[Tutorial] Building your own DIY E-Stim Stereo Device

[TheHenchmen]

Great Thread, I essentially joined because of this.

I build my Estim as per the "seraah mk4b" setup from seraah dot net. It´s fundamentally similar, but has some design differences:

- It utilizes a 18W KEMO amplifier module for each channel, so 36W theoretical power
- the resistor is 5k Ohms at 10Watts ceramic resistor, not 1k with cooling. The resistor is placed at the output end of the transformers.

It uses the same transformers in the same configuration.

The following lessons have been learned thus far:
- KEMO Amp modules are quite bulky and take a lot of space due their fixed cable attachments
- The power requirement is a bit high, utilizing battery power in the first try killed the output in a short notice

The most frustrating part:
- The system doesn´t work properly with high frequency files (like tek files) or high-frequency sections in webteases (like tower). The tease says "now feel THAT!", my LEDs blink happily.... and I feel nothing.

[Electro]

Great Thread, I essentially joined because of this.

I build my Estim as per the "seraah mk4b" setup from seraah dot net. It´s fundamentally similar, but has some design differences:

- It utilizes a 18W KEMO amplifier module for each channel, so 36W theoretical power
- the resistor is 5k Ohms at 10Watts ceramic resistor, not 1k with cooling. The resistor is placed at the output end of the transformers.

It uses the same transformers in the same configuration.

The following lessons have been learned thus far:
- KEMO Amp modules are quite bulky and take a lot of space due their fixed cable attachments
- The power requirement is a bit high, utilizing battery power in the first try killed the output in a short notice

The most frustrating part:
- The system doesn´t work properly with high frequency files (like tek files) or high-frequency sections in webteases (like tower). The tease says "now feel THAT!", my LEDs blink happily.... and I feel nothing.

You should be able to feel it, I do with my StereoStim setup. I didn't see anything over 1000hz in Estim Tower.

[Caius1]

What kind of amperage is going through the system in these builds? More precisely I'm thinking about the required A rating for the triphase switch on the secondary/output side of the transformers.

Still following mantrid's guidelines, if you get for example the MA-180 amplifier that has input power 12VDC 5A, Digikey 237-1146-ND transformer, 10Ohm 2W resistor and 100µF capacitor, then is the current on both sides of the transformer 0,25A?
This value I got from mantrid's data sheet, which only has current on primary side, row 12. I'm also assuming that the primary side means the input side of transformer?

[Electro]

What kind of amperage is going through the system in these builds? More precisely I'm thinking about the required A rating for the triphase switch on the secondary/output side of the transformers.

Still following mantrid's guidelines, if you get for example the MA-180 amplifier that has input power 12VDC 5A, Digikey 237-1146-ND transformer, 10Ohm 2W resistor and 100µF capacitor, then is the current on both sides of the transformer 0,25A?
This value I got from mantrid's data sheet, which only has current on primary side, row 12. I'm also assuming that the primary side means the input side of transformer?

For the price of that Digikey transformer, you could get a better transformer that has a higher frequency range and more taps to adjust turns ratio if you want, the Speco T7010 is $15 on Amazon. ..but if you are going to be building his build, I wouldn't interchange parts, go all out with his if you are going that route.

For estim purposes, we are using these transformers backwards. The audio line transformers we use are normally used as step-down transformers where the power goes into the primary winding where there are more turns around the iron core(ohm side) and out the secondary winding where there are less turns. For estim we are putting power into the secondary winding and wiring electrodes to the primary side with more turns.

A tri-phase switch goes on the electrode side. Amperage is normally around 50mA to the electrodes, spikes probably around 100mA would be too much for a person to handle (I think). Normally this is a very small amount of current for a switch. There is going to be current on both sides of the transformer. 100ohms is a different value than most StereoStim builds, Mantrid's build doesn't match most StereoStim designs. Not to say it doesn't work because obviously Mantrid is satisfied with his setup, but it's different because most people also aren't using capacitors and HV blockers in their circuits which changes the values required for the other components.

[lolol2]

I build my Estim as per the "seraah mk4b" setup from seraah dot net. It´s fundamentally similar, but has some design differences:

- It utilizes a 18W KEMO amplifier module for each channel, so 36W theoretical power

My design is inspired or nearly 99% taken from the first version of the seraah estim box.
I also tried the one with the KEMO single channel amp blocks but had no luck with a good output.
It felt very spiky and unpleasant, even after I tripple checked everything, not sure if maybe one of the AMPs had a problem.

At the end after a few hours of try and error I switched back to the first design and still very happy with it.

[Caius1]

For estim purposes, we are using these transformers backwards. The audio line transformers we use are normally used as step-down transformers where the power goes into the primary winding where there are more turns around the iron core(ohm side) and out the secondary winding where there are less turns. For estim we are putting power into the secondary winding and wiring electrodes to the primary side with more turns.

Oooh! The transformer's used backwards! Now that's a thing I missed completely
And now some things make more sense.

For the price of that Digikey transformer, you could get a better transformer that has a higher frequency range and more taps to adjust turns ratio if you want, the Speco T7010 is $15 on Amazon. ..but if you are going to be building his build, I wouldn't interchange parts, go all out with his if you are going that route.

Yeah, the Speco type seems quite popular, and it's way cheaper, and adjustable. Would it be possible to combine these, the high pass safety features mantrid's proposing and the adjustability of the Speco transformer? I'm guessing that it wouldn't be very feasible or worth the effort, but I like safety features, but I also like not using a lot of money :DD
Obviously people haven't had many problems without the filters but why not go safe if you can?

[mantrid]

So the current consensus is that a fuse is not a necessary safety feature but I also take it that a fuse wouldn't have any negative effects either.

A fuse indeed isn't a bad idea, because it protects against malfunction of the transformer. Especially shortcuts in the primary coil can cause current increase.

The fuse must be on the primary side, i.e. between transformer and amplifier. The fuse current can be calculates with the spreadsheet you already found. It's the I_P value, i.e. about 1 A.

The TVS diode at the output of the amplifier is a better protection against malfunction of the amplifier or power supply because it reacts faster. But it does not protect against transformer failure.

The only problem of the fuse is that it reacts so slow. Unless you play with selfbondage you will be faster than the fuse.

An even better protection is a TVS diode at the transformer output, i.e. parallel to the electrodes. But value depends on electrodes. You have to measure the voltage between electrodes using a sinusoidal signal calibrated to a comfortable level and then choose a bidirectional TVS diode (the ones ending wit "CA") with a breakdown voltage that is at twice as large.

Such a diode also protects against hot spots which occur if the conductivity becomes to bad. (Output is current controlled. If the contact surface becomes smaller current density and voltage increases.)

I
EDIT
I also started to think more about the high pass filters.

So my recommendation:
Low cut-off frequency option: 100µF + 20 Ohm or 220µF + 10 Ohm
High cut-off frequency option: 100µF + 10 Ohm
In both cases, highest winding ratio that deliver sufficient power should be used.

I have 0 experience with the estim files/teases (only xp with basic Mystim device) so I don't really know if I should build a higher or lower cut-off. So I thought, why not go crazy and build them both?

You will only feel a difference with files that play with different frequencies. Then the low cut-off option is the better choice.

The problem is that many estim creators use low frequencies signals for pain. And that's IMHO a bad idea.

I'm trying to work out a shopping list now, if I look at mantrid's list in the spoiler below, where's the 5,1 Ohm resistors supposed to go?

It is intended as an alternative for the 10 Ohm resistor. You only need one resistor per channel. In my setup I have one channel with a lower series resistor.

But see below.

What kind of amperage is going through the system in these builds? More precisely I'm thinking about the required A rating for the triphase switch on the secondary/output side of the transformers.

Still following mantrid's guidelines, if you get for example the MA-180 amplifier that has input power 12VDC 5A, Digikey 237-1146-ND transformer, 10Ohm 2W resistor and 100µF capacitor, then is the current on both sides of the transformer 0,25A?
This value I got from mantrid's data sheet, which only has current on primary side, row 12. I'm also assuming that the primary side means the input side of transformer?

I do not recommend the red transformer anymore because I noticed clipping. I described that in the post with the spreadsheet you already found. The clipping is also the reason why I was less sensitive to certain volume effects others considered as harsh.

Use the black 1:22.4 transformer instead. On the end of the post you find a new shopping list with components for 4 channels. Only issue is that the power rating of the resistors could be larger. With the dummy load they became quite hot, but not too hot because max. operational temperature is 175°C or so.

Power and current calculation are in the spread sheet. With 1A per channel for power supply you are on the save side. The current at the electrodes is just a few mA. Almost all power is transformed into heat in order to make the signal current controlled.

In our case primary side is the transformer side with the low turn number.

For the price of that Digikey transformer, you could get a better transformer that has a higher frequency range and more taps to adjust turns ratio if you want, the Speco T7010 is $15 on Amazon.

I cannot find a datasheet for it. But with a proper datasheet and the spreadsheet (there is a manual for it) you can easily taylor values of all components.

I just prefer the smaller transformers from digikey because I can easily make modules from it and isolate them with shrink tubings. Together with boxed amplifiers this avoids complicated housings.

[hosenguy]

My main stim box is built with 18 watt Kemo amps. I built it a bunch of years ago, don't remember how many. It has served me very well. It is from a design taken from the SmartStim site.

I have built several other designs and tested all of them and compared them.

The build with the Kemo amps was plagued with problems when first completed as it wanted to provide an unwanted high frequency ringing after any signal peak. That made the signal seem muted and covered much detail of the signal and drained the battery quickly. I tracked the problem with an oscilloscope and ended up fixing the trouble by adding proper signal separation utilizing correct wire routing along with parts placement, adding filters to limit the input signal frequency at the input to each amp, and adding bypass caps on the power immediately next to the modules.
The Kemos are very nice package amplifiers and have a very wide frequency response.

I have not had this issue with any other box I have build with any other amp module. Of course I learned to be serious about cleanliness with the build layout and wire routing since that first box!

[TheHenchmen]

The build with the Kemo amps was plagued with problems when first completed as it wanted to provide an unwanted high frequency ringing after any signal peak. That made the signal seem muted and covered much detail of the signal and drained the battery quickly. I tracked the problem with an oscilloscope and ended up fixing the trouble by adding proper signal separation utilizing correct wire routing along with parts placement, adding filters to limit the input signal frequency at the input to each amp, and adding bypass caps on the power immediately next to the modules.

Interesting.
Please provide a sketch of these adaptions.

As you build with kemo amps, I need to ask again:
Is the resistor type I´m using the right one? I´m using a 1kOhm 5W ceramic resistor, other builds use a resistor with a metal heat sink.

[Caius1]

The fuse must be on the primary side, i.e. between transformer and amplifier.
...
In our case primary side is the transformer side with the low turn number.

For estim purposes, we are using these transformers backwards. -- For estim we are putting power into the secondary winding and wiring electrodes to the primary side with more turns.

So let me get this straight. The transformer is used backwards, so the power goes in through the secondary side and the electrodes are on the primary side of the transformer? But when talking about the whole thing, amplifier is still on the primary side of the whole build because that's where the current is coming from.
Sorry, I'm a little confused.

You will only feel a difference with files that play with different frequencies. Then the low cut-off option is the better choice.
The problem is that many estim creators use low frequencies signals for pain. And that's IMHO a bad idea.

I can live with that, I'm not that into pain. And if I ever start craving for more stinging zaps, I know what to do.

I do not recommend the red transformer anymore because I noticed clipping.

Oh right, forgot about that. Thanks for the reminder!

For the price of that Digikey transformer, you could get a better transformer that has a higher frequency range and more taps to adjust turns ratio if you want, the Speco T7010 is $15 on Amazon.

I cannot find a datasheet for it. But with a proper datasheet and the spreadsheet (there is a manual for it) you can easily taylor values of all components.

What kind of info would be needed to get tailored info out of the spreadsheet? Power transformers don't seem to have turns ratios or impedance in their datasheets (only a quick search) but they do seem to have in and out voltages, can you just get the turns by dividing the voltages?
This page has the Speco T7010 and tell the Ohms, but it doesn't have secondary voltage, can it be 70V, the same as input?

[Caius1]

According to the spreadsheet, using the black transformer Digikey 237-2018-ND 22,4:1.
2,5 Ohm resistor should give similar results as the red one (25:1) with 10 Ohm resistor.
With load of 500 Ohm cutoff is at 235,9Hz and load of 200 Ohm cutoff at 268,3. Both cases output current at 50mA.

5W a good a power rating or is there really even a max, or is it just how much you wanna pay for the power rating?
The shopping list has 10W but I'm just thinking about things.

[mantrid]

The fuse must be on the primary side, i.e. between transformer and amplifier.
...
In our case primary side is the transformer side with the low turn number.

For estim purposes, we are using these transformers backwards. -- For estim we are putting power into the secondary winding and wiring electrodes to the primary side with more turns.

So let me get this straight. The transformer is used backwards, so the power goes in through the secondary side and the electrodes are on the primary side of the transformer? But when talking about the whole thing, amplifier is still on the primary side of the whole build because that's where the current is coming from.
Sorry, I'm a little confused.

With primary side I mean the input side. IMHO that assignment is also the convention.

Datasheets often describe the side with the higher turn number as primary side, if they use that term. But because there is no standard direction for transformers that term is a little bit misleading.

In general, both sides can be used as input, i.e. can be the primary side. And in our case it is the side with the smaller turn number.

The problem is that many estim creators use low frequencies signals for pain. And that's IMHO a bad idea.

I can live with that, I'm not that into pain. And if I ever start craving for more stinging zaps, I know what to do.

The problem is that low frequencies are dangerous and that the pain probably cannot be explained just by stimulation of neurons ...

According to the spreadsheet, using the black transformer Digikey 237-2018-ND 22,4:1.
2,5 Ohm resistor should give similar results as the red one (25:1) with 10 Ohm resistor.
With load of 500 Ohm cutoff is at 235,9Hz and load of 200 Ohm cutoff at 268,3. Both cases output current at 50mA.

The Digikey transformer can also be used with a 2.5 Ohm series resistor. That reduces the power loss at the resistor but makes the output current a little bit more dependent form output load.

I added that calculation to the spreadsheet (columns P and Q)

Columns R and S now contain a calculation for the Speco transformer. Primary side is Common and 8 Ohm tap, secondary side is Common and 8000 Ohm tap. Impedance transformation of 8:8000 corresponds to a winding ratio of sqrt(8/8000)=1:31.62. DC resistances (marked red) are guessed. If they would be correct a 2.5 Ohm resistor and 330µF capacitor would fit well.

5W a good a power rating or is there really even a max, or is it just how much you wanna pay for the power rating?
The shopping list has 10W but I'm just thinking about things.

The operating temperature is often up to 175°C. That means at the rated power the temperature raises by about 120K. If they dissipate only half as much they still get about 80°C hot.

[Electro]

A few things about resistors. One of the big reasons why some of these DIY stim box designs use 25 watt chassis mount resistors is that they have a flat bottom with mounting feet making them easy to attach to a metal plate and they also provided a round bar shaped terminal that could used that a nylon terminal block could be screwed down onto for a solderless design, sometimes they needed to be squished a little to fit but it works. Prior to that many were 10+ watt resistors. When attaching mine to a metal plate I used a thin layer of thermal paste to help pull away more of the heat, I also didn't have a perfectly flat plate either. The 25w chassis mount resistors on Amazon often give you 2 to 5 of them for $10, at least for the 3.9 ohm ones that work for my combination of components. When they are sized this big they don't get as hot since they can reject more heat if they are properly mounted and have free air flow. I think 10w should work just fine, although it's possible if you end up with a short circuit and don't realize it, there's a higher possibility that they will cook. I've had my entire metal plate warm up before with a heavy session and touching the resistors was a situation where you wouldn't be comfortable holding your finger on them. If there is no difference(or very little) in price and you have the space for them, getting a 20+ watt resistor that's easy to mount might be worth it to you. The design I used specified a 10 watt resistor before the design was modified to an easier to build design.

[strommann]

Yesterday I drilled and soldered and finally my box was ready to go. I had ordered all the parts lolol2 had listed in his how-to paper and the testdrive was impressive. I already got a ET312 and a 2B for many years and this new box is really different. Seems as if the audio is more clearly defined and at the same time it feels smooth and strong. To my surprise it has lots of power. so thanks to lolol2 and all the other helpful people around here.

[phoopha]

Yesterday I drilled and soldered and finally my box was ready to go. I had ordered all the parts lolol2 had listed in his how-to paper and the testdrive was impressive. I already got a ET312 and a 2B for many years and this new box is really different. Seems as if the audio is more clearly defined and at the same time it feels smooth and strong. To my surprise it has lots of power. so thanks to lolol2 and all the other helpful people around here.

Did you follow the original plans? Were any changes made? Where did you source the parts?