Petr, whose first language is Czech, asks…
12.3 2013
Ramping capacitor – possible problem
Yesterday i was building some tDCS with LM334 and try for first time use ramping capacitor. When i test device with load (5KOhms) all was ok ramping when i turn device on and ramping down when off. But when i change load (5kOhms potentiometer) during stimulation (testing) it create current peak up to 5 mA. I test it with few different capacitor and behavior is always the same (only different value of peak and the time to return to normal ). Device without capacitor work without problem. In result of this i use instead of capacitor serial load ( linear potentiometer 100kOhms ) allows me to do manual ramping (0,07mA to setup current).
My question is can anybody test this capacitor problem maybe i do something wrong, bad multimeter etc.. . If this problem is real, it’s a very bad idea to use capacitor for ramping in use the resistance change is not too quick but still can cause pretty high current peaks.
What is a ramping capacitor? I have a MSEE and worked in this field for 35 years and I never heard of it
Sorry for reviving an old post, but I have a question about the “cap across current regulated circuit”
Is this circuit an example of this?
https://nocko.se/assets/tdcs-article/tdcs-schem-small.png
Call me Mr. Bean — I wish I had read this before putting together that “goflow-style” circuit. It was like being hit in the head with a hammer 🙁
I have added a (4.7K) resistor in series with the cathode as a voltage limiter, unfortunately the capacitor was AFTER that (directly across the electrodes) in the circuit.
I have since removed the capacitor and it seems to work, although I get a phosphene flash when I apply the power.
Also note how I said across the terminals, not between the regulator and the anode! The way you turn off a current regulated loop is to short the anode to ground, not something stupid like pull the plug or cut power!
The reason for this is that ideal current regulators do not exist. The regulator will have a transients, as does any mechanical switch. Flipping a switch, or pulling / plugging a contact results in something like a square wave at a rather high frequency. The regulator, naturally, cannot keep up with this, resulting in spikes.
So just use a pot to short the anode to ground when you turn off, then unplug the trodes, then turn off power.
Indeed!
There is nothing wrong with the capacitor per se, however there is a whole lot wrong with putting a capacitor in a current-regulated loop!
What does a capacitor do? It resists changes in voltage!
Now, imagine the following: Mr. Bean tries to do tdcs… He has a current regulated loop, which is broken, Mr. Bean has just turned it on, but the trodes are lying on the floor.
Well, the regulator tries to maintain current, which means it opens wide, and now you get 18 Volts at the terminals in a few seconds, while the capacitor charges up!
Now, Mr. Bean takes the trodes and applies them to his head. Let’s imagine an ideal current source, which has no transients. Well since the output resistance just fell from a couple of teraOhms to 3k, it steps in and reduces the voltage to get 2mA…
…but, the capacitor is sitting there right across the output! It absolutely Hates changes in voltage, so it does its best to keep the 18 Volts on the output, and Mr. Bean gets a taste of XIX. century psychiatric treatment while the cap exhausts itself (through his head!), and finally allows the voltage to drop.
Now, no particular component is at fault. The ideal voltage regulator (which doesn’t exist btw) did its job, it never once allowed more than 2mA through. The cap also did what a cap is supposed to do. It was just the idea of putting these two in such a configuration that eventually led to a rather painful current of 6mA being shorted through Mr. Bean’s head.
So end of the story, DO NOT PUT a capacitor across a current controlled loop! It’s stupid and dangerous! A capacitor stores power, and that power has an unregulated path through your brain!
I don’t care where you saw those schematics, it’s just plain wrong and stupid!
What you need is an audio taper potentiometer, which is connected up between the two terminals, allowing you to short out the loop, and gradually apply the current to the trodes.