How to design a single supply EMG, EKG, or EEG circuit

[@mohankaushik2943]

Almost 5 and half years later I still find this is the most useful video for biological interfaces. Thank you man.

[@jakobhalskov]

Wow, that intro got me.. I remember watching this 8 years ago and I am now back for making my own EOG circuit for REM sleep tracking :) Thanks Jimmy

[@andriykhrupa]

How is the progress on the EOG circuit?

[@Tubesn4ke]

Nice combination of detail and brevity, thanks for the info

[@freeelectron8261]

Sallen-Key second order low-pass filter is the active filter shown. Great tut JD - thanks!

[@relaxationTherapy001]

whenever you are getting uncontrolled amount of DC offsets with your main AC signal, I think the best idea is to offsetting it with another opamp before sending the signal to the next stage of the filters. :)

[@polymath4402]

Thank you, this was exactly what I was looking for.

[@pguillem3]

Great video man. You look kinda worried on half the vid (lol)... Just wanted to add that the opamps don't have positive or negative pins... They are called inverting and non-inverting inputs. I'm not going to screw your great effort with a "wiseass electro-geek comment" hahaha. keep up the damn good work, get better every day, accept debates and never stop sharing your growing wisdom. Cheers from Universidad el Bosque (Colombia)

[@BPMbiosignals]

With a single supply that circuit presented won't work because the input range of the INA118 is from (V-) + 0.95 to (V+) - 0.65. Having V- clamped to GND means that your ECG electrodes need to generate a DC potential of about 1V to make this setup work. It's not enough to shift the output but you need to make sure that the inputs are also within the valid operating rage of the amp. You could, example, inject a bias voltage into the subject to bring the active electrodes within the operating range. It's usually called right leg drive and is also used to reduce 50/60Hz interference. Personally I would just go for negative supply with a charge pump though...

[@jpdieffe]

Thanks for the input! My latest design does have a right leg drive in addition to a negative supply from a charge pump. I was having difficulty getting my charge pumps to work at the time of this video due to a large impedance at my solder joints (I was using tin finished PCBs). I noted a DC offset on my electrodes at this time (just low frequency noise I presume), which is likely why this circuit worked well for me!

[@BPMbiosignals]

in case you haven't got luck with a charge pump you can also get an (isolated) pos/neg supply with an NMA0505 from a single 5V supply.

[@anshultyagi3684]

BPM biosignals how about shifting the entire ground level to some intermediate voltage say 2.5V for a 5V circuit using tle2426

[@anshultyagi3684]

Jimmy Dieffenderfer how does large impedence solder joint affect charge pump circuit??

[@BPMbiosignals]

What's missing in the drawing is the 3rd electrode anyway which connects to GND and establishes the common mode voltage. You could lift that 3rd electrode to 2.5V if you wish.

[@Jetelametbienprofond]

Thank you for the video :)

[@jshscott]

dood u r like so smart and stuff, teach me yo ways and i wanna b jus like u when i growed up, stay coo homie

[@beefyogurt]

its been 7 years, have you grown up coo? xD

[@thehandsomenipple3623]

dark lol

[@thehandsomenipple3623]

dark apparently he’s too coo for youtube now

[@Gaayathri_Ganesh]

Does it matter what order you put the filters? I.e high pass to low pass VS. Low pass to high pass.

[@rahardikramadibrata8675]

Hi, i just wonder why dont u use band pass filter instead of lpf and hpf ? and why do u use non-inverting amp instead of inverting amp ? thank you

[@cyrusIIIII]

Thank you for the video here I have some questions: 1.When you pre amplify it, basically you amplify noise too. Why that does not create problems? I say this because you said for the second amplifier we put it after filter. why here it is not the case for the first amplifier? 2.There are some commercial signal conditioning modules out there. Are they similar to yours? 3.why did you have to use high pass filter. Low pass filter make sense to me but after that why do we need high pass? I assume after the first filter there is no high frequency left to be cut off. 4.If the signals are in nano volt order, or picoamp, then what would be your strategy? I think the range of EMG voltage is uV to mV, my question is what if it was nVto uV?

[@pguillem3]

You have to. If you plug the filters directly to a microvolt signal without preamp, the source signal will be so weak it will likely be below the filter's amplitude threshold... You will be effectively listening for the filter's noise instead of the target signal.... First you amplify the whole original signal and then you filter whatever you need from it.

[@pguillem3]

Question number two... Joining two filters (highpass/lowpass) is done to build a BANDpass filter... You are only interested in frequencies between a certain range... Not the rest.

[@cyrusIIIII]

Pedro Guillem Thank you but what is the difference between PreAmp and Amp itself? both theoretically and in action? For example can you use the same AD741 to do both amplifying and preamplifying?

[@pguillem3]

+Silas Saw excellent question. The "Pre" is just a convention word relating to the fact the amplification is being performed before the important stuff happens. Amp would be the fact you are amplifying a signal which already has some process in it (filtering, for instance). Each step of the process is called a "stage".. Normally you do Preamp->Filtering->Amp->Process. The 741 and Ad620 are very cheap noisy amps and in this field of electronics, price means reliability. I would go for the OP series for filters, INAs are very good indeed... And for christ sakes use an isolation amp in your circuit before hooking yourself to this things, such as the ISO124 or an optocoupler... You don't want to be your computer's ground!!!

[@cyrusIIIII]

Pedro Guillem Thank you I understand now. What is Isolation ampt exactly? I remember I used optocoupler with relay to control switching off and on a pump. Would you please explain which part of my circuit I should use the ISO124 or the optocoupler. I am a noob. I do these stuff for hobby.

[@elle7272]

Hi! I want to design and make an EEG circuit that detects brain frequencies (alpha, beta, theta, gamma, delta) but I have no knowledge or experience of coding, circuits or anything! Can you please help me or give me some tips and advice to design a circuit!!

[@sohaibhayy7105]

Hi! I wanted to make a EEG circuit that can filter for theta waves which are between 4-7 Hz. Suggestions how to make that? I was thinking of doing digital filtering but i dont know how to do that. also will the analog part be similar?

[@jpdieffe]

The analog part will be similar. As a rule of thumb, you want to get rid of most low frequency and high frequency noise before you "digitize" this signal. The reason for this is that you can easily apply REAL gain after analog filtering. The most basic way to digitally filter is using matlab. If you want something that is doing this in real time, I would have LabVIEW acquiring your data and then running a matlab filtering script every 10 seconds. Good luck

[@sohaibhayy7105]

Jimmy Dieffenderfer so digitizing is used to make it more accurate because i believe that analog has some limitations and so it will not be filtering something just between 4-7. instead could be doing something like 2-10 or something...would you agree with me on that?

[@sohaibhayy7105]

did you use a simple lf347 in the opamps and how would you digitally filter the signal? FFT? I dont know how to do that

[@jpdieffe]

sohaib hayy You're on the right track. Digital filters are used because they can be easily adjusted. So you would have analog filters keeping the range of 1Hz to 150Hz (for example) and then have a digital filter to keep 4 to 7.

[@jpdieffe]

sohaib hayy You can make a digital filter by going through the whole process of switching to the frequency domain and multiplying by a filter... Or you could just use a couple lines of code in Matlab. Do a search for digital filters in matlab.

[@haddadmarcelo]

is it necessary to have the emg cables? or it will work just fine with normal wires?

[@phoenixgamescomp.7932]

Assuming that I have calculated the cut off frequencies that I want my filters to have, what about the gain of them?

[@iliahadzhiev8023]

Hi, in single supply configuration where I should connect the reference electrode ?

[@UseLogicPlease123]

Interesting filter design. Haven't run into those too often. Will have to try them out. I've always gotten away with passives or attaching Z components to my gain stages. I was confused when you were purposing the single supply design that you didn't explore a virtual ground built by simply buffering your resistor divider with an op amp and referencing your filters against that. Charge pump / SMPS would also work for a negative rail but those can be dangerous to small signals because of ripple. With 5V you have a ton of common mode range for a virtual ground. @ 3.3V it requires a lot more care.

[@juandavidarroyave3537]

Hi man, can i use the AD620?

[@victorronaldaf1514]

i think than its the same

[@mohammedadel9959]

In ECG the voltage is maximum when ventricle is half depolarized & half repolraized (Half positive & negative) ' peak of R wave in ECG' Why?! Shouldn't voltage be the same as long as there's a potential difference between charges under electrodes whether it's half or not!?? Could you explain!

[@abhiroopkaginalkar]

Hey, I have made the sensory circuit and it works fine with an oscillator generated signal but, there is some issue with the actual electrodes. Can I use standard ECG electrodes or are there any specific EMG electrodes? Just the electrodes pressing against the skin leads to a signal that might be covering up the actual EMG. Is something like that possible?

[@jpdieffe]

Electrodes can play a part in distorting your signal, although it's usually not the problem with people's circuits. With 3M electrodes, you should be able to get good results. This leads to a good point though. Having a two lead ECG/EMG is very sensitive to noise as well as changes in impedance. You will probably notice that the DC bias your signal is riding on will change as a function of the electrode pressure. This is simply because the more pressure applied to your electrode, the less impedance, the greater the signal. Not sure if this answered your question or not.

[@hareshk2]

i m having problem in getting output from ina 128 burbrown ic ... would u plz help me through this.. which one is good for getting ecg signal either ina 128 texas instruments or burbrown ???

[@10bokaj]

for the high-pass filter just use 2,5V as ref (ground), then the dc will not be removed? (why not?)

[@jpdieffe]

Jakob Furbo that won't remove the DC drift, it will just add a bias.

[@nullyyllun6374]

hi there, first of all i'm newbie to electronics and i have simple question. if i was to make this with a single battery, is my gnd also the negative of my battery ( i mean the side with the minus , i'll just have to use this as gnd?) , i was thinking of making something portable and dont want to tie it to a radiator or heater all the time.

[@jpdieffe]

Yes, the "minus" of the battery would be connected to GND in this case. In my latest video, I use a specialized IC to make a positive and negative voltage (and gnd) from a single battery.

[@freeelectron8261]

And the same second order filter in high pass.

[@rohansalunkhe9127]

Pls tell me can I use this circuit with arduino, Any one??

[@user-fh5le5fc4l]

Can u make a video for eog ??

[@safaaelwaleed8598]

Up

[@kevinsimpson6147]

Sallen-Key Filters

[@alaaalmigdad7457]

What is the purpose of the last non inverting amplifier?

[@sergiop753]

alaa almigdad it'd be some sort of final amplification. Given that if you amplify the signal too much at the beggining, you'd amplify noise too. So, that'd be it

[@rodrigoloza4263]

I will have to agree with BPM biosignals, this won't work okay. I think that you should first try to implement your idea.

[@bhuvaneshs.k638]

Cool.....!!!

[@rozoramirez]

which amplifiers did you use?

[@juandavidarroyave3537]

INA106

[@deeprj3710]

suggest me combined EEG ECG EMG CIRCUIT

[@_sadaf_]

Hey, Did you make the combined circuit?

[@sourcesource8750]

After lowpass, why sending it into high pass?

[@jpdieffe]

Source Source gets rid of drift

[@vazduhvatra7949]

Here no circuit value -this is not complete!?

[@engineerhaydar18]

hello sir :) please i have a question -- why a lot of these kind of circuits they use a/d converter and then present it on pc ? can they use an oscillator like ( usb oscillator ) --- im waiting for your answer :) thank u for ur kindness ..

[@muhamadkarzan3213]

Please please tell me who I can made my own brain erg. By clear and practical way.

[@tillatbatool5297]

you are honestly so cute :)

[@engineers6967]

kdr ka ha rah tou

[@Rich_Villian]

Is it possible to send EMG signals wirelessly?

[@phrenologisto]

It's not what youre saying it's how youre saying it. You're coming off with a professorial vibe, and introducing your video as a reason to stop emails. Maybe you created this video to stop emails but if youre looking for views, dont introduce the video. Youre not presenting to a class. Your title is descriptive enough to explain its contents and the intro is dropping viewers. Get rid of the intro and your view count will increase. People dont want to be a burden. If your introduction isnt absolutely necessary (your title doesnt describe the contents) DONT USE ONE; get on with it! Otherwise, useful video. Hope youve solved your email problem! And thanks for sharing your process

[@mohmandayaz4150]

I am also working on a exo skeleton and i have already started my designs, i never thought i would come across someone who was working on the same project as me, My plan is to use phneumatics for the limb drives and extra pistons at across the fingers, and use a mixture of carbon fibre and titanium for the frame. Pm me on facebook at Ayaz mohmand (i have over 3000 subscribers so i should be easy to find, im 17 years old