Need Help!- DIY Driver, LMV339 chopper, 74HC08 not chopping.

[vim.au]

Thanks mostly to the fantastic information available on these forums, I'm doing pretty well with my DIY stepper driver.

I've run up against a problem I don't have the tools or experience to deal with, and I'm hoping someone here might be able to point me in the right direction.

So far I've got a microprocessor handling the step logic, and a LMV339 (quad comparator) regulating the current via a sense resistor.

In order to combine the step logic with the LMV339 chopper, I've added an 74HC08 AND gate - and it's giving me trouble.

When I AND the step logic and the chopper the 74HC08 drives the fets, but isn't limiting the current.
During testing I wired the AND gate to switch from just the LMV339 chopper and the result is _almost_ no current limiting at all, but I'm at a loss as to what that means...

anyone got any bright ideas??

Thanks,
vim

[pminmo]

Need to see you schematic. How much off time do you have, blanking time, sense trip levels, sense resistors.....etc

[vim.au]

Phil,
Thanks for the reply.

I'll work on a schematic/diagram and post it as soon as I'm done.

For simplicity of testing I've stripped back the circuit to just one comparator, one AND gate and one mosfet. All this runs a small DC motor.

off time, blanking time: I've no idea. The comparator is running without hysteresis (I'm ok with the concept, but lost practically), so it's all controlled by the sense resistor.

sense trip levels: 0.01 - 0.08v.

sense resistors: just one 1w 1ohm for this test.

Thanks.

[pminmo]

Blanking depends on drive type, unipolar or bipolar, and other factors. Unipolar can be faster, 6us or slower. For example the Allegro SLA7078's vary fro 6us to 12us depending on step size and angle.

[vim.au]

Phil,

I take it blanking is used to allow the inductor (motor coil) to discharge?

Attachment one shows just the chopper. This works fine regulating the speed of a small DC motor.

Attachment two shows the addition of the AND gate. This works but gives very little in the way of speed control (it runs very fast)

Sorry for the quality of my schematics.

[Motorcoach1]

@ Pminmo - been studying the step motor and the driver and looking at all the driver circuts out there. is it possable to use the back BEMF(of the coil) to triger the step from the motor to the logic drive so as not to get harises or gosht in the vector rastor. low drain and gate in the IRF 520 and the 820 series does not due well in the inviroment.

[neilw20]

74HC08 can only drive +/- 4ma so it can't pull the charge out of the gate of the mosfet, hence it stays on.

This sugestion will simplify your circuit.

Try using a proper current mode controller.
UC3843A - LOW COST, 8 pins
This has inbuilt oscilator, 5v reference output, comparitor, High current totem pole drive for MOSFET and current sense input and udervoltage sensing so you don't blow up MOSFET when there is not enough drive.
Have a close look at all the application notes for this chip. Just replace the output transformer with your motor.

The current sensing drops 1v across resistor, but this can be offset to 400mv with a diode in series with and a 1K resitor to pin 3 of the IC.
Cathode to sense resistor.
From the anode of the diode connect a 10K to pin 8.
From pin 3 to ground you will probably find that a small capacitor to ground, maybe 1n will make the circuit more reliable. This will stop spikes inadvertently causing early turnoff.

[kreutz]

Originally Posted by vim.au Phil, I take it blanking is used to allow the inductor (motor coil) to discharge? Attachment one shows just the chopper. This works fine regulating the speed of a small DC motor. Attachment two shows the addition of the AND gate. This works but gives very little in the way of speed control (it runs very fast) Sorry for the quality of my schematics.

Attachment 1 works almost as a linear current regulator (not as a chopper drive). Attachment 2 does not work because you don't leave time for the current to decay so there is no current control. Neilw20 is right about the poor driving current of the 74HC08 for Mosfet control.

Take a look at the theory behind chopper based current control on the following link: http://www.cs.uiowa.edu/~jones/step/


Regards,

Kreutz.

[pminmo]

Kreutz is correct, you have no off time delay. In addition the mosfet isn't a logic level mosfet, with the sense resistor you can't guarantee turn on. It's capacitive characteristcs aren't bad.

NiewlW20, 74HC08 can source and sink 20ma and they don't do too bad of job with low capacitance logic level mosfets. Not as good as a real mosfet driver, but do work for the price.

[vim.au]

neilw20: initially I had no idea what you meant regarding the mosfet gate. After a lot of reading I'm now aware of gate capacitance and have some idea about it's effect at high switching frequencies. Thanks for pointing me in the right direction!
Regarding the UC3843A - The reason I'm trying to build my own controller (apart from the educational value) is that I've given up trying to get hold of 'exotic' components. If an Australian supplier has it they likely want $25 each, and if I give in and pay the p&h from the US I just know I'm going to have to do it again for the other $1 component I need to complete the circuit... </rant>

kreutz: the Jones documents are the best and most comprehensive collection of information I've come across. While I get more out of them all the time I still only really understand a little of what is presented there.
Fortunately the information here is a little more practical!

kreutz & pminmo: so the output from the 339 needs to be massaged into a nice regular square wave, and to a frequency the HC08(and the mosfet) can handle?
I guess the output should also have a variable duty cycle, though that sounds a bit complicated. A fixed duty cycle with minimum off times should work, shouldn't it?

pminmo: the datasheet I have indicates a Gate Threshold Voltage of 2-4v.
oh. argh - I guess the catch is that the "test conditions" for that are VDS = VGS... ?
so much to learn.

I'm a bit confused about the sink current on the HC08, but seeing as all the outputs together failed to pull the mosfet down and the datasheet indicates 20mA I'm going to assume that's per input so it's not such an immediate problem.

Adding hysteresis to the 339 would slow it's switching speed down, wouldn't it?

Thanks all for your input, I'm starting to feel like this might work yet.

[neilw20]

vim.au...

$2.70 That's cheaper than a pile of discrete parts!
Under $2 for +25.

UC3843BN Farnell 988-2847

http://au.farnell.com/jsp/search/bro...equestid=69610

Set top boxes and other cheap consumer goods use these in their power supplies.

[vim.au]

Originally Posted by neilw20 UC3843BN Farnell 988-2847

Thanks. It seems farnell have changed since I gave up on them a few years back.
$10 p&h isn't good, but it's not bad either.

If there isn't an inexpensive way to complete the circuit I have I'll look into the UC3843BN option.