Joule Thief 101

[tinman]

Mmm

Yep--definitely something wrong with my JT circuits operation--hey MH.

I bet you that Mag's would see that very same !!death spike!! that you see in my scope shot's. Maybe Mags could find the time to do the same test as i did. Then we shall see if my circuit is operating as it should,or in some sort of spastic mode as you say.

Funny the scope shot below,when i use the same test points as you had Mag's use.
Nice try,but not good enough.


Brad

[tinman]

Tinman,

What you want to end up displaying is the voltage difference across the 10R resistor. 

Connecting the two probes as you have done is the first correct step in doing so.  With both channels' VPD settings (and probe cals) identical (try 2 VPD for the 10R), you should see identical waveforms on CH1 and CH2.

Next, you will want to invert one channel as you have also correctly done.



This method effectively gives your scope a single channel differential input and is valid for both analog and digital scopes.  Inverting one channel and then adding them together is the same as subtracting the inverted channel from the other channel.



You may have to increase the vertical sensitivity (VPD), but as you do so you must keep both channel VPD settings identical at all times.  As you change the VPD settings, switch between the normal and ADD display to ensure that you are not clipping the channel inputs with too much VPD gain.  If necessary, increase the value of the 10R to 100R to increase the voltage difference across it.

PW

The next step, which you omitted, is to then ADD the two channels together.  You will now have just one trace displaying the voltage difference between the two ends of the 10R resistor.

Nope-this scope cannot do that unless the math function is used,and then the math trace is in with the other two traces. CH1 + CH2 is only available in the math function.

Using your digital scope, you should be able to use one of its math functions to directly perform the subtraction of one channel from the other.  However, being unfamiliar with your scope, I do not know if a difference trace will be displayed, which I would want to see (I am less interested in just "numbers in the boxes").

I can bring up the math trace-as in scope shots below. But the scope will not display the math trace value. This has been an issue than many have tried to solve for the past 3 year's,and to date,no one has been able to figure out how to get the math value displayed on the scope.


Brad

[picowatt]

Nope-this scope cannot do that unless the math function is used,and then the math trace is in with the other two traces. CH1 + CH2 is only available in the math function.

I can bring up the math trace-as in scope shots below. But the scope will not display the math trace value. This has been an issue than many have tried to solve for the past 3 year's,and to date,no one has been able to figure out how to get the math value displayed on the scope.


Brad

If you can display the math trace, forget about inverting one of the channels and just display a math trace of CH1-CH2...

Try to understand what it is you are trying to make your scope display, the voltage difference between the ends of the 10R.  If you cannot get better vertical display resolution of the math channel, increase the 10R to 100R if necessary to get more difference signal.

Why do these traces look so much different than your previous traces made across the 10R?  Is this the same test condition as in your earlier attempt?

ADDED:  I don't recall that harsh step in the CH2 waveform being present in your previous attempt.

Also, can the CH1 and CH2 displays be turned off leaving only the math trace?

PW

[tinman]

If you can display the math trace, forget about inverting one of the channels and just display a math trace of CH1-CH2...

Try to understand what it is you are trying to make your scope display, the voltage difference between the ends of the 10R.  If you cannot get better vertical display resolution of the math channel, increase the 10R to 100R if necessary to get more difference signal.



ADDED:  I don't recall that harsh step in the CH2 waveform being present in your previous attempt.



PW

Why do these traces look so much different than your previous traces made across the 10R?  Is this the same test condition as in your earlier attempt?

No-this is a different JT i was testing at the time of your request,so rather than swap back to the other JT,i simply did the same thing on this JT to give you examples of what i can and cannot do with the scope as quickly as i could.

Also, can the CH1 and CH2 displays be turned off leaving only the math trace?

No they cannot-the math trace switches off as soon as you switch of one channel.

Brad

[tinman]

If you can display the math trace, forget about inverting one of the channels and just display a math trace of CH1-CH2...

Try to understand what it is you are trying to make your scope display, the voltage difference between the ends of the 10R.  If you cannot get better vertical display resolution of the math channel, increase the 10R to 100R if necessary to get more difference signal.

Why do these traces look so much different than your previous traces made across the 10R?  Is this the same test condition as in your earlier attempt?

ADDED:  I don't recall that harsh step in the CH2 waveform being present in your previous attempt.

Also, can the CH1 and CH2 displays be turned off leaving only the math trace?

PW

The math function on this scope is nothing but useless,and why i never attempt to use it.

Below is what we end up with. I have manually dimmed both channels to have the math trace stand out. I now have a 100 ohm resistor on the base as our CVR.
No matter what i try,the math trace is always just a bunch of noise.
I have both channels non inverted,and CH1 - CH2.
I know what we are trying to achieve,but it must be done without using the math function(if you can call it a function at all).

Why can we not simply do it by subtracting the value of one channel from the value of the other to obtain our voltage drop across the CVR ?

Brad