programmable frequency generator

Question

Can anyone help me with this.

I need programmable frequency generator which will work like described below.

Something I can setup like this

program 1) 20Khz sinus, 20 minutes -ON

program 2) 460Khz sinus, 25 minutes - ON

program 3) 20Hz sinus, 40 minutes - ON

program 4) 201Hz sinus, 20 minutes - OFF

program 5) 300Hz sinus, 20 minutes - OFF

program 6) 500KHz sinus, 40 minutes - OFF

program 7) 1,5Khz saw, 16 minutes - ON

program 8) 200Hz sinus, 40 minutes - OFF

program 9) 15Khz square, 16 minutes - ON

program 10) 25Khz square, 16 minutes - ON

program 11) 600Hz square, 10 minutes - ON

program 12) 100Khz saw, 5 minutes. - ON

Every program I can set some frequency and wave form and time how much time it will run and On or OFF for certain program. One program will run and when it ends goes to next one which is ON and skips one which is OFF until the last one with ON, after that goes to start and repeats again. But If I setup that all programs goes just one loop it will stop when last one ends

but If I setup it must take 10 or 20 loops it will goes from start to the end as many times I setup, If I setup infinity it will run whole time.

I hope it is possible to have at least three program loops (from program 1 to program 12) like this so I can choose which one to run, and hope I can set this over internet also, but it also needs to have a display where I can setup all of that if internet is not availible. Exit must go to some amplifier which give out about 10W.

I know this is a complicated project but can someone help me. Don't know what to use for this arduino or some Atmel or something else.

And nex I don't know nothing about programming so I will need some extra help here.

Something like this but with more functions like this.

[URL=www.ebay.com/.../111438305665 generator[/URL]

Thank you.

michaelkellett · Answer

I'm intrigued - what is this for ? (If you are able to tell).

The range of frequencies is such that the DDS based approach is by far the simplest. The board Shabaz suggests looks fine. I've used the AD5984 and it isn't very hard to control.

Rigol sell a 10W power amp which will work over the frequency range you need - I've never seen a cheaper one to buy than this (but I could make one much cheaper if I had to )

https://www.rigol-uk.co.uk/Rigol-PA1011-10W-Power-Amplifier-p/pa1011.htm

You might still need a low power amplifier between the DDS board and the Rigol power amplifier.

Do you have any funds for this project or does it all have to be ultra cheap ?

MK

michaelkellett · Answer

Here's a really quick and dirty 10MHz audio amp:

It's a rather evil circuit because:

1) several of the semis are operating at the max power they can stand with 10V pk-pk output

2) The output devices are no longer recommended for new designs by Rohm (but they were in LTSpice so quick to use)

The input pair runs at 10mA per transistor, 20mA down Q3 and Q4 (ie 300mW quiescent - good job it's only a model)

The output devices are only rated for 2W and that isn't enough.

However, the approach looks viable.

I'll locate some more suitable parts and spice models and see how it looks. Layout will matter, the response peaks horribly without the 3.3pF cap.

MK

shabaz · Answer

Hi Dennis, The frequency generation bit is really easy to do, but with a DDS board attached to your Arduino (or any other microcontroller). Something like this, can attach with just a few wires: https://www.ebay.co.uk/itm/AD9834-Signal-Generator-Module-DDS-Signal-Source-Board-Sine-Triangle-Square-Wave/282729270172?hash=item41d3fa4b9c:g:0z0AAOSwi0xaBlA2 A small bit of code is needed to configure it. I'm sure someone will have done it for Arduino already, and you could modify it to suit your requirements. I've used it but not with an Arduino. Modifying an existing program to suit your requirements is not difficult, but will involve you spending a couple of days doing Arduino tutorials to get familiar with its capabilities and how to write simple programs. Anyone can code for it, but does need a day or two of reading and practice. Otherwise, if this is a work project, then almost any function generator will work, if you check it has USB or Ethernet connection - then you can instrument all this from your PC to meet your requirements. Jan Cumps has a series on using LabView for example, which you can check by clicking on his name and then going to his content. The amplifier bit is not easy to do, you might get lucky and find something off-the-shelf but it may not be cheap. You're outside of the frequency range of most audio amps, otherwise that could have been an option..

Jan Cumps · Answer

yes, if you include SCPI functionality in your design, you can use the usual lab automation tools to control your instrument. You could use the PWM module of your microcontroller to generate square waves. Fairly simple opamp circuits can be used to derive sawtooth and sinus signals from that square wave. I have a full functional SCPI square wave generator here on element14, using a TI LaunchPad. You could start from there and add the analog parts to the output. Create a Programmable Instrument with SCPI - Part 1: Parser Library

Jan Cumps · Answer

michaelkellett wrote: I'm intrigued - what is this for ? (If you are able to tell). ... Do you have any funds for this project or does it all have to be ultra cheap ? MK If it is a one-off, using development boards for the microcontroller and the signal generator, and an off-the-shelf amplifier like the Rigol, I’d assess hardware costs to be somewhere around 500€. Because blisk doesn’t have program skills, someone would have to build the firmware (microcontroller talking to the AD chip and exposing an api that can be used to set the frequency and wave form). Someone (else) would also have to write the specific control program to drive the instrument to step through the particular set of steps that blisk needs. Building an API that’s as flexible as he wants it, and allows him to set up himself with no program skills, would require another level of abstraction so that it can be configured flexibly by a non-coder. If the person that makes that firmware, and the wrapper that allows blisk to flexibly configure it, asks a normal market price for the hours spent developing all of it, I assess a figure in the several 1000s of €.

rachaelp · Answer

blisk wrote: I did search about communication with PC but didn't find any specifications what can I do over PC with generator! From the pictures on the listing I can see it is a JDS6600. Entering this into google I find the operating manual: https://supereyes.ru/img/instructions/JDS6600%20Quick%20Guide.pdf From this I can see the USB presents as a standard serial device with a baud rate of 115200. It also states it uses an "open protocol" A little more googling (10 seconds or so) and I find a python library for controlling this exact unit on Github: https://github.com/on1arf/jds6600_python Best Regards, Rachael

blisk · Answer

Yes it should be ultra cheap and it is not for something for selling but for healing purpose with frequency for me.

I know that project is a bit difficult but I think it can be made step by step, that means upgrading it until I get what I want.

500€ is alot for this project.

I am here for help if that is not possible than it is not possible I will accept that.

I thought with arduino that will not be that hard to do that.

I have experiences with electronics but not with micro controllers.

Jan Cumps · Answer

If you have hardware skills, you can save a lot by making your own 10W amplifier. That one was more than 300 € so you can save on that. You can also save on the signal generator because for frequency therapy, precision and noise aren't important. You would not need the coax connectors and converters. Is there a specific need in the healing functionality for sinus and sawtooth? If not, yoj'd only need the pwm of a microcontroller, and the 10W self-built amp. And the firmware could use an SD card to get the instructions. SD card readers are available for the arduino. You would create a text file on your computer and write it to the card. Then insert the card in the arduino and power on.

Jan Cumps · Answer

Do we know the impedance of the load? It’s going to drive the power amplifier design. If will need enough headroom to deliver its 10 W into the load.

Jan Cumps · Answer

To load 10 W into a high impedance load, a high voltage is required. For DC, you’ll need 100 V if your impedance is 1 K. More volt if the impedance is higher. There are online calculators that give you the required voltage for square, triangle and sinus to achieve a load of 10 W rms.

michaelkellett · Answer

Not entirely nowhere - that generator you found is the cheapest boxed up waveform generator I've ever seen. I may well not be able to resist buying one just to check it out.

I have $11 in my Paypal account - not quite enough

And I've done some good thinking about 10Mz 10W amplifiers - mainly along the lines of how far one can take classic simple audio power amplifier design, I'm aiming for 10W into 8 or maybe 4 ohms.

MK

Jan Cumps · Answer

I suggest that you create a file with the desired patterern/sequence on an sd card, and read the contents on your arduino at power on.

shabaz · Answer

Hi Michael,

That's very interesting, it would be cool to have a nice design for a wideband amplifier that could drive such loads. It could be a nice lab instrument, and also really useful for physics experiments. I guess you're thinking about a discrete design?

Another thing of interest was a general-purpose amp x10 or x100 with switchable input offset removal and adjustable output offset. And perhaps expansion in future with 50/60Hz notch.. i.e. a lab amp. I did sketch up a circuit, but have not got any further. It was only for a 50 ohm load. I think I was shooting for 1MHz.

The use-case for it was just that it is tiresome when experimenting with sensors to have to build up an amplifier circuit each time, just to try it out. And lots of modern sensors have a 2.5V etc offset since they are single-supply.

Recently I picked up from ebay a 'biomedical amp', it is ancient, but I was hoping to tear it apart and see how they designed it.

beacon_dave · Answer

Have you tried looking at some similar projects on the likes of Instructables ? A quick search returns this one which looks like it could get you most of the way there: Arduino Waveform Generator

Jan Cumps · Answer

that’s an fpga. Way overkill for what you need, and the learning curve to build a design with it is significant. Your requirements don’t call for a display, even if run without Pc. you don’t need a button either. you need electronics that can generate sine, sawtooth and square, a programmable switch that can select one of those signals, and a broadband 10W amplifier.

shabaz · Answer

Hi Dennis,

The more software features you add, the harder it will be - you've already stated you know little about programming.

It may be better (just a suggestion) to work on some of the core functionality, and then when you are happy with that, then work on your improvements or version 2, with the extra features.

I feel you may need to pay someone to design and write the software for you, and that's going to be more expensive than an off-the-shelf function-gen. Initially I thought you just wanted sine waves, but now I see you've got square and sawtooth on the list too.

shabaz · Answer

I'm not suggesting you get this one specifically (I have never tried this one - it was just a quick google search), but if you search around, you may find something suitable. It has a display too.

https://www.ebay.co.uk/itm/Dual-channel-DDS-Function-Signal-Generator-Arbitrary-Waveform-1Hz-100MHz-X2Z4/153026952830?ha…

This is less than $100, and has an interface to connect your computer.. once you're happy it all works you could remove the computer and replace with an Arduino.

I'm sure there are cheaper options too, if you search long enough.

shabaz · Answer

Indeed.. If you want to learn programming, then there are lots of free online courses if you google for this. That will then give you the confidence to embark on this or any other project. Most people start off by flashing an LED with some code. That's only a few steps removed from adding some timing to turn bits of hardware on and off at the times you want.

There are lots of ideas in this thread, and at some point you may wish to decide on a path of action. It all hinges on learning a bit of programming though, so you can feel confident enough to then decide your path of action.

programmable frequency generator

Top Replies