High Voltage Amplifier and Frequency Divider

[TChi]

Hello everyone,

I could not find the solution for my problems by searching the forums. Thus I am writing them here so would you guys please help me.
There are two problems but it is not good to create two topics for this so I combine them here:

1. High Voltage Amplifier:
We often work with Piezoelectric actuators (http://www.thorlabs.com/newGroupPage9.cfm?objectGroup_id=1076). Usually, this kind of actuator needs a driver with the voltage range of 0 -> 150 Volts and we drive it by using the Piezo driver to amplify the input signals: 0 V -> 10 V to: 0 -> 150 V. Practically, that driver is simply the high voltage amplifier (the amplified ratio is ~ 15). Thus we want to make our own amplifier instead of buying the commercially expensive product.
I searched the internet for the high voltage operational amplifiers but most of them can only offer the voltage swing ~ ±40V (http://www.national.com/ds/LM/LM3875.pdf , http://www.mskennedy.com/client_images/catalog19680/pages/files/0004r_.pdf , http://www.microsemi.com/datasheets/SG143.pdf) that do not work for this case.
Would you please give me any idea that I can actually make this kind of amplifier.
The output current does not need to be high. The output power consumption is small.

2. Frequency Divider:
I have a train of square pulses: pulse width is 3 µs, pulse height is 3.7 V, repetition rate is 1 kHz. Actually, I have a circuit that uses the HD74HC14 (Hex Inverters: http://pdf1.alldatasheet.com/datasheet-pdf/view/248446/RENESAS/HD74HC14P.html) and 74LS290 (Decade Counters: http://pdf1.alldatasheet.com/datasheet-pdf/view/27418/TI/SN74LS290D.html) to divide the frequency to 20 and 50 Hz (because the 74LS290 is the divide-by-ten(five, two) counter). For example, my circuit can offer the 20 Hz signal whose width is 5 ms and the interval between two successive pulses is 45 ms (20 Hz). However, I want to make the frequency divider that has the frequency range from 0 to 100 Hz with 10 Hz step. Thus I am struggling with the 30, 60 and 70, 80, 90 Hz.
I found the topic "burst of 40 kHz" (https://www.physicsforums.com/showthread.php?t=405313&highlight=counter) in which there is a valuable post by vk6kro. Nevertheless, I could not use any of his provided information for my problems.
Could you give me any idea, any name of IC that I can make this type of circuit.
I understand that practically for a given 1 kHz signal, it is impossible to get the ideal 30, 60, 70, 80, 90 Hz signals but any idea is welcome, even for making the similar frequencies.

Thank you.
Regards,
Tran Trung Luu

 

[vk6kro]

There are some audio amplifiers that are like opamps but with much higher power ratings.
These have voltage limits of about +/- 40 volts, but they can be connected as a bridge circuit to double this. I saw one that had two amplifiers in the one package, but I don't recall the number on it.
Maybe TDA... something. The TDA7294 may be an overkill, but it is typical of the devices available.
You would have two such amplifiers and feed them out of phase then connect the speaker between the outputs, not to ground.


You would have to use phase locked loops to get oddball frequencies.
You could do this at 10 to 100 Hz, but it would be better to do it at 1000 to 10000 Hz and divide down by 100 to get the final frequency.

So, you have a voltage controlled oscillator covering 1000 to 10000 Hz. You divide this down to give 1000 Hz out.
You compare this with the reference oscillator and send an error signal back to the VCO to pull it onto exactly the right frequency.
For example:
The VCO is on 8000 Hz. Divide by 8 to get 1000 Hz. Compare this with a precise 1000 Hz and if there is an error, send a DC voltage to the VCO to pull it onto exactly 8000 Hz.
Then divide the 8000 Hz to 80 Hz by dividing by 100.

Sounds easy when you say it quickly, but PLL design is fairly difficult. Especially, the filter that produces the DC for the control line. It often becomes a matter of build it and get it going by trial and error.

If this is for a serious application, you would be better off spending the money on a DDS kit. There are plenty of these on Google. This is Direct Digital Synthesis and typically allows you to set the output to any frequency between 1 Hz (or lower) and 30 MHz with a resolution of 0.1 Hz or better.

 

[TChi]

Dear vk6kro,

Thank you for your message.

About the high voltage amplifier, maybe you said in a fast way so I could not catch you. If I understand what you suggested correctly then by using two TDA7294 amplifiers, I can get ±40 V for each amplifier and double it to get 160 V (peak to peak) by "feeding them out of phase" (I do not understand this phrase. Do you mean that I should make one pi shift?) Would you please tell me clearer?
Actually, I just found one site that has quite a good information:
http://www.physics.udel.edu/~mitchell/Download/High_Voltage_Amplifier/HVA.htm
They use the high voltage operational amplifier PA84A:
http://pdf1.alldatasheet.com/datasheet-pdf/view/87168/APEX/PA84A.html
Do you think that it is an easy way?

Your suggestion about the PLL application is too good for me. Thank you about it, but as you have said, PLL is rather hard than any particular straight circuit. I just want to look for a less complicated circuit and cheaper. (the convenience and price have more priority than precision in my circuits)
We also bought several Function Generators from SRS and they worked well. However, our aim is to filter these oddball frequencies out from the given 1 kHz. Would you please tell me about any possibility. We do not need such those precise frequencies, even 58, 62, 68 Hz, ... are enough for us.

Thank you very much.

 

[Mike_In_Plano]

Amplifiers for Piezo appplications can be much more difficult to design than you would think. The Piezo stacks look to be large capacitors. This loads the output of the amplifier, making it unstable.

Perhaps, you can generate your 1 kHz signal from a 2kHz, or 4kHz source. Then divide the source down using a CMOS counter such as a 74HC4040 to get your lower frequency with a bit more precision.

 

[TChi]

Amplifiers for Piezo appplications can be much more difficult to design than you would think. The Piezo stacks look to be large capacitors. This loads the output of the amplifier, making it unstable.

Perhaps, you can generate your 1 kHz signal from a 2kHz, or 4kHz source. Then divide the source down using a CMOS counter such as a 74HC4040 to get your lower frequency with a bit more precision.

Thanks for your suggestions.
Would you please tell me your recommendation about the Piezo application?

About the frequency divider, the 1 kHz is the given frequency that I could not alter it. However, I tried to use the counters as you said and I got some good frequencies. I will try more about this.

 

[Mike_In_Plano]

For resonant frequency applications, there are circuits that self oscillate. Forgive me, becuase I cannot remember any of them, but I do know that they are commonly used in ultrasonic cleaners.

For a laboratory amplifier, you might consider purchasing a large audio amplifier (If the frequency is under under 20 KHz). Most large audio amplifiers can generate +/-45 volts or more. If you use a unity gain, inverting op amp prior to the input of one channel of the amplifier and connect the other channel directly, than the difference voltage between the output of the channels can reach +/-90 volts.
If you do this, you'll need to stabilize the output of the audio amplifier. Otherwise, it will go into high frequency oscillation, and will likely burn out. Stabilization can be as simple as a large wattage 33 ohm resistor in series with the stack. In addition, you may wish to place large wattage 15 ohm resistors on the output of each channel. This will help the amp see a fairly resistive load on the output.
Confirm the amps stability with a scope, and don't leave it on long it oscillation appears.

Best of luck,

Mike

 

[TChi]

Hi,

For my frequency divider, I found out a divide by three circuit based on the 555 timer here:
http://www.national.com/ds/LM/LM555.pdf -> It is good enough for me.

To Mike_In_Plano: thank you for your idea about the stabilization of the output of the amplifier. It's worthy and I will keep that in mind.

You guys told me about several ideas and I would like to try all of them. I will share with you guys once I get it done.

Have a nice new week.
Tran Trung Luu