EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: BershaM on July 24, 2016, 03:22:14 pm
-
the circuit i made opamps and a darlington pair to drive the motor causes lots of oscillations (parasitic maybe ?) which shorts out both power supply rails through the transistor , a solution to that which i found very reasonable is to limit the bandwidth of the opamps so any oscillations are attenuated/removed, i tried inserting 220nf caps across R24 but it didnt help alot, i tried a zobel network of 100nf and 10 ohms but still, i tried loading the feedback to ground via a 100nf cap but again nothing happen and i still get the oscillations , any idea how to track them down and filter them out ?
PS :1- for some reason the PNP transistor gets affected by the oscillations way more than the NPN one .
2- Removing the 1uF cap (C42) causes alot more oscillations than before (perhaps there are several sources of oscillation?) , i read online that the culprit is mostly the negative feedback to pin 13.
(http://i.stack.imgur.com/BJEso.png)
-
The problem most likely is the second stage excess phase shift due to PNP
causing problem. And Darlington aggravates the issue due to much lower Ft.
Some comp techniques shown here - http://www.ka-electronics.com/images/pdf/Op_Amp_Booster_Stages_pt1.pdf (http://www.ka-electronics.com/images/pdf/Op_Amp_Booster_Stages_pt1.pdf)
http://forum.allaboutcircuits.com/threads/output-power-boost-for-op-amps.117595/ (http://forum.allaboutcircuits.com/threads/output-power-boost-for-op-amps.117595/)
http://cds.linear.com/docs/en/application-note/an18f.pdf (http://cds.linear.com/docs/en/application-note/an18f.pdf)
Regards, Dana.
-
Why does the circuit have a PNP transistor at all?
-
Why have two stages? As a starting point I would remove the first stage and put the gain of 10 on to the second stage - such that its loop gain is lowered - giving (hopefully) more phase margin.
-
@danadak i'll try them out thanks for your feedback.
@Mosher because i wanted to control the direction too.
@Andy what if a single stage with a gain of 10 isn't enough ? what else can be done ?
-
i'm suspecting motor noise get into the feedback line generating nasty effect, try putting diodes from gnd to +15 and gnd to -15v psu.
Why does the circuit have a PNP transistor at all?
i guess he want the motor to be reversible, so instead of building h-bridge from single supply, he's using bipolar 2 transistor bipolar psu from dual rail supply.
-
This might be a stupid question - but is there a connection to 0V of the (hopefully split rail) power supply?
-
@Mosher because i wanted to control the direction too.
OK, just checking you know why the PNP is there and not just copied some circuit you saw somewhere.
i made opamps and a darlington pair
I would not describe that NPN and PNP as 'Darlingtion'
They are in what I would call 'push-pull' arrangement. (anyone care to elaborate or disagree)
i tried inserting 220nf caps across R24
Yes, that will not help because that is not putting a filter on the op-amp.
That is putting an integrator on the op-amp.
And it is the wrong op-amp
I would go for the attached, ignore Rout, the filtering element is C1
2- Removing the 1uF cap (C42) causes alot more oscillations than before
So, I think the 1uF cap is along the right lines for the filter.
Incidentally, congratulations, what you have built is a basic variable voltage regulator (which is also able to go into -ve) :-+
Try looking up fundamentals of voltage regulators and why/where the filtering need to be inserted.
What your circuit is lacking, is short circuit protection. careful not to short the output, it will destroy the transistors.
-
@Mosher because i wanted to control the direction too.
OK, just checking you know why the PNP is there and not just copied some circuit you saw somewhere.
i made opamps and a darlington pair
I would not describe that NPN and PNP as 'Darlingtion'
They are in what I would call 'push-pull' arrangement. (anyone care to elaborate or disagree)
The wrong symbols are used on the schematic. The TIP142 & TIP 147 are Darlington transistors.
What's the maximum voltage required by the motor?
Even if the oscillation can be eliminated, it's quite likely the output stage will drop too higher voltage to power the motor properly.