I am going to drive a coil (about 100 uH and 5 ohm) to transmit sound signal (ex. swept) to a vibrator. The frequency should be around 100 kHz. The first issue was that the card does provide 250 kS/s which is only twice the sampling range I need. As a result, I don't have a nice Sine wave as a carrier so I decided to use a triangular wave which is read from an excell sheet with freq. 120KHz and sample it with 240 k. So sampling is one of the issue I have had!
Second, with a serie tuning circiut (RC) on the transmitter side before the coil, I don't get more than 40mV at the vibrator which is still low and I need at least 100mV. Then I decided to use a parallel circuit to provide more current but I ended up a clipped voltage in transmitter side and as a consequence a distorted sine at the output. I don't know if it is practically possible to use parallel RC for tuining!
DAQ I am using is 6218 BNC!
I appreciate any help in this issue!
All the Best,
Here is the thing I would give a try:
You get 5V 50mA from the +5V output (or use a external power supply)
You use a digital output in a class C scheme to drive a transistor that run the coil /capacitor circuit.
With the digital output you can use a PWM to AM modulate the 120kHz (or 180kHz 🙂 ) The DO is much faster than AO, and at 120kHz you are close class C anyway 😉
You say AM modulation but whant to sweep the frequency (FM) ??
Is the frequency of high importance? If you want to transmit AC POWER at resonance : Build a Royer converter with your sending coil and use the AO to control the DC power supply....
And you should try LC in parallel with a 5k in series (to protect the AO) to the AO.
Thank you Henrik!
If I understand well, your idea is using the PWM for modulation and a transistor outside the card as class C amp. top drive the coil. I have not used PWM so I need to read about that! But what I did is simple multiplication in LV for the modulation (attached pic) and the best is to not have an external module unless I have to. Becasue the card is an HV isolated and if I use anything else then I'll have go under so much regulation stuff.
About Sweep, my modulating signal is going to be swept sine ( up to 10 kHz) but now , I just use a pure tone of 2 kHz!
Sieht aus als wenn Du LV in Deutsch verwendest 😉 .....however...
have you tried this ?
I you run it at 100kHz measure the voltage at the coil .... And run a sweep over the exitation frequency to find the resonance.
(If you have a sine genrator and a scope... even better.)
If you are a student or/and have the sound and vibration toolkit available. just give the Frequency Response with Swept sine a try 😉
Still didn't get the HV part. is your hole usb device isolated?
It will still need power, so you can't use that 5V powersupply to?
(the spec say that the USB621x have a fused 400 mA maximum load , standard USB isolators should provide 500mA (or more) so still 100mA for a decend coil driver 🙂
what is the difference between a coil hooked up to the 6218 and a coil with a small circuit?
Seems to you want to send power and information (modulated) to another coil....
Provide more information and I migth give you some more hints...
The reactance of the coil at 120 kHz is about 75 ohms. The total voltage across the coil including the internal resistance at 2 mA is about 151 mV. Most likely you are hitting the current limit of the AO circuit. You really are going to need a buffer amplifier.
It is not clear what you are trying to do with the AO. You cannot get any kind of "waveform" when generating a signal at 1/2 the sampling frequency. You get exactly two samples per cycle. What two samples do you want? If you sample at the zero crossings for example, all the samples are zero! At any other point you get samples which can only be meaningfully interpreted as peak values because you have no phase information.
Why do you want a carrier frequency of 120 kHz? If you are trying to drive the device at resonance, what happens if the resonant frequency is above the Nyquist frequency for your DAQ device? How will you sense and adjust for resonance?
I think that before we try to solve the drive and tuning issues, we need to clarify what the real requirements for the system are. What determines the carrier frequency or range of acceptable frequencies? How much drive current, voltage, or power is required? What determines the modulation frequency, amplitude, and waveform?
hi Lynn, it's again the cristal ball issue 😄
Grüßtig Henrik! Richtig, das ist ein Dtsch LV 😉 and Hi Lynn!
The serie resistance with parallel Cap. looks good in simulation! I will try it again, I tried before with an extra resistance parallel with capacitance. Measuring the resonance freq would be easy since I have access to FFT analyser.
About the isolation, the whole USC card is 1000 v-isolated device so I don't have to deal with regulation stuff but if I add anything even with a battery then I should do some regulation! However, I think somehow, I will have to!
About the samplig, since I cannot use a sine wave as carrier (At least I need the 5 times of my carrier freq 120 KHz to get a proper sine) so I use triangular wave! I have for example two-level voltage in my excell data (+-1) and I sample with 240000 as the rate. As a result I will have a triangular wave with frequency of 120k with defined number of sample (f.ex. 2048 sample), then I make a loop to have this continuously on the analogue output coupled (AM) with my desired signal (a low freq pure sine as a sound). I have a problem in programming I guess since there is jump in the signal when they are multiplied to each other as you can see in the first figure but technically transmission works!
About the resonance Freq requirement, I have possibility to choose from 100 kHz to 130 KHz but whatever else I built so far, is based on 120 kHz and my demoulator works in this range as well! the load at the receiver-side after demodulator and tunning can approximately be simulated by a 50 ohm resistance plus 30uH inductance. I need at least to provide a 100 mVrms to be able to move the vibrator.
Thank you guys for the time you spend on this,
If you are sampling at 240 kHz and that data is two levels alternating at 120 kHz, then you do NOT have a triangle wave. You have a square wave. If it looks like a triangle wave, then something must be filtering or integrating the signal. The default setting for a LabVIEW graph is to draw a line interpolating between data points so it will look like a triangle but there is no data between the two levels.
Time: t0 t1 t2 t3 t4 t5
Signal: +1 -1 +1 -1 +1 -1
There are no intemediate values so the only possible waveform is square!
Theoretically, to get a decent triangle waveform requires even more samples per cycle than a sine wave because of the harmonic content.
Do I understand correctly that you need >= 100 mV rms of the low frequency sine wave to be delivered to your 50 ohm resistance plus 30 uH inductance load?
I think you are right that's a filtered squar wave which looks like triangular ones but since my load (vibrator=can be electrically simulated with a 50 ohm and 33 µH) after demodulation does not sense any frequency above 10 KHz so I do not care about the harmonics of squre wave or triangular waves. Any wave that I can use as a carrier would work fine!
You are right my load can ellectrically be simulated with a resistor and inductance but it is magnetic vibrator.