What exactly occurs in the circuit?

You'd have to ask the manufacturer. My guess would be that noise from one part of the circuit gets shunted to ground rather than passing directly to some other part of the circuit.

I queried ChatGPT about safety for your situation where there is inadequate grounding. Here's the chat:-

Me: Can mains-powered EEG equipment provide complete electrical safety to a patient if neither the electrical mains supply nor the building in which the equipment is used has electrical grounding?

ChatGPT: No, mains-powered EEG equipment cannot provide complete electrical safety to a patient if neither the electrical mains supply nor the building has proper electrical grounding.
Here's why:

Electrical Grounding is Critical for Safety

1. **Protective Earth (PE) is a Safety Feature**
Medical devices, especially those that are mains-powered and patient-connected like EEG machines, rely on a proper ground connection to safely divert any leakage currents or fault currents away from the patient and operator.

2. **Class I Medical Equipment Requires Grounding**
Many EEG machines are **Class I medical devices**, meaning they are designed to use the ground connection for safety. Without grounding:
* Leakage currents may **flow through the patient** instead of being diverted safely.
* A fault (e.g. insulation failure) could result in **electric shock**.

3. **Double Insulation (Class II Equipment) Doesn't Fully Solve the Problem**
Even if the EEG device is **Class II (double insulated)** and does not rely on grounding for electrical safety, **the patient connections (electrodes)** still pose a risk due to their direct connection to the body. Inadequate grounding increases:
* The risk of **common-mode voltages** appearing on patient leads.
* Potential interference or **voltage build-up**, risking safety and data integrity.

4. **Patient Isolation (BF/CF Rated Circuits)**
EEG systems typically have **isolated patient circuits (e.g., BF or CF applied parts)**, which help protect the patient from electrical hazards. However, these circuits **still depend on a safe system design**, which assumes a grounded reference elsewhere in the system.

### Why Grounding Matters in EEG Use

* EEG signals are in the **microvolt range**, so electrical noise (especially from ungrounded systems) can **overwhelm the signal**, making the EEG useless.
* A lack of grounding can lead to **floating voltages** and **stray currents**, which are not only dangerous but can also cause **artifacts** in EEG readings.

### Conclusion:

> **No**, a mains-powered EEG system **cannot be considered electrically safe** for the patient **without proper electrical grounding**. Operating such equipment in an ungrounded environment violates standard medical electrical safety practices (IEC 60601), and poses **a serious risk of electric shock** or faulty readings.

### Recommendations:

* **Never use patient-connected mains-powered medical devices** in ungrounded environments.
* Consider **battery-powered EEG systems** with proper isolation for use in environments without grounding.
* Ensure all medical equipment complies with **IEC 60601** standards, especially regarding leakage currents and grounding requirements.

 

You'd have to ask the manufacturer. My guess would be that noise from one part of the circuit gets shunted to ground rather than passing directly to some other part of the circuit.

I queried ChatGPT about safety for your situation where there is inadequate grounding. Here's the chat:-

Me: Can mains-powered EEG equipment provide complete electrical safety to a patient if neither the electrical mains supply nor the building in which the equipment is used has electrical grounding?

ChatGPT: No, mains-powered EEG equipment cannot provide complete electrical safety to a patient if neither the electrical mains supply nor the building has proper electrical grounding.
Here's why:

Electrical Grounding is Critical for Safety

1. **Protective Earth (PE) is a Safety Feature**
Medical devices, especially those that are mains-powered and patient-connected like EEG machines, rely on a proper ground connection to safely divert any leakage currents or fault currents away from the patient and operator.

2. **Class I Medical Equipment Requires Grounding**
Many EEG machines are **Class I medical devices**, meaning they are designed to use the ground connection for safety. Without grounding:
* Leakage currents may **flow through the patient** instead of being diverted safely.
* A fault (e.g. insulation failure) could result in **electric shock**.

3. **Double Insulation (Class II Equipment) Doesn't Fully Solve the Problem**
Even if the EEG device is **Class II (double insulated)** and does not rely on grounding for electrical safety, **the patient connections (electrodes)** still pose a risk due to their direct connection to the body. Inadequate grounding increases:
* The risk of **common-mode voltages** appearing on patient leads.
* Potential interference or **voltage build-up**, risking safety and data integrity.

4. **Patient Isolation (BF/CF Rated Circuits)**
EEG systems typically have **isolated patient circuits (e.g., BF or CF applied parts)**, which help protect the patient from electrical hazards. However, these circuits **still depend on a safe system design**, which assumes a grounded reference elsewhere in the system.

### Why Grounding Matters in EEG Use

* EEG signals are in the **microvolt range**, so electrical noise (especially from ungrounded systems) can **overwhelm the signal**, making the EEG useless.
* A lack of grounding can lead to **floating voltages** and **stray currents**, which are not only dangerous but can also cause **artifacts** in EEG readings.

### Conclusion:

> **No**, a mains-powered EEG system **cannot be considered electrically safe** for the patient **without proper electrical grounding**. Operating such equipment in an ungrounded environment violates standard medical electrical safety practices (IEC 60601), and poses **a serious risk of electric shock** or faulty readings.

### Recommendations:

* **Never use patient-connected mains-powered medical devices** in ungrounded environments.
* Consider **battery-powered EEG systems** with proper isolation for use in environments without grounding.
* Ensure all medical equipment complies with **IEC 60601** standards, especially regarding leakage currents and grounding requirements.

Many thanks for the above tips.

Can AC adaptor leak the 220V AC to the 5V DC? CHATGPT assumes the 220V is plugged into the unit. It is not as it is the 5V plug that plugs into the unit. Please tell me before we close the thread. What you mean by "screened wires" or screened input and output? Did you mean coaxial or just very thick rubber power cord?

I'll just use the 5V powerbank. If you know of commercially available 3.7V + 3.7V with regulator to make very clean 5V. Let me know. Also some BCI experiments are done in electromagnetically shielded room. It is enough for me if I can fit inside the faraday cage. So if you know a person size faraday cage, let me know. Thanks.

 

What you mean by "screened wires" or screened input and output? Did you mean coaxial or just very thick rubber power cord?

Coaxial or equivalent, where an outer conductor is grounded to screen an inner conductor. Normal mains cables are not screened.
Can't help re Faraday cages.

 

Deducting common mode noise in software is certainly an interesting theory. Also, not especially possible for all noise. Interfering signals may be fairly well reduced, but consider that the noise spectrum may not be exactly the same at both inputs, because of coming from different locations. So cancelling external noise in software is a false premise. Cancelling interfering SIGNALS is usually a different case.

Btw is there not a technique to read all ADCs at same time instead of serially? A user of the g.USBamp needs to spend $20,000 minimum. I cant believe it has to read 16 channels serially. People can notice it.

 

I cant believe it has to read 16 channels serially.

It depends what you mean by 'read'. There are 16 ADCs operating simultaneously, but there appears from the description to be only one DSP handling the conversion results.

 

This is a comment about the "chat" statements about patient safety: Consider that the input impedance is "greater than 100 MEGOHMS, how much current could flow to a connected person, even in an instance of a direct connection to the 120 volt mains, thru that resistance??? Recall that current= voltage/resistance. so 120/100,000,000 works out to be much less than ONE MICRO-AMP. Not much of a shock hazard.
For the alternate 5 volt DC battery power source, a higher capacity and less expensive option could be a Gell Cell six volt battery with a series linear voltage regulator. You may also consider the option of a good quality mains powered LINEAR regulated DC power supply.

As for the reality of sequential data transmission : "Btw is there not a technique to read all ADCs at same time instead of serially?"
consider the basic definition of "serial".
It was not mentioned in the manual if the amplifier did all of the conversions at the same time, and then sent the data after the sample. THAT scheme is the way it could be done, but then the readings would not be updated at the same rate .
What I am not able to state is how much effect the data skew will have on the results of the experiment. THAT is outside of my area of my experience.

AlecT mentions in post #85 that there are simultaneous reads, but only one DSP. So you will need to get more detail from the company, if they are willing to explain how their system delivers the data. What appears to be clear is that the data is not quite delivered in real time.
IF other users have not included that reality in their analysis of data, there may be a bit of confusion.

 

Consider that the input impedance is "greater than 100 MEGOHMS, how much current could flow to a connected person, even in an instance of a direct connection to the 120 volt mains, thru that resistance??? Recall that current= voltage/resistance. so 120/100,000,000 works out to be much less than ONE MICRO-AMP. Not much of a shock hazard.

True if the fault current is through that resistance, but perhaps the chat (or relevant safety standard) was envisioning something like dielectric breakdown, or MOSFET gate breakdown, whereby the resistance could be much lower?

 

If the

True if the fault current is through that resistance, but perhaps the chat (or relevant safety standard) was envisioning something like dielectric breakdown, or MOSFET gate breakdown, whereby the resistance could be much lower?

Some of the advice from chat indicates a limited insight. Or perhaps hallucinations.

 

I found the following data. There are no delays between the 16 channels. All readings were simultaneous with perhaps some microseconds delay between what appears on screen and what you see which is for all intent and purposes simultaneous? I wonder if the Brainmaster Discovery brand also do this.

001-D32SensorsSignalsSoAv121.pdf

a excerpt below:

 

Yesterday I was checking this AC based Nirvana SMPS designed to power audio devices with the lowest noise. It has 5V 2.85A output. My EEG unit is only 1.3A maximum 5V. Is the Nirvana better than say the Globtek medical power supply? Does lower noise equate to better safety too? Is it better than the powerbank 3.7V to 5V buck or boost converter in noise? What other ultra low noise 5V AC power supply do you know that is better than a powerbank?

Nirvana SMPS

 

Does that SMPS comply with medical safety regulations? Note the passage re leakage current and grounding (which you don't have) :-

 

Question about the above description " Any switched mode power supply (smps) has a small transformer that creates EMI because of the abrupt on/off of mosfets. To counter it, we have a capacitor across the input and output of the transformer (that has a special rating Y since failure can be catastrophic). "

Is this capacitor between primary and secondary also put in all the 12V adaptor we used in router, cctv, etc.? I never heard of AC adaptor that can become 220V in the DC. What special rating Y is it? because since all 12V adaptor is made in china, I wonder if they mass produced those special rating Y. What rating is it exactly? And how can catastrophic failure occur in the above description?

 

What special rating Y is it?

hi,

Link.
while Y capacitors are connected between the line and ground. These capacitors are designed to fail in a predictable way (open circuit) when subjected to a surge, unlike standard capacitors that might short and cause a fire or shock.

E
https://www.google.com/search?client=firefox-b-d&q=X+Y+ratings+of+capacitors

 

we have a capacitor across the input and output of the transformer

Not entirely clear whether there is one cap linking primary and secondary, or one cap across the primary and another cap across the secondary. As for the Y rating, have a look here.

 

Once again, I pose the question as to how much shock current can a "perfectly grounded" person receive thru a ten megohm resistance??

 

Once again, I pose the question as to how much shock current can a "perfectly grounded" person receive thru a ten megohm resistance??

An antistatic wrist band has also 10 Megaohm resistor and so even 220V at the other end won't hurt you, isn't it? I asked about the safety of wearing antistatic wrist band when other end is attracting fault currents that run under the earth.

But I haven't been using the AC adaptor because the gtec technical support said he has customers where surge from the AC adaptor damaged the g.USBamp. I only got a few technical support because I only bought it at Ebay. And there is no warrantee so I can't avoid any surges. Anyway, for the included medical AC adaptor. Is it not there is very high leakage current protection between primary and secondary, how can surges get into the secondary and damage the unit?

 

An antistatic wrist band has also 10 Megaohm resistor and so even 220V at the other end won't hurt you, isn't it? I asked about the safety of wearing antistatic wrist band when other end is attracting fault currents that run under the earth.

But I haven't been using the AC adaptor because the gtec technical support said he has customers where surge from the AC adaptor damaged the g.USBamp. I only got a few technical support because I only bought it at Ebay. And there is no warrantee so I can't avoid any surges. Anyway, for the included medical AC adaptor. Is it not there is very high leakage current protection between primary and secondary, how can surges get into the secondary and damage the unit?

An AC power supply (ac adapter) that is designed and built adequately will not deliver any voltage surges, except possibly from a nearby lightning strike. And certainly there is a vast spread in the quality of those external DC power supplies. I observed in the photos and illustrations that the amplifier uses a quite different DC power input connector from the type used for most equipment. Certainly that will make using an alternate power source more challenging.
One available option will be to add an external UPS system that includes surge suppression in line with the mains power to the present AC adapter. THAT should also reduce the noise from the AC mains supply.

 

An AC power supply (ac adapter) that is designed and built adequately will not deliver any voltage surges, except possibly from a nearby lightning strike. And certainly there is a vast spread in the quality of those external DC power supplies. I observed in the photos and illustrations that the amplifier uses a quite different DC power input connector from the type used for most equipment. Certainly that will make using an alternate power source more challenging.
One available option will be to add an external UPS system that includes surge suppression in line with the mains power to the present AC adapter. THAT should also reduce the noise from the AC mains supply.

Why do you still recommend AC adaptor versus just using 5V Powerbank? Do you mean AC adaptor can still be cleaner than the powerbank 3.7V to 5V buck/boost converter noise? In the research setup i'm replicating, they used all batteries as well as used electromagnetically shield room (something I'm still figuring out how to build although it can be small enough just to fit one person)

About the input of my unit, I made a ODU plug to usb converter last year to connect my powerbank to the unit:




8 years ago when I installed the prototype all Siemens GFCI main breakers. I also installed the First Surge protector (see photo below). I mentioned prototype because my American advisers then said no one installed all GFCI breakers in US because the grounding in the appliances can cause leak current and trip them. And in my country I am the only one who used this.

 

I have seen a few of similar surge suppressors installed in different entrance panels, each connected to their own circuit breaker. Consider that the only possible way that a parallel surge suppressor can reduce a surge power is by conducting that surge to "ground" or the system neutral, when that happens with any size of surge that triggers the protector to conduct will thus cause that circuit breaker to trip open, which disconnects it from the system. So the suppressor only is protecting until the circuit breaker opens, which might be much less time than the length of the surge voltage.

In the application information for industrial surge protectors, they ALWAYS connect the suppressor directly across the circuit to be protected, AND they ALWAYS connect a circuit breaker to disconnect the protected load when the surge protector starts conducting. Mostly, those circuit breakers are the automatic reset type, that close a short time after the protection action.

 

I review about surge protectors I learnt 8 years ago. There are 3 kinds of type as summarized below. I only have type II. I don't have type III which is equipment level. And with no grounding, I rely on the MOVs sacrificing themselves in the Siemens Firstsurge instead of diverting the surge current to ground. My breakers never trip on surges. This needs special kinds of breakers? normal breakers don't trip on surges. Anyway with no type 3 and competent grounding. I will not use the AC adaptor again. I'll juse battery. The reason for this thread is knowing why the harmonics are present. This is to to know if the unit is defective or working well. Now that knowing the harmonics are due to lack of grounding at ac adaptor. I know the unit is functioning and that is enough and will go back to battery power.

https://viox.com/surge-protective-device-type-1-vs-type-2-vs-type-3/


"Understanding the Surge Protection Hierarchy
Before diving into the specifics of each type, it’s important to understand how surge protective devices work together in a coordinated system:

Type 1 SPDs: The first line of defense, installed at the service entrance
Type 2 SPDs: The secondary protection, installed at distribution panels
Type 3 SPDs: The final layer of protection, installed near sensitive equipment
This cascading approach to surge protection, known as “defense in depth,” provides comprehensive protection throughout your electrical system.

Type 1 Surge Protective Devices: The First Line of Defense
What Are Type 1 SPDs?
Type 1 surge protective devices are the heavy-duty front-line defenders in your electrical system. They are specifically designed to handle high-energy surges, including direct lightning strikes, and are installed between the utility service and the main electrical service panel.

Key Characteristics of Type 1 SPDs:
Installation Location: Service entrance, upstream of the main breaker
Voltage Protection Rating (VPR): Typically 700-1500V
Surge Current Capacity: 50,000 to 200,000 amperes or higher
Technology: Usually employs silicon avalanche diodes or metal oxide varistors with thermal disconnects
Testing Standard: Tested with the 10/350μs impulse waveform, simulating direct lightning strikes

Type 2 Surge Protective Devices: Distribution Level Protection
What Are Type 2 SPDs?
Type 2 surge protective devices provide the second tier of protection and are the most commonly installed SPDs in both residential and commercial applications. They protect branch circuits and equipment from surges that have been partially mitigated by Type 1 devices or lower-energy surges generated within the building.

Key Characteristics of Type 2 SPDs:
Installation Location: Distribution panels, sub-panels, or branch circuits
Voltage Protection Rating (VPR): Typically 600-1200V
Surge Current Capacity: 20,000 to 100,000 amperes
Technology: Usually employs metal oxide varistors (MOVs) with thermal and overcurrent protection
Testing Standard: Tested with the 8/20μs impulse waveform, simulating indirect lightning effects

Type 3 Surge Protective Devices: Point-of-Use Protection
What Are Type 3 SPDs?
Type 3 surge protective devices provide the final layer of protection at the equipment level. They are designed to handle the residual surges that may still exist after Type 1 and Type 2 protection, as well as small surges generated within the building’s wiring system.

Key Characteristics of Type 3 SPDs:
Installation Location: Outlets, power strips, or directly integrated into equipment
Voltage Protection Rating (VPR): Typically 330-600V
Surge Current Capacity: 5,000 to 20,000 amperes
Technology: Usually employs MOVs with additional filtering components
Testing Standard: Tested with the 8/20μs impulse waveform, with specific installation location requirements
Installation Requirement: Must be installed at least 10 meters (30 feet) from the service entrance"