?

I have no clue. How about measuring the current?

But remember when measuring using ECG or EEG. the skin is abraded to lower the resistance (from average surface resistance of 50kOhm or more). What if there is wound not seen, then the resistance can go to 100 Ohm. and using the 9V battery example...  I =V/R = 9/100Ohm = 90mA. If you use 15V battery. Even high amps. What if it gets into contact with the electrode leads for example, an insect crawl inside the unit and Vs got shorted to electrode.  Why is this not caused for concern?

Btw.  When I used the main amp directly injecting it with 90Hz 5mV signal generator using Netech ECG simulator. It is ok with switch of  100Hz or even 1000Hz in main amp. But when I switched it to 3000Hz filter. I noticed the sine wave is more jagged line. I can't decide if it is the sine wave generator that can do that. Does sine wave generator also produce noise in the component with uniform high frequency noise even if the sine wave is only 90Hz?  Or is it the main amp noise? But at only 3000Hz. Why does noise come out already? It used the low noise AMP01 Instrumentation amplifier.. See    https://www.analog.com/media/en/technical-documentation/data-sheets/amp01.pdf                 the noise seems to be at nV only.   What is the real noise of AMP01 at 3000Hz?

[Andy Chee]

Most uses of Isolated DC-DC Converter is to isolate the main power ground. But my use is to protect from 90mA from batteries. A sailor who used multimeter died from 9V battery after it reach 90mA (maybe because of defective multimeter and he punctured his skin), see https://darwinawards.com/darwin/darwin1999-50.html It's in the official Navy guidelines, true?

Dying from a 9V block battery? That's BS. A fairytale probably told to gullible new navy cadets for some reason. 

But remember when measuring using ECG or EEG. the skin is abraded to lower the resistance (from average surface resistance of 50kOhm or more). What if there is wound not seen, then the resistance can go to 100 Ohm. and using the 9V battery example...  I =V/R = 9/100Ohm = 90mA. If you use 15V battery. Even high amps. What if it gets into contact with the electrode leads for example, an insect crawl inside the unit and Vs got shorted to electrode.  Why is this not caused for concern?

You need to think about battery internal impedance.  You can test this yourself!  Put a 100ohm resistor between +ve and -ve and measure the battery voltage, and watch it go down.  Lower voltage means lower current.

[loop123]

Most uses of Isolated DC-DC Converter is to isolate the main power ground. But my use is to protect from 90mA from batteries. A sailor who used multimeter died from 9V battery after it reach 90mA (maybe because of defective multimeter and he punctured his skin), see https://darwinawards.com/darwin/darwin1999-50.html It's in the official Navy guidelines, true?

Dying from a 9V block battery? That's BS. A fairytale probably told to gullible new navy cadets for some reason. 

But remember when measuring using ECG or EEG. the skin is abraded to lower the resistance (from average surface resistance of 50kOhm or more). What if there is wound not seen, then the resistance can go to 100 Ohm. and using the 9V battery example...  I =V/R = 9/100Ohm = 90mA. If you use 15V battery. Even high amps. What if it gets into contact with the electrode leads for example, an insect crawl inside the unit and Vs got shorted to electrode.  Why is this not caused for concern?

You need to think about battery internal impedance.  You can test this yourself!  Put a 100ohm resistor between +ve and -ve and measure the battery voltage, and watch it go down.  Lower voltage means lower current.

I read this thread where the professor himself (IanB)  wrote:

https://www.eevblog.com/forum/beginners/eneloop-aa-battery-short-circuit-current-and-other-questions/



I used the same 1.2V Eneloop rechargeable batteries NiMH 2500mAH. He wrote it could even reach 24A! 90mA and above could already be lethal. So I cant dare short the battery. I guess it is one of those "Do no try this at home".  Please share what current can go to your finger if  you have micro wounds and your resistance can get 100 Ohm. If the voltage is say 15V D.C.. It can reach I=V/R = 15V/100ohm = 150mA enough to stop your heart.  Please share the counterarguments because it seems I  heard little of electronic enthusiasts who got electrocuted by their circuit powered by batteries only.

Also if I'd put the resistor in series with the resistor. Let's say it is to avoid amperage greater than 5mA. Then the resistor has to be R=V/I = 15/0.005 = 3kOhm? Should it be put in all 3 input wires (+, -, ground/common) of the bioamp?

How do you know if the existing device has resistor in series in the input already? how to measure the presence of any resistor and values? Thanks.

[RFDx]

But remember when measuring using ECG or EEG. the skin is abraded to lower the resistance (from average surface resistance of 50kOhm or more). What if there is wound not seen, then the resistance can go to 100 Ohm. and using the 9V battery example...  I =V/R = 9/100Ohm = 90mA. If you use 15V battery. Even high amps. What if it gets into contact with the electrode leads for example, an insect crawl inside the unit and Vs got shorted to electrode.  Why is this not caused for concern?

You're confusing the impedance between electrode and skin, which is imo irrelevant in this context and at best a few kOhms considering wet electrodes, and the high to very high impedance between two arbitrary points on the human body. Never ever would a 9V potential difference be able to drive 90mA through the body nor can a few Volts more (15V) drive "high amps". Who told you this nonsense?

Btw.  When I used the main amp directly injecting it with 90Hz 5mA signal generator using Netech ECG simulator. It is ok with switch of  100Hz or even 1000Hz in main amp. But when I switched it to 3000Hz filter. I noticed the sine wave is more jagged line. I can't decide if it is the sine wave generator that can do that.

What is there to decide? It's very easy to find out by measuring the signal generator on it's own.
By the looks of it, the sinwave is generated digitally and then converted to analog by a low resolution DAC. A proper reconstruction filter also seems to be missing. You can even count the discrete steps on the sinewave.

[Andy Chee]

How do you know if the existing device has resistor in series in the input already? how to measure the presence of any resistor and values? Thanks.

You have already identified the resistors here:

[Andy Chee]

How do you know if the existing device has resistor in series in the input already? how to measure the presence of any resistor and values? Thanks.

You have already identified the resistors here:

No. The above was the schematic of the ISO-Z where the ISO122P is located. It's not the main amp which I haven't trace yet.

If you want electric shock protection, then use the same resistor connection as the schematic.  That's how the unit is designed for safety.

[Andy Chee]

How do you know if the existing device has resistor in series in the input already? how to measure the presence of any resistor and values? Thanks.

You have already identified the resistors here:

No. The above was the schematic of the ISO-Z where the ISO122P is located. It's not the main amp which I haven't trace yet.

If you want electric shock protection, then use the same resistor connection as the schematic.  That's how the unit is designed for safety.

My bad. The pcb in my last message was still the ISO-Z. The following is the main amp pcb I haven't traced yet. Maybe it already has the resistor?

As the two units are designed to connect together and work together, I am going to guess no.  In other words, the ISO-Z does all the safety protection.

[loop123]

How do you know if the existing device has resistor in series in the input already? how to measure the presence of any resistor and values? Thanks.

You have already identified the resistors here:

No. The above was the schematic of the ISO-Z where the ISO122P is located. It's not the main amp which I haven't trace yet.

If you want electric shock protection, then use the same resistor connection as the schematic.  That's how the unit is designed for safety.

In the ISO-Z.  The resistors have color code yellow-white-white-brown (which you can clearly see in the pcb, the blue resistors in bottom left). which is code for 4990 Ohm. Now if the input is 15V. the current is 15V/4990 = 0.003.. or 3mA. So you are right. it may be designed just to limit it to 3mA. But why didn't they just use 5000 Ohm instead of 4990 Ohm.

The circuit of the main amp is complicated. If there is also a resistor inside, then it would become series to existing and become 15/10k = 1.5mA. But wait, at 1mV input differential voltage, what is the current?

Btw.. the main amp is for use in rats. Yes. It's a biopotential for use in rats. They designed the ISO-Z for use in humans. But since I'd use batteries only. I can use the main amp without the ISO-Z. That's why I have to put safety in place because it made for rats!

[loop123]

But remember when measuring using ECG or EEG. the skin is abraded to lower the resistance (from average surface resistance of 50kOhm or more). What if there is wound not seen, then the resistance can go to 100 Ohm. and using the 9V battery example...  I =V/R = 9/100Ohm = 90mA. If you use 15V battery. Even high amps. What if it gets into contact with the electrode leads for example, an insect crawl inside the unit and Vs got shorted to electrode.  Why is this not caused for concern?

You're confusing the impedance between electrode and skin, which is imo irrelevant in this context and at best a few kOhms considering wet electrodes, and the high to very high impedance between two arbitrary points on the human body. Never ever would a 9V potential difference be able to drive 90mA through the body nor can a few Volts more (15V) drive "high amps". Who told you this nonsense?

So the impedance requirements of 5kOhm (from 50kOhm using abrasive gel) is between electrodes and skin. But if there is a wound between the 2 electrodes. What is the lowest resistance the internal body can have, can't it be 100 Ohm? In the Darwinian article, it says I=V/R=9Volt/100 = 90mA.
 

Btw.  When I used the main amp directly injecting it with 90Hz 5mA signal generator using Netech ECG simulator. It is ok with switch of  100Hz or even 1000Hz in main amp. But when I switched it to 3000Hz filter. I noticed the sine wave is more jagged line. I can't decide if it is the sine wave generator that can do that.

What is there to decide? It's very easy to find out by measuring the signal generator on it's own.
By the looks of it, the sinwave is generated digitally and then converted to analog by a low resolution DAC. A proper reconstruction filter also seems to be missing. You can even count the discrete steps on the sinewave.

Oh. I didn't post the 1000 Hz noise and 3000 Hz noise when the Netech simulator is connected directly to the main amp (without using the ISO-Z).

In the following, the 1000Hz has less noise. The 3000Hz is noisy. You are saying both are made of discrete steps using low resolution DAC. I haven't thought of that. Thanks. But why is it that at 3000Hz. the discrete steps show more?

1000Hz:



3000Hz:



This is important because as I wrote in the message to Andy. I mentioned there is an INA114AP just before the ISO122 and I was able to use resistor to produce gain of 100 just like you said. And it has no visible broadband noise at 1000Hz. But at 3000Hz. There is a noise which seemed to be from the discrete steps. When I used the Netech directly into the main amp. There is the above discrete step noise at 3000Hz showing more than the 1000Hz. Again why? Note the LF411CP, 6800pf, 3300pf producing cutoff of 4.65kZ still has the 3kHz inside it so it shouldn't be producing broadband noise at 3000Hz.

With just batteries powering the main amp and ISO-Z. The ISO-Z may not be useful anymore except for the resistors which I can just add in the main amp. But just to be sure the ISO-Z can just add noise paths to the signal. What noises are in the isolated DC-DC converter. I heard it can produce noises and some avoid it like the plagues, but I haven't seen the noise. It is what can overwhelming make me not use the ISO-Z anymore. Many thanks!

[Andy Chee]

But remember when measuring using ECG or EEG. the skin is abraded to lower the resistance (from average surface resistance of 50kOhm or more). What if there is wound not seen, then the resistance can go to 100 Ohm. and using the 9V battery example...  I =V/R = 9/100Ohm = 90mA. If you use 15V battery. Even high amps. What if it gets into contact with the electrode leads for example, an insect crawl inside the unit and Vs got shorted to electrode.  Why is this not caused for concern?

You're confusing the impedance between electrode and skin, which is imo irrelevant in this context and at best a few kOhms considering wet electrodes, and the high to very high impedance between two arbitrary points on the human body. Never ever would a 9V potential difference be able to drive 90mA through the body nor can a few Volts more (15V) drive "high amps". Who told you this nonsense?

So the impedance requirements of 5kOhm (from 50kOhm using abrasive gel) is between electrodes and skin. But if there is a wound between the 2 electrodes. What is the lowest resistance the internal body can have, can't it be 100 Ohm? In the Darwinian article, it says I=V/R=9Volt/100 = 90mA.
 

The lowest resistance is dependent on electrode separation and applied voltage. 

But for reference, the IEC states that at 25 volts dry skin, the hand-to-hand resistance for 95% of subjects it is <6100 ohms, for 50% of human test subjects is <3250 ohms, and for 5% of subjects it is <1750 ohms.

[loop123]

But remember when measuring using ECG or EEG. the skin is abraded to lower the resistance (from average surface resistance of 50kOhm or more). What if there is wound not seen, then the resistance can go to 100 Ohm. and using the 9V battery example...  I =V/R = 9/100Ohm = 90mA. If you use 15V battery. Even high amps. What if it gets into contact with the electrode leads for example, an insect crawl inside the unit and Vs got shorted to electrode.  Why is this not caused for concern?

You're confusing the impedance between electrode and skin, which is imo irrelevant in this context and at best a few kOhms considering wet electrodes, and the high to very high impedance between two arbitrary points on the human body. Never ever would a 9V potential difference be able to drive 90mA through the body nor can a few Volts more (15V) drive "high amps". Who told you this nonsense?

So the impedance requirements of 5kOhm (from 50kOhm using abrasive gel) is between electrodes and skin. But if there is a wound between the 2 electrodes. What is the lowest resistance the internal body can have, can't it be 100 Ohm? In the Darwinian article, it says I=V/R=9Volt/100 = 90mA.
 

The lowest resistance is dependent on electrode separation and applied voltage. 

But for reference, the IEC states that at 25 volts dry skin, the hand-to-hand resistance for 95% of subjects it is <6100 ohms, for 50% of human test subjects is <3250 ohms, and for 5% of subjects it is <1750 ohms.

I read the resistance between ears is 100 Ohms. What would happen when one wires a 9V or 12V battery to the ears, it should produce 12/100Ohms = 120mA current to his brain? Would you feel a zap? 



https://www.asc.ohio-state.edu/physics/p616/safety/fatal_current.html#:~:text=It's%20The%20Current%20That%20Kills&text=Individuals%20have%20been%20electrocuted%20by,as%2042%20volts%20direct%20current

[MrAl]

Yes, it ran. But do you see the vertical scale?

At -160dB you are simulating parasitic coupling through the feedback network or things like that...

There is a V1, AC1 in the left side. What is it? I thought it was the power supply. Where and how do you add the power supply? I don't know how to use LTspice Pls download the file in original thread and add it and upload it back so I can run the frequency response plot. Thanks.

V1 is the input signal.

LT Spice, in their wisdom, always use a DC generator symbol with + & - polarity markings on it, unlike the ac generator symbol which I am familiar with from decades of seeing it used.
That is why LTSpice annoys me so much.

If the generator is so marked, it is easy for someone to think it is a DC power supply source.

Just to note, AC generators also must have a polarity.  That's because they have a specific phase reference.
If you flip an AC generator, you reverse the phase so it would be 180 degrees out of phase with an AC generator that is not flipped.
If there were no polarity markings, you would not be able to tell which phase it was.

[MrAl]

Consider this Sullen-Key low pass filter.

(Attachment Link)

If you change the amplifier to LF411CP (with the same pins). How much would the frequency response change?

When I increase the value of the capacitor from 1000pF to 6800pF, the frequency get lowered. Why is that?

I found a Ltspice file using the third order filter (see attached). I changed it to the 2nd order Sullen-key by deleting some parts and changing the values. Please check if the entry is correct. Also the voltage source is reverse. This is ok for AC, isn't it? Second. What amplifier is in the Ltspice attached Ltspice file?. How do you change it to the LF411CP? I want to see if the frequency response would change. If it's hard to change. Please change it to LF411CP and attached the edited Ltspice.

(Attachment Link)

(Attachment Link)

What does ".ac oct 20 10 10000" mean? If I removed it, I can't run the simulator anymore.

Thank you.

When you use an op amp you also have to think about the slew rate.  The minimum slew rate for an op amp is:
sr=A*w
where
sr is the slew rate in volts per second,
A is the maximum peak amplitude of the sine wave,
w is the angular frequency 2*pi*f, with f in Hertz.

[Andy Chee]

I read the resistance between ears is 100 Ohms. What would happen when one wires a 9V or 12V battery to the ears, it should produce 12/100Ohms = 120mA current to his brain? Would you feel a zap? 

https://www.asc.ohio-state.edu/physics/p616/safety/fatal_current.html#:~:text=It's%20The%20Current%20That%20Kills&text=Individuals%20have%20been%20electrocuted%20by,as%2042%20volts%20direct%20current

Please examine the voltages on your picture of the human body.  It is 110 volts, not 9 volts. 

The resistance of the human body does not behave like a standard component resistor.  At 9 volts, the human body resistance will be higher than your diagram.

[loop123]

I read the resistance between ears is 100 Ohms. What would happen when one wires a 9V or 12V battery to the ears, it should produce 12/100Ohms = 120mA current to his brain? Would you feel a zap? 

https://www.asc.ohio-state.edu/physics/p616/safety/fatal_current.html#:~:text=It's%20The%20Current%20That%20Kills&text=Individuals%20have%20been%20electrocuted%20by,as%2042%20volts%20direct%20current

Please examine the voltages on your picture of the human body.  It is 110 volts, not 9 volts. 

The resistance of the human body does not behave like a standard component resistor.  At 9 volts, the human body resistance will be higher than your diagram.

Are you saying that for any ECG equipment this is battery operated There is 0% risk? No scenerio at all it can be dangerous? For example if there are wounds in the chest and the electrodes got short to Vs. There will still be high resistance at all time in the chest inches from heart even with micro wounds at skin surface? But if the dc voltages keep getting higher. what dc voltages would make it hazardous?
 

[Andy Chee]

I read the resistance between ears is 100 Ohms. What would happen when one wires a 9V or 12V battery to the ears, it should produce 12/100Ohms = 120mA current to his brain? Would you feel a zap? 

https://www.asc.ohio-state.edu/physics/p616/safety/fatal_current.html#:~:text=It's%20The%20Current%20That%20Kills&text=Individuals%20have%20been%20electrocuted%20by,as%2042%20volts%20direct%20current

Please examine the voltages on your picture of the human body.  It is 110 volts, not 9 volts. 

The resistance of the human body does not behave like a standard component resistor.  At 9 volts, the human body resistance will be higher than your diagram.

Are you saying that for any ECG equipment this is battery operated There is 0% risk? No scenerio at all it can be dangerous? For example if there are wounds in the chest and the electrodes got short to Vs. There will still be high resistance at all time in the chest inches from heart even with micro wounds at skin surface? But if the dc voltages keep getting higher. what dc voltages would make it hazardous?

Please study the document standard IEC 60601 which relates to medical equipment.  This standard contains many of the answers to your questions regarding medical equipment electrical safety.

https://webstore.iec.ch/preview/info_iec60601-2-26%7Bed3.0%7Db.pdf
https://numlor.fr/tp/docs/IEC60601-1.pdf

https://en.wikipedia.org/wiki/IEC_60601

IEC 60601-2-25 Medical electrical equipment - Part 2-25: Particular requirements for the basic safety and essential performance of electrocardiographs
IEC 60601-2-26 Medical electrical equipment - Part 2-26: Particular requirements for the basic safety and essential performance of electroencephalographs
IEC 60601-2-40 Medical electrical equipment - Part 2-40: Particular requirements for the basic safety and essential performance of electromyographs and evoked response equipment

[loop123]

RFdx. To illustrate what I meant below. Here are the waveforms using the ISO-Z with the 100gain before the ISO122. There is no more bandwidth noise at 1000Hz. But at 3000Hz the discrete steps can be seen just like using the sine wave generator directly on the main amp. Maybe because of increased resolution that makes the steps more visible?





Setting is 90Hz, 1mV in Netech simulator. Gain in main amp is set to 50X because 100X can make it clip.
For the 1mV input, and gain of 100X before ISO122. The voltage in the output of the ISO-Z is 0.1V. With 50X gain in main unit. It's like the maximum voltage it can display without clipping is about 0.1x50 = 5Volt. I wonder if this is the limitation of Audacity or the main unit. I can't use 1000 gain before the ISO122 because with ISO-Z 1Volt output and 10X mininum gain in the main unit, 10V will clip, or maybe there is setting in Audacity to make it display 10V?

But where are the infamous noises of the isolated DC-DC converter in all the waveforms? If it doesn't appear, maybe the single 0.1uF capacitor at Vs of the isolated DC-DC converter removed all noises? If it's that simple, why so many complains about the noises of isolated DC-DC converters in general?

But remember when measuring using ECG or EEG. the skin is abraded to lower the resistance (from average surface resistance of 50kOhm or more). What if there is wound not seen, then the resistance can go to 100 Ohm. and using the 9V battery example...  I =V/R = 9/100Ohm = 90mA. If you use 15V battery. Even high amps. What if it gets into contact with the electrode leads for example, an insect crawl inside the unit and Vs got shorted to electrode.  Why is this not caused for concern?

You're confusing the impedance between electrode and skin, which is imo irrelevant in this context and at best a few kOhms considering wet electrodes, and the high to very high impedance between two arbitrary points on the human body. Never ever would a 9V potential difference be able to drive 90mA through the body nor can a few Volts more (15V) drive "high amps". Who told you this nonsense?

So the impedance requirements of 5kOhm (from 50kOhm using abrasive gel) is between electrodes and skin. But if there is a wound between the 2 electrodes. What is the lowest resistance the internal body can have, can't it be 100 Ohm? In the Darwinian article, it says I=V/R=9Volt/100 = 90mA.
 

Btw.  When I used the main amp directly injecting it with 90Hz 5mA signal generator using Netech ECG simulator. It is ok with switch of  100Hz or even 1000Hz in main amp. But when I switched it to 3000Hz filter. I noticed the sine wave is more jagged line. I can't decide if it is the sine wave generator that can do that.

What is there to decide? It's very easy to find out by measuring the signal generator on it's own.
By the looks of it, the sinwave is generated digitally and then converted to analog by a low resolution DAC. A proper reconstruction filter also seems to be missing. You can even count the discrete steps on the sinewave.

Oh. I didn't post the 1000 Hz noise and 3000 Hz noise when the Netech simulator is connected directly to the main amp (without using the ISO-Z).

In the following, the 1000Hz has less noise. The 3000Hz is noisy. You are saying both are made of discrete steps using low resolution DAC. I haven't thought of that. Thanks. But why is it that at 3000Hz. the discrete steps show more?

1000Hz:



3000Hz:



This is important because as I wrote in the message to Andy. I mentioned there is an INA114AP just before the ISO122 and I was able to use resistor to produce gain of 100 just like you said. And it has no visible broadband noise at 1000Hz. But at 3000Hz. There is a noise which seemed to be from the discrete steps. When I used the Netech directly into the main amp. There is the above discrete step noise at 3000Hz showing more than the 1000Hz. Again why? Note the LF411CP, 6800pf, 3300pf producing cutoff of 4.65kZ still has the 3kHz inside it so it shouldn't be producing broadband noise at 3000Hz.

With just batteries powering the main amp and ISO-Z. The ISO-Z may not be useful anymore except for the resistors which I can just add in the main amp. But just to be sure the ISO-Z can just add noise paths to the signal. What noises are in the isolated DC-DC converter. I heard it can produce noises and some avoid it like the plagues, but I haven't seen the noise. It is what can overwhelming make me not use the ISO-Z anymore. Many thanks!

[RFDx]

RFdx. To illustrate what I meant below. Here are the waveforms using the ISO-Z with the 100gain before the ISO122. There is no more bandwidth noise at 1000Hz. But at 3000Hz the discrete steps can be seen just like using the sine wave generator directly on the main amp. Maybe because of increased resolution that makes the steps more visible?

The switchable filter in the main unit acts as reconstruction filter for the digitally generated, low samplerate sinewave from the Netech simulator. The sinewave becomes smoother as you lower the bandwidth.

Setting is 90Hz, 1mV in Netech simulator. Gain in main amp is set to 50X because 100X can make it clip.
For the 1mV input, and gain of 100X before ISO122. The voltage in the output of the ISO-Z is 0.1V. With 50X gain in main unit. It's like the maximum voltage it can display without clipping is about 0.1x50 = 5Volt. I wonder if this is the limitation of Audacity or the main unit. I can't use 1000 gain before the ISO122 because with ISO-Z 1Volt output and 10X mininum gain in the main unit, 10V will clip, or maybe there is setting in Audacity to make it display 10V?

Audacity is just a software. The limitation lies in the soundcard used to make the measurements. The max. input voltage the soundcard can handle is unknown.

But where are the infamous noises of the isolated DC-DC converter in all the waveforms? If it doesn't appear, maybe the single 0.1uF capacitor at Vs of the isolated DC-DC converter removed all noises? If it's that simple, why so many complains about the noises of isolated DC-DC converters in general?

The noise of the DC-DC converter is already quite low, even with small value decoupling capacitors. As the switching frequency of the converter is very high (400kHz) and there are multiple active filters with low cuttoff frequencies fitted to the ISO-Z/main amp, you won't see any noise at the output.

[loop123]

RFdx. To illustrate what I meant below. Here are the waveforms using the ISO-Z with the 100gain before the ISO122. There is no more bandwidth noise at 1000Hz. But at 3000Hz the discrete steps can be seen just like using the sine wave generator directly on the main amp. Maybe because of increased resolution that makes the steps more visible?

The switchable filter in the main unit acts as reconstruction filter for the digitally generated, low samplerate sinewave from the Netech simulator. The sinewave becomes smoother as you lower the bandwidth.

Setting is 90Hz, 1mV in Netech simulator. Gain in main amp is set to 50X because 100X can make it clip.
For the 1mV input, and gain of 100X before ISO122. The voltage in the output of the ISO-Z is 0.1V. With 50X gain in main unit. It's like the maximum voltage it can display without clipping is about 0.1x50 = 5Volt. I wonder if this is the limitation of Audacity or the main unit. I can't use 1000 gain before the ISO122 because with ISO-Z 1Volt output and 10X mininum gain in the main unit, 10V will clip, or maybe there is setting in Audacity to make it display 10V?

Audacity is just a software. The limitation lies in the soundcard used to make the measurements. The max. input voltage the soundcard can handle is unknown.

But where are the infamous noises of the isolated DC-DC converter in all the waveforms? If it doesn't appear, maybe the single 0.1uF capacitor at Vs of the isolated DC-DC converter removed all noises? If it's that simple, why so many complains about the noises of isolated DC-DC converters in general?

The noise of the DC-DC converter is already quite low, even with small value decoupling capacitors. As the switching frequency of the converter is very high (400kHz) and there are multiple active filters with low cuttoff frequencies fitted to the ISO-Z/main amp, you won't see any noise at the output.

We were talking about milliVolt input mostly prior. But now I'm focusing on microVolt input signal so I can zoom in the milliVolt to microVolt at will when changing between EMG and EEG in controlling the augmented cytoskeleton. You said above the noise of DC-DC converter is quite low. But this is with respect to milliVolt. In terms of microVolt, what is the noise of the DC-DC converter? And what kind of noise waveform to look for?

I plan to just use the DC-DC converter directly on the battery pack of the main amp. So it's bypassing those "multiple active filters with low cuttoff frequencies fitted to the ISO-Z/main amp".

Btw.. I received the DIP version of the OPA602 for the 2 pole Sallen-key but don't seem to need it now. But I'll try it when my 2nd sets of units arrived. I bought the duplicate to experiment on upgrading the main amp. But it seems the conclusions of the amp expert is there may be nothing significant to replace it with. Maybe you know something better than the AMP01. Thanks.

[Andy Chee]

But now I'm focusing on microVolt input signal so I can zoom in the milliVolt to microVolt at will when changing between EMG and EEG in controlling the augmented cytoskeleton.

If you are not doing so already, I suggest you begin familiarising yourself with existing BCI methodology.  You don't have to reinvent the wheel, you can learn from other people's experiments.

Use the academic search engine  https://scholar.google.com  to look for relevant scientific papers, like those attached below.

[MrAl]

Consider this Sullen-Key low pass filter.

(Attachment Link)

If you change the amplifier to LF411CP (with the same pins). How much would the frequency response change?

When I increase the value of the capacitor from 1000pF to 6800pF, the frequency get lowered. Why is that?

I found a Ltspice file using the third order filter (see attached). I changed it to the 2nd order Sullen-key by deleting some parts and changing the values. Please check if the entry is correct. Also the voltage source is reverse. This is ok for AC, isn't it? Second. What amplifier is in the Ltspice attached Ltspice file?. How do you change it to the LF411CP? I want to see if the frequency response would change. If it's hard to change. Please change it to LF411CP and attached the edited Ltspice.

(Attachment Link)

(Attachment Link)

What does ".ac oct 20 10 10000" mean? If I removed it, I can't run the simulator anymore.

Thank you.

Hi,

The most obvious answer is that this depends on the operating frequency vs the power bandwidth of the op amp.
The power bandwidth is a lower frequency number than the gain bandwidth product and that is because the slew rate comes into effect also.
If you do not want to calculate this then a very rough guide is to divide the bandwidth by 30 to 50.
For example, an LM358 has a gain bandwidth of 1MHz, but its power bandwidth is only around 30kHz with a 5v peak AC output.
It does depend on the output voltage amplitude too though, so if the output amplitude was lower the power bandwidth would go up.

If you choose a higher bandwidth op amp then it is not likely to alter the frequency response, but then you have to think about oscillations.

In short, op amps have their own frequency response and to know the effects of changing one out for another you have to do a few little calculations.  The power bandwidth calculation is the most important or else you could get a lot of distortion on the output.

The maximum frequency for any or the usual op amps is:
Fmax=2*pi*A
where
A is the peak amplitude of the output voltage of the op amp in volts.

It's interesting that the peak output of the op amp is more important to know than the gain bandwidth product.  That's because the slew rate is usually the limiting factor not the gain bandwidth product, and the slew rate works with the output amplitude in determining the max allowable operating frequency, so you also have to determine what your max output voltage peak will be.