Step up transformer, from 120VAC to 220VAC, to a 5Mhz pulse. Seeking Advice

Thread Starter

mbferguson

Joined Apr 23, 2017
94
I am trying to create a high voltage pulse at approximately 200 to 220V, at 5MHz. I have explored a few ways to do this using op amps, and multistage amplifiers using transistors, however I am trying to go about this using the most simple method possible.

My idea is to use a transformer from a wall outlet.

The outlet power in North America is 120VAC 60Hz, so I would need a transformer with Pri-120VAC to Sec-220VAC, but I cannot find one like this anywhere after searching on google for a while.

I read that transformers can be used backwards, so I'm hoping the following will work for me. It has multiple primaries and secondaries listed. How does one go about selecting between those?

https://www.alliedelec.com/solahd-e...qy8JinNYUNS2_hBEDDtWo8NQQniK6wZIaAqfPEALw_wcB

Because I am trying to get a specific frequency, I think it is necessary to convert the 220VAC 60Hz, into 220VDC so that I can control a square wave form at a specific frequency using a switch. I found the following article which seems to be able to do this, but I am clueless as to what component values would need to be manipulated.

https://www.eleccircuit.com/the-variable-high-voltage-power-supply-0-300v/

I've drawn a cartoon to help show what I'm trying to do. It'd be nice if I could get away with simply using some type of battery, however I doubt they make transformers that go from ~9V to 220V. Any advice is appreciated.

 
Last edited:

danadak

Joined Mar 10, 2018
4,057

Thread Starter

mbferguson

Joined Apr 23, 2017
94
A standard wall wart transformer totally unbsuitable for work at 5 Mhz.

You trying to do something like this at 5 Mhz ?

http://www.mmjp.or.jp/peec/ehvpulse.htm

A discussion here -

https://www.edaboard.com/showthread.php?340478-High-Voltage-200V-high-frequency-1MHz-pulse-generator

High frequency transformer techniques -

https://www.agilemagco.com/transfor...MIpYWh3szE3gIVykwNCh1ltQT1EAAYAiAAEgIQdPD_BwE



Regards, Dana.
Yup, those seem to be good resources to look at. I'm hoping that the 5MHz frequency will not matter for the transformer, atleast I think so, because it is being rectified after the fact. Why would the microcontroller switching closed/open have an effect on the transformer?

I can see how it would have an effect if current is halted, thus preventing the transformer from working.
 

danadak

Joined Mar 10, 2018
4,057
Unless I do not understand, you are trying to use a transformer design for 50/60 Hz,
the design of its core and windings, in a 5 Mhz application. The core and windings
design approach quite different for LF vs HF applications. One cannot push RF thru
LF transformers, except by parasitics.

Transformers at HF are typically used for impedance matching. Cores typically ferrite
vs laminated iron/steel.


Regards, Dana.
 

Alec_t

Joined Sep 17, 2013
14,280
What will you do with this 220V 5MHz pulse train? How will you suppress all the nasty RF noise it will create? How much current must the pulse provide to the (unspecified) load?
 

ArakelTheDragon

Joined Nov 18, 2016
1,362
The circuit that you are looking for is a bridge with transistors. The transistors can change the DC current and create an AC 5MHz signal. After the transformer's secondary coil, there is a bridge rectifier with filter capacitors, after the capacitors are the transistors.
 

ArakelTheDragon

Joined Nov 18, 2016
1,362
What will you do with this 220V 5MHz pulse train? How will you suppress all the nasty RF noise it will create? How much current must the pulse provide to the (unspecified) load?
Mostly this is used for big industrial converters and power supplies I think. My other guess is some communication with the signal.
 

ebp

Joined Feb 8, 2018
2,332
220 VAC RMS when rectified and filtered will yield a peak voltage of about 305 VDC. The peak will change little with loading but the average voltage will drop according to the load and the capacitance of the filter.

Transformers can generally be used in either direction. However, many are designed to produce the specified voltage at the full specified load current, so a small adjustment is made in the turns ratio to compensate for losses in the winding resistance. This means that if you use what as intended as a secondary as the primary and apply the nominal voltage to it, the winding originally intended as the primary will deliver a voltage lower than the nominal rating.

To produce the output you require, the most sensible devices to use are power MOSFETs driven with high-speed high-current gate driver ICs. Components would have to be chosen carefully. Circuit layout would be critical. This is not something a novice is going to find easy and it isn't something that can be built on a plug in breadboard with any expectation of decent performance. There are potentially lethal shock hazards.
 

MisterBill2

Joined Jan 23, 2018
18,175
If you produce much power at any frequency near 5 megahertz it will be radiating far more power than you can do legally in the USA, and the FCC will quickly ask you to cease . If you don't, then comes the notices of monetary fines, followed by federal marshals, who can be very serious folks.
And if you start pumping much power at 5 megahertz into an iron core transformer designed for 60 hertz, you may well wind up with the local fire department putting out the fire, if you are lucky.
So I am asking you to explain what you are hoping to make happen with this radio frequency energy. Many of the folks here are able to provide very good advice that is quite useful, but almost none of us is able to read minds at any distance. Thus we require a much more complete description so that we can provide good advice.
In addition, the warnings about shock hazards given in post #8 are real. You are talking about voltages and frequencies that can not only kill but also produce really nasty burns.
 

Thread Starter

mbferguson

Joined Apr 23, 2017
94
If you produce much power at any frequency near 5 megahertz it will be radiating far more power than you can do legally in the USA, and the FCC will quickly ask you to cease . If you don't, then comes the notices of monetary fines, followed by federal marshals, who can be very serious folks.
And if you start pumping much power at 5 megahertz into an iron core transformer designed for 60 hertz, you may well wind up with the local fire department putting out the fire, if you are lucky.
So I am asking you to explain what you are hoping to make happen with this radio frequency energy. Many of the folks here are able to provide very good advice that is quite useful, but almost none of us is able to read minds at any distance. Thus we require a much more complete description so that we can provide good advice.
In addition, the warnings about shock hazards given in post #8 are real. You are talking about voltages and frequencies that can not only kill but also produce really nasty burns.
I apologize if I was not clear enough in my post. I am an undergraduate student working a project for school. There are many things left unexplained to me in relation to how transformers work. It's clear to me that modeling it as a simple ratio of coil turns is not enough to achieve what I need.

I am confused why many other comments have stated that the 5MHz will be a problem. If I input 60Hz 120VAC, the output would be 220VAC at 60Hz. The rectifier would create some DC voltage above that. The switch would then simply let a small amount of voltage/current out at a specific event, not continuously. Is the current leaving through the closed switch what will cause a problem with the 60Hz transformer?
 
Last edited:

Thread Starter

mbferguson

Joined Apr 23, 2017
94
What will you do with this 220V 5MHz pulse train? How will you suppress all the nasty RF noise it will create? How much current must the pulse provide to the (unspecified) load?
This pulse will be used to drive an ultrasonic transducer. I do not think that the current is an important factor in the operation of the transducer. It is powered by the upwards motion of the waveform, which then resonates at the specified frequency of 5MHz oscillating until the energy leaves the transducer.
 

MisterBill2

Joined Jan 23, 2018
18,175
A transformer that will work at 5,000,000 Hz is totally different from one that works at 60Hz. Certainly such transformers have been created but they are totally different.. The wire, the core materials, and the physical arrangement would be quite different. Next, the 60 Hz current is a sine wave with very few harmonics present, while the switched wave is a square pulse with a whole lot if odd harmonics out to frequencies many times the 5 megahertz fundamental frequency. Normally, to improve efficiency and reduce the number and magnitude of harmonics, the 5 megahertz transformer is made to resonate at that frequency. That in itself brings complexity.
The switch is a whole another complex system, especially if the transformer system is resonant. Switching on and off at any higher frequency involves charge building and dissipating at a very high rate. And having the power to quickly change from off to on is a big challenge as well..

On top of that, my warning about the FCC getting involved is still applicable. Those are laws, not just suggestions.

And we still have no idea as to what the purpose of this is intended to do.
 

Thread Starter

mbferguson

Joined Apr 23, 2017
94
A transformer that will work at 5,000,000 Hz is totally different from one that works at 60Hz. Certainly such transformers have been created but they are totally different.. The wire, the core materials, and the physical arrangement would be quite different. Next, the 60 Hz current is a sine wave with very few harmonics present, while the switched wave is a square pulse with a whole lot if odd harmonics out to frequencies many times the 5 megahertz fundamental frequency. Normally, to improve efficiency and reduce the number and magnitude of harmonics, the 5 megahertz transformer is made to resonate at that frequency. That in itself brings complexity.
The switch is a whole another complex system, especially if the transformer system is resonant. Switching on and off at any higher frequency involves charge building and dissipating at a very high rate. And having the power to quickly change from off to on is a big challenge as well..

On top of that, my warning about the FCC getting involved is still applicable. Those are laws, not just suggestions.

And we still have no idea as to what the purpose of this is intended to do.
The purpose is to be used for an ultrasonic transducer. It is powered by a 220V pulse at 5MHz, but only half of the period (lifting up a trampoline and letting it go).

I understand a little bit more now that you mentioned the charge building and dissipating. Would this cartoon drawing I made be possible if the pulse was only create for one single half cycle? (100ns)
 

MisterBill2

Joined Jan 23, 2018
18,175
A single half cycle is a possibility, but not a trivial thing. When fed a signal resonant systems as well as non-resonant systems tend to ring. More later, I am falling asleep.
 

ebp

Joined Feb 8, 2018
2,332
With 200 V applied, the current into the transducer might be as high as 4 or 5 amperes. If it is the sort used for non-destructive testing (e.g. thickness gauge or weld inspection), the usual way to excite it would be to apply a very narrow pulse, well under 1/2 cycle in width and let the transducer ring at its resonant frequency. Pulses would be quite widely spaced, so the average power would be a small fraction of a watt.

I think people have some gotten the impression that you intended to use an iron core transformer at 5 MHz, though how I don't know. Your diagram looks pretty clear to me.

Because your average power will be very low, if you have trouble finding a transformer with the required 1:2 ratio, you could use two small transformers. For example, there are lots of small transformers with dual primary windings - you connect the windings in parallel to use the transformer at 120 V or in series to use it at 240 V. If you had two such transformers, with any convenient secondary voltage, you could connect the secondaries of the two together, the primaries of one for 120 V and the (intended) primaries of the other for 240 V, and use the latter as your output. You could probably find two otherwise-identical (but they wouldn't need to be) transformers, one with say a 12 V secondary and one with an 18 V secondary to give you an overall step-up of 12/18 x 240/120 = 4/3 for about 160 V RMS, which would have a peak voltage of about 224 V after rectifying and filtering. (that is you put 120 V into the xfmr with the 12 V secondary, connect the 12 V secondary to the 18 V secondary of the second transformer and take 160 VAC from the primary of the second transform; "primary" and "secondary" here meaning the originally-intended purpose). You would need transformers rated at only a few "volt- amperes" - perhaps 3 to 6 VA (a 6 VA transformer with at 12 V secondary would be rated for 0.5 A secondary current).

I'm not sure this will work as intended, but if it does, the link will get to to Digi-Key with a few 12 V types that could be used:
https://www.digikey.com/products/en...tity=&ColumnSort=0&page=1&stock=1&pageSize=25
There will be other secondary voltages available. Most of these small transformers are made for printed circuit board mounting.
 

Thread Starter

mbferguson

Joined Apr 23, 2017
94
I think people have some gotten the impression that you intended to use an iron core transformer at 5 MHz, though how I don't know. Your diagram looks pretty clear to me.
Thank you. I was paranoid when people kept going on tangents I didn't understand how they arrived at. I posted the merchant link for the specific transformer I'm looking at as well.
 

Thread Starter

mbferguson

Joined Apr 23, 2017
94
With 200 V applied, the current into the transducer might be as high as 4 or 5 amperes. If it is the sort used for non-destructive testing (e.g. thickness gauge or weld inspection), the usual way to excite it would be to apply a very narrow pulse, well under 1/2 cycle in width and let the transducer ring at its resonant frequency. Pulses would be quite widely spaced, so the average power would be a small fraction of a watt.
I was under the impression this entire time that the transducer's resonant frequency was necessary for any input pulse. It makes sense though, any voltage should excite the "trampoline" part of the transducer to begin bouncing up and down, which would settle into it's resonant frequency of 5MHz. This seems to simplify a great deal of some of the problems I'd been trying to remedy.

This raises another question though, if the pulse duration can be independent of the half-period length, do I really need to bother with rectifying the AC into DC? I thought it was necessary originally to ensure I could control the pulse was entirely one polarity of voltage.

I would have no way of knowing what part of the period the AC waveform would be in, meaning I could potentially close->open the switch when the waveform is in the middle at 0V. It's probably best to stick with the rectifier I suppose.

This is the actual probe I'm trying to operate. It's intended to be used with a guage meter typically, and there is only a limited amount of data sheet information available. I was able to find out that a 220V pulse is used from the manufacturer though.

https://www.ebay.com/itm/PT-12-Prob...hickness-Gauge-/123458946992?oid=122466251951
 
Last edited:

MisterBill2

Joined Jan 23, 2018
18,175
MBBF, I seldom follow links and so I had not seen anything that pointed to a different type of second transformer. Now that I understand the application I can offer some advice that should be more useful. A single pulse can come from a rather straight forward capacitor discharge circuit. If you do not need to have it triggered, but to just pulse every second or so, that circuit can be as simple as a cheap strobe light circuit. That technique was actually used in some of the very early navy sonar systems, the ones that used all tubes. All you need to do is put the transducer in series with the flash tube and trigger the flash. That will deliver one pulse, and when the one flash is completed the tube returns to it's off state of a very high impedance, allowing the transducer to ring at it's resonant frequency. The reason for not suggesting the use of a FET device is that the flash tube has much less capacitance in the off state. Probably there are some FETs that would work, but I have not researched them. Also, the flash tube is much less likely to be damaged by a higher voltage transient. The only hard part is selecting the size of the capacitor and picking out the flash tube that is higher powered enough.
It could also possibly work with an SCR device.
AND, now I recall that many types of depth sounder systems use a pulse driven transducer to generate the "Ping" that they transmit. So there are some more ideas. I know that the flash tube one works.
And this thread is a perfect example of why presenting enough information early will lead to useful suggestions much sooner.
 

Thread Starter

mbferguson

Joined Apr 23, 2017
94
And this thread is a perfect example of why presenting enough information early will lead to useful suggestions much sooner.
You are right. When multiple people have misunderstandings it is more than likely because the post is ambiguous. I should have described the load so that it was more obvious about my intentions with the pulsing of the high voltage.
 
Last edited:
Top