The OP doesn't want to use a microcontroller That's the problem.
...

Yep. It's a common problem on this forum.

Some posts are like people saying "I need to fly to another country but I don't want to use an airplane... Can someone tell me how to build a fast flying machine made from old newspapers and rubber bands?"

 

To every one - Yeah I'm a Neanderthal when it comes to computers and uC's!

The variation in out put voltage is just a by product of the PWM in a VFD. The voltage is/has to be limited to the DC bus. With lower frequency it just conforms to the natural variance of PWM. And at higher frequency it is limited by the bus voltage.

If this was a product to be sold I agree totally that a uC would be the way to go. But this is just a project for my own use. Don't have a place or time to learn programing. You guys that do it, how long has it taken to get to the point your at, knowledge wise? Don't know that I have that much time left, thus trying to do it down and dirty.

 

Not sure if someone already stated it but I guess a stepper motor control with microstepping feature could be used too.
Like the A3977 for about $6 to $10.
Not as complicated as learning to program microcontrollers but it's a long datasheet too. The output stages would need to be connected to some drivers and two H-bridges since the DC bus voltage for shortbus' motor will be much higher than the max rating of this chip though.

This is what I've been saying from the beginning of the thread. Instead of the A3977 I just ordered a A4988 break out board from Pololu, thats its replacement. Data sheet - http://www.pololu.com/file/download/a4988_DMOS_microstepping_driver_with_translator.pdf?file_id=0J450 The break out board has all the surface mount stuff done and changes it to a DIP interface .

Tried to find, without success a micro stepping driver without built in H-bridges but couldn't. Will have to use the internal H-bridges at a low voltage to switch half bridge drives for the out put voltage.

Here are a out put drawing of the quarter stepping wave form and the internal block of the chip. If this works it could even, by using one of the out puts, to make a simple DC to AC inverter circuit for people that keep asking on the forum.

 

how are you going to get these voltage levels as a percentage of bus voltage as shown in the output drawing (38.27%, 70.71%, 92.39%, 100.00%)? a series of regulators?

 

how are you going to get these voltage levels as a percentage of bus voltage as shown in the output drawing (38.27%, 70.71%, 92.39%, 100.00%)? a series of regulators?

As far as I understand it it's PWM too. It's just a way to draw the resulting current waveform.

 

As far as I understand it it's PWM too. It's just a way to draw the resulting current waveform.

maybe, but the current wouldn't change in steps like that, since the motor is an inductor; it would be an actual sine wave AFAIK. That's what makes me question it as a voltage waveform.

 

Tried to find, without success a micro stepping driver without built in H-bridges but couldn't. Will have to use the internal H-bridges at a low voltage to switch half bridge drives for the out put voltage.

I had a quick look at the datasheet and I'm not so sure about the current limiting section (page 9). If it's really necessary you will need to adapt the current signal from your motor , possibly isolate it. If not, you could possible just short the Rsense inputs to GND.

Also if, as discussed earlier, it becomes necessary to change the PWM in terms of modulating voltage, I don't see how that can be done. Or is the reference input voltage for the DAC supposed to do that?

 

I'm tempted to burn a 2 phase magic sinewave program onto an 18 pin PIC and send it to him as a "Quadrature Bipolar Stepper Motor Controller, external Switches Required".

 

I'm tempted to burn a 2 phase magic sinewave program onto an 18 pin PIC and send it to him as a "Quadrature Bipolar Stepper Motor Controller, external Switches Required".

How much money would we be talking? How much is needed to link to the rest of the circuit? I'm a complete dummy about using a uC. Well not complete some parts aren't finished yet.

 

I had a quick look at the datasheet and I'm not so sure about the current limiting section (page 9). If it's really necessary you will need to adapt the current signal from your motor , possibly isolate it. If not, you could possible just short the Rsense inputs to GND.

Also if, as discussed earlier, it becomes necessary to change the PWM in terms of modulating voltage, I don't see how that can be done. Or is the reference input voltage for the DAC supposed to do that?

Why would a current limit be needed? A standard AC motor doesn't use it. The current limit sets the current for the stepper, so the driver can be used with all different size motors. Since stepper motors are almost always used 'over voltage' the current in the windings needs to be limited. Current is what kills stepper motors. The only thing the driver chip will be doing in my circuit is providing the switching 'logic' to the mosfet drivers.

 

Why would a current limit be needed? A standard AC motor doesn't use it. The current limit sets the current for the stepper, so the driver can be used with all different size motors. Since stepper motors are almost always used 'over voltage' the current in the windings needs to be limited. Current is what kills stepper motors. The only thing the driver chip will be doing in my circuit is providing the switching 'logic' to the mosfet drivers.

I meant I'm not sure about how well the chip will behave without having any current feedback. I assume it works anyway though..

 

how are you going to get these voltage levels as a percentage of bus voltage as shown in the output drawing (38.27%, 70.71%, 92.39%, 100.00%)? a series of regulators?

Those are current levels from the PWM of the driver. It is like the PWM wave form you posted yesterday from the motor drive out put. The internal logic of bipolar stepper drivers is always shown like that. But it is an actual PWM wave to the motor coil itself. Each amount of current causes the small movement of the motor.

 

Yeah Yeah I know. Don't ask if I know the meaning of "easiest".
You're going to go the stepper motor driver way...

However I found your initial request of building it without a uC quite interesting. I simulated a few hours until my computer couldn't handle it anymore. This is just for your entertainment. The circuit will provide two stepping sinewaves , 90 degree shifted that change in amplitude with frequency. Frequency goes from about 5Hz to about 120Hz.

Some components are in there because otherwise the simulation wouldn't run. Others would need to be adjusted (it's the general idea)
Removing any DC offset from the sinewaves and generating a PWM is the easy part...
I realize that all this fits into a stepper motor driver (apart from the amplitude change maybe), here it took already 9 ICs and there will be 2 to 3 more plus drivers and H-bridges...

 

Thank you prandevou. How is all of this done in a stepper driver? At one time there were stepper circuits using nothing but Cmos logic on the net. I lost them to a hard drive crash and now it is all done either in an IC or uC. I guess the old analog and linear way is gone like the dinosaurs.

Thank you and everyone else for all the answers/help in the thread. Merry Christmas and a Happy New Year to you and yours.

 

Shortbus,
There are still CMOS drivers "out there on the net", but they are generally pretty limited in functionality. If you're lucky, you can find one that has half-step and full-step. Many of them just output one phase at a time. With just one phase output, you get about 1/2 the rated power of the stepper. Full step is two phases at a time. Half step alternates between one and two phases.

 

SgtWookie - finding any thing without the internal H-bridge is just limited to full and half step. Got a few L297 chips left from another project, to play with. Just really wanted to use quarter step, although all the 'modified sine wave' inverters use the half step wave form. Oh well, just have to play within my comfort zone, started this stuff too late in life.

 

Shortbus, one method I saw a few years back used a 4017 logic IC (the one that has 10 sequential outputs for a "step" pulse input) and used some simple resistors attached to the 10 pins to produce a 10 step analog voltage waveform cycle.

It may be crude but would be easy to build with off the shelf parts. I think you can cascade 2 chips too to get a 20 step cycle if you need more resolution.

 

Shortbus, one method I saw a few years back used a 4017 logic IC (the one that has 10 sequential outputs for a "step" pulse input) and used some simple resistors attached to the 10 pins to produce a 10 step analog voltage waveform cycle.

It may be crude but would be easy to build with off the shelf parts. I think you can cascade 2 chips too to get a 20 step cycle if you need more resolution.

RB, that was one of the first circuits I found. In the Cmos Cookbook. But couldn't figure how to shift the output 90° for the second phase. I know of your Linistepper from CNCzone. It is just for unipolar, right? Do you have one for bipolar? With quarter stepping.

 

No I don't have an open-source bipolar stepper driver design. It requires the complication of 2 full H bridges (to drive the 2 motor phases) and hi/lo side driving etc, it's a heck of a lot of complication and things to go wrong at hobby level especially if you want to throw in PWM chopping which creates all those nasty HV transients.

Maybe I missed it earlier in the thread, but exactly what are you wanting to drive with this? It's critical to know the voltages and current, and motor type (is it a stepper motor?). There are lots of decent bipolar chips but all are limited to lower PSU voltages and lower currents. Then there are a number of higher power H bridge solutions but that can get complex, expensive and a lot of things to can go wrong. And if you are talking about a 2 phase variable speed mains voltage motor that's heading into expert designer territory.

 

And if you are talking about a 2 phase variable speed mains voltage motor that's heading into expert designer territory.

He is going to change a 3-phase asynchronous motor into 2-phase one. He wants a VFD from 25Hz to 120Hz.
Common VFDs change their PWM in a way that the resulting V applied to the motor applied to the motor changes with frequency. (see my and strantors previous posts). The principle of what I showed in one of my last posts, similar to stepwave 4017 proposition .

If the V/F curve is necessary I don't know if the stepper motor driver/controller can provide that.

Depending on the DC bus voltage it's probably better to use IGBT, you get them easily with higher voltage ratings.

As for the control part, creating the PWM would really be easiest done with a uC, but that is not what the OP wants.