# Constant AC current source

Discussion in 'General Electronics Chat' started by aq_blues, Apr 30, 2012.

1. ### aq_blues Thread Starter New Member

Jul 20, 2011
12
0
Hi All,

I am tasked to design an AC constant current source working at 400Hz and can deliver a current upto 200mA to a transformer. The transformer is a step down (voltage) transformer which converts the current to 1A. Power Supply remains at +/-15V and load ranges from 0-2 Ohm (on the secondary side of the transformer)

I am simulating the circuit and simulation results look good. But before I build this, would like to know if having transformer as a load would place any issues with the design ?

Also all these discussions on AC current sources never talk on bandwidth. Is bandwidth not an issue at lower frequency ? Does the circuit respond to it really quickly ?

What about current regulation for varying load conditions ? Like how would the circuit respond to a 50% load regulation ?

Thanks for reading the post this far and a reply would be so appreciated

AQ

2. ### t06afre AAC Fanatic!

May 11, 2009
5,939
1,222
What will your voltage requirements be on the primary side. Have you any idea of this. From a quick estimate I got around 75 volt amplitude value at least. Not taking any loss at all into the estimation.

3. ### aq_blues Thread Starter New Member

Jul 20, 2011
12
0

If 2 Ohm * 1 A = 2V is on the secondary, then primary would be 10V. (A step down transformer of ratio 5 is expected here. It would step current from 200mA to 1A).

But apart from the signal levels expected at either side, would you be able to suggest any topology which I can use here. I tried the improved howland current pump, but the load regulation is not so good (current changes a little with changes in load resistances).

4. ### MrChips Moderator

Oct 2, 2009
12,642
3,457
A transformer or any power source with low internal resistance will provide voltage. A step-down transformer with a 5:1 turns ratio will transform 10VAC to 2VAC. The current drawn will depend on the load attached.

To create a constant current source you need a supply with high internal resistance. This can be implemented with an active circuit.

5. ### t06afre AAC Fanatic!

May 11, 2009
5,939
1,222
Ok so you want to take 10 volt 0.2 ampere AC and step down to 2 volt 1 ampere AC. You wrote,So it looked like you wanted a +/- 15 volt on the secondary side
But now the whole thing make more sense. I see some issues here. First if the transformer is made for 50 to 60 Hz. It may(will) be less ideal at 400Hz. Another thing is that the transformer models in most ECAD system. Are often very idealized. So transforming your results here to the real world. Must be done with care. As a driver I would have used some audio amplifier. They are ideal for this purpose.

6. ### aq_blues Thread Starter New Member

Jul 20, 2011
12
0

In order to do what you suggest, you would require very high voltage levels. Also your load regulation would not be very good (the current would change with changing load resistance, but maybe by a small amount).

What I require is a true current source. A way to do it is based on voltage feedback. You adjust your output voltage based on a feedback signal. This essentially makes it a current source. An example of such approach is the Howland pump.

With all that said, still I simulated howland pump and the results are still not upto the specs.

Any more suggestions are warmly welcomed here

7. ### aq_blues Thread Starter New Member

Jul 20, 2011
12
0
Thanks t06afre for the reply again.

I can get the required transformers for 400Hz. Also Ecad may not be able to simulate the parasitic inductances/capacitance. But apart from that, I want to see the concept working with ideal sources first. Once that's done, I would see if i can factor in the parasitic elements as discreet elements and then monitor the effects.

But again, first thing first, which topology offers me the best load regulation and lowest THD ??

8. ### MrChips Moderator

Oct 2, 2009
12,642
3,457
Well if you know so much about current sources why do you need help?

9. ### aq_blues Thread Starter New Member

Jul 20, 2011
12
0
I still am unable to meet my specs of load regulation (0.3%). Am looking for a better approach to do this.

10. ### Ron H AAC Fanatic!

Apr 14, 2005
7,050
657
Could you put the transformer inside the feedback loop, i.e., take your feedback from the secondary side?

11. ### aq_blues Thread Starter New Member

Jul 20, 2011
12
0
Even if I just use a resistive load (and not to use secondary), the difference in current is quite visible.
Attached pictures for reference.

Where R3 is considered as the load resistance

12. ### Ron H AAC Fanatic!

Apr 14, 2005
7,050
657
I think your op amp is saturating in the first example. 295uA RMS through 44.8k is over 18V peak. What are your op amp's supply voltages?

13. ### aq_blues Thread Starter New Member

Jul 20, 2011
12
0
It's actually showing the peak values. My supply is +/-15V. So the peak is at some 13V.

14. ### Ron H AAC Fanatic!

Apr 14, 2005
7,050
657
What is the value of VG1?

15. ### aq_blues Thread Starter New Member

Jul 20, 2011
12
0
Sinewave ... 1mV amplitude

16. ### Ron H AAC Fanatic!

Apr 14, 2005
7,050
657
Last edited: May 2, 2012
17. ### aq_blues Thread Starter New Member

Jul 20, 2011
12
0
Thanks for the reply Ron H,
Indeed you were right and there was saturation on the output. I increased the sense resistance and circuit is altogether better. Although it's not regulating to 100% accuracy, but the improvement is worthwhile.
I am reading more on improved Howland design and I can feel that 100% accuracy may not be achieved with this topology.
So I am again in a quest now to see if I can use transistor with an opamp and voltage as a feedback mechanism to do better regulation.
Would keep the forum posted on progress (if any )
Analog electronics is a tough game. Thanks to simulators, I would have lost it otherwise

18. ### Ron H AAC Fanatic!

Apr 14, 2005
7,050
657
What is the range of constant current you are trying to achieve?

19. ### aq_blues Thread Starter New Member

Jul 20, 2011
12
0
My specs are :