# Boosting a 9V 5MHz pulse waveform to 180V

#### WattsUpWithThat

Joined Nov 9, 2018
3
Hi all,

I'm looking for some guidance on how to accomplish the task in the title. Loss of current isn't much of a concern, I really just need the Waveform I have at a higher voltage. Willing to purchase or build within a reasonable price (\$50 or less).

Pretty much everything I can find online is either DC-DC / DC-AC / AC-DC, there isn't a lot out there concerning AC to AC boost converters. At this point I'm wondering if my only option is to build a transformer, and I can't imagine that a DIY transformer would leave my frequency intact at 5MHz.

Any input helps, thanks!

#### crutschow

Joined Mar 14, 2008
33,363
Getting 180V @ 5MHz is not trivial.
What is the minimum pulse width?
How close does the output waveform need to conform to the input waveform?

#### fernandopv

Joined Oct 20, 2018
10
Allow me to ask you why 180 Vac exactly?

You can get that voltage at 5MHz using LC resonant circuits, being part of an oscilator or driven by a C class amplifier.

Want to radiate electromagnetic waves ? bad idea at exactly 5 MHz.
Is for a sonar ?

#### WattsUpWithThat

Joined Nov 9, 2018
3
Getting 180V @ 5MHz is not trivial.
What is the minimum pulse width?
How close does the output waveform need to conform to the input waveform?
I'm shooting for a 200ns pulse width, and the output waveform needs to be similar to the input waveform but not exact. If 180V isn't doable I could probably get away with closer to 100V.

#### ebp

Joined Feb 8, 2018
2,332
You have provided far too little information. Are you saying the input is an AC (that is, reversing polarity, not just varying magnitude) with 9 V peak to peak amplitude (so ±4.5 V relative to circuit "ground") and you want ±90 V output? Is the input a square wave or ...? Assuming it is rectangular in some fashion, what rise and fall time do you require at the output? How much current? What is the nature of the load? Are you expecting to produce the output directly from the input or are you expecting to generate the output from high voltage power supplies?

What does "loss of current" mean?

#### LesJones

Joined Jan 8, 2017
4,115
If youy Google "transistor oscilloscope deflection amplifier" I think you will find some circuits that should meet your reqirements. This is one circuit that may do what you want.

Les,

#### WattsUpWithThat

Joined Nov 9, 2018
3
You have provided far too little information. Are you saying the input is an AC (that is, reversing polarity, not just varying magnitude) with 9 V peak to peak amplitude (so ±4.5 V relative to circuit "ground") and you want ±90 V output? Is the input a square wave or ...? Assuming it is rectangular in some fashion, what rise and fall time do you require at the output? How much current? What is the nature of the load? Are you expecting to produce the output directly from the input or are you expecting to generate the output from high voltage power supplies?

What does "loss of current" mean?
Sorry for the lack of info I'm new here

Yes AC, reversing polarity, ±9V 18Vpk-pk wanting ±100V 200Vpk-pk. Input is a square wave. Unsure about rise/fall time. Nature of the load is that it requires high voltage and very little current, and I'm looking to produce the output directly from the input.

"Loss of current" means if I have to go with a transformer to scale up my voltage I'm fine with lowering current.

#### ebp

Joined Feb 8, 2018
2,332
Given the requirements, a transformer is your only option. You need a turns ratio of about 1:11. Assuming your claim that you require very little (but unquantified!) current is true, a small toroid, probably of ferrite but possibly of powered iron, would be suitable for the core. With the high frequency, if a suitable low-loss core material with reasonably high relative permeability can be found, you probably won't need very many turns - which is good because winding capacitance can be a problem.

I've wound many pulse transformers for quite high current for driving power MOSFETs, including with pulse widths down to less than 100 ns, but losses were not a big deal.

- sources I would suggest:
• Amidon (used to be Amidon Associates, but they may have changed the name) - a supplier of ferrites and powder cores, the former mostly from Fair-Rite, the latter mostly from Micrometals
• Fair-Rite
• Micrometals
• Magnetics, Inc.
I think you can find some information on high frequency transformer design at the websites of all of these companies, and possibly even some on-line or downloadable aps to assist

High frequency transformers used to be common things in amateur radio transmitters. I don't know if the ARRL has a website with useful things, but its worth a search.

#### ebeowulf17

Joined Aug 12, 2014
3,307
Given the requirements, a transformer is your only option. You need a turns ratio of about 1:11. Assuming your claim that you require very little (but unquantified!) current is true, a small toroid, probably of ferrite but possibly of powered iron, would be suitable for the core. With the high frequency, if a suitable low-loss core material with reasonably high relative permeability can be found, you probably won't need very many turns - which is good because winding capacitance can be a problem.

I've wound many pulse transformers for quite high current for driving power MOSFETs, including with pulse widths down to less than 100 ns, but losses were not a big deal.

- sources I would suggest:
• Amidon (used to be Amidon Associates, but they may have changed the name) - a supplier of ferrites and powder cores, the former mostly from Fair-Rite, the latter mostly from Micrometals
• Fair-Rite
• Micrometals
• Magnetics, Inc.
I think you can find some information on high frequency transformer design at the websites of all of these companies, and possibly even some on-line or downloadable aps to assist

High frequency transformers used to be common things in amateur radio transmitters. I don't know if the ARRL has a website with useful things, but its worth a search.
I have very, very little experience with transformers, so I'm probably just misunderstanding something, but...

I thought transformers were pretty bad for square waves. Not only do you need bandwidth many harmonics above your square wave frequency in order to get good rise times and square-ish corners, but they don't want to hold the flat line tops and bottoms. What should be flat lines will be all ramps towards the middle, right? Maybe at such high frequencies there isn't enough time for ramping down to be noticeable?

#### MisterBill2

Joined Jan 23, 2018
16,603
This sounds like a system to drive a piezo type transducer. For any sort of non-symetrical waveform, look at "horizontal deflection systems" relative to television picture tubes.
The problems with creating square pulses at a high frequency are rise times and switching times.
Actually, many radar systems use bursts of much higher frequency power, so checking some of those references may provide the information that you seek.And sonar systems, likewise, but at lower frequencies. I only know enough about those two types to get into trouble, so I can't help any more than what I just mentioned.

#### MisterBill2

Joined Jan 23, 2018
16,603
Use DIY air core toroidal transformer.
It will work at 5MHz just fine, with efficiency about 80%.
Square pulses will transform without their form distortion.
http://www.nessengr.com/technical-data/toroid-inductor-formulas-and-calculator/
With any transformer, the power out can not exceed the power in. So the one step missing is to find the power needed to excite that crystal to ring. one amp of current at 500 volts equals 500 watts. The transformer must also be able to deliver the current, not just the voltage.