# Build a CT

Discussion in 'General Electronics Chat' started by tinkerman, Jan 9, 2013.

1. ### tinkerman Thread Starter Member

Jul 22, 2012
136
34
Was hoping to build a CT using a torodial core from a power supply. Primary current will be 25 amps. Secondary needs to be in range of 15 volts. How many turns would be required on the secondary? I could do this by trial and error but there should be a way to calculate turns.

TIA

2. ### crutschow Expert

Mar 14, 2008
14,813
3,808
Need to know the size of the toroid and the characteristics of the core material.

3. ### SPQR Member

Nov 4, 2011
379
48
What's a CT?
I saw the title and said to myself - "Oh great! Someone's going to build a CAT scanner - this will be fantastic!"

4. ### mbohuntr Active Member

Apr 6, 2009
413
32
A Current Transformer is a transformer that wraps around high current cable and steps it down to be displayed at a panel etc. Kinda looks like a square doughnut with the large cable running thru it.

5. ### tinkerman Thread Starter Member

Jul 22, 2012
136
34
The core I've salvaged is from a computer PS and measures approx 6 mm x 12 mm. It would probably be a ferrite?? I should add the output is bridge rectified and feed into a 1K pot.

6. ### tinkerman Thread Starter Member

Jul 22, 2012
136
34
Yes one loop of heavy primary wire. Similar circuits are found in gen sets for load detection. ie. idle when no load is connected; full speed when it detects a load.

7. ### KnRele New Member

Jan 7, 2013
20
8
So the 25A primary should correspond to 15V across 1 KOhm, or 15 mA secondary.

The turns ratio for a current transformer corresponds to the ratio of currents:

$I_{1}*N_{1} = I_{2}*N_{2}$

or

$I_{1}/I_{2} = N_2/N_1$

Here $I_1$ is the primary current, 25A, $N_1$ is the number of turns on the primary, which is likely to be some small number like 1 or 2; $I_2$ is secondary current, 15 mA, so the turns ratio

$N_2/N_1$ would be

$N_2/N_1 = 25A/15mA = 1667$.

Meaning that the secondary should have 1667 turns for each turn of the primary.

Next concern would be with saturation. Since this is to be used for current measurement, we will not want any saturation, as that will destroy any expectations of linear operation -- the above formulas for currents and turns ratios depend on the transformer's core behaving linearly, which it will not do if forced into saturation.

As crutschow mentioned, we will need to know more about the core. The magnetic driving force on the core is dependent on the primary current, the number of turns, and the frequency. The effects are more severe for larger currents, larger turns counts, and lower frequency.

Absent specific information, I would guess that the transformer here is being made to run at mains frequency, 50Hz or 60Hz -- and if the original use was in a switching supply running at maybe 50 kHz or so, means that the lower frequency by a factor 1000 would require a corresponding reduction by a factor 1000 of the $I_1*N_1$ product, compared to whatever the maximum value of this was in the original service.

One way to find out more about this would be to build the transformer with one primary turn and 1667 secondary turns, hook it up, then increase the primary current from 0 on up, and see what happens to the voltage measured across the 1KOhm on the secondary.

tinkerman likes this.
8. ### SPQR Member

Nov 4, 2011
379
48
Ok, one of THESE.
Thanks!

tinkerman likes this.
9. ### tinkerman Thread Starter Member

Jul 22, 2012
136
34
Yes similar. The one shown looks larger and has more turns. Yes is would run at 60 HZ. Saturation did occur to me. In gen sets they detect current at one amp I believe and beyond that saturation wouldn't matter. But in this case the CT is to be used to in a charger circuit to limit output at 25 A to protect the rectifiers and transformer. I've tried a small 120 volt transformer and added one turn of heavy wire for a primary but I get too much voltage output. Perhaps a transformer with a 40-60 winding would work better. It's also bulky.

10. ### mbohuntr Active Member

Apr 6, 2009
413
32
I think so, the ones I'm used to seeing look like this

File size:
18.8 KB
Views:
17
File size:
114.3 KB
Views:
25
11. ### debe Well-Known Member

Sep 21, 2010
964
186
This one came out of a scraped 5Kw UPS from a local windfarm.

• ###### CURRENT TRANSFORMER.jpg
File size:
266.2 KB
Views:
19
tinkerman likes this.
12. ### tinkerman Thread Starter Member

Jul 22, 2012
136
34
Yup seen lots of those too.

13. ### tinkerman Thread Starter Member

Jul 22, 2012
136
34
Hey that looks like it would be just what I need. I'm going to wind more turns on this core and see how that turns out. I've looked for other small transformers that might adapt but most have no room to pass a heavy wire through the core.

Dec 4, 2009
274
37
If you can reduce your output voltage requirement of 15V on 25A, you can reduce the no. of turns. Say 1V, would bring down the turns by 15. This can be amplified if permissible in the application. Ensure that you have the resistor always connected to the secondary, and overrated in power.

15. ### tinkerman Thread Starter Member

Jul 22, 2012
136
34
Amplified would be good if I knew how to adapt the circuit. Using a transistor? I have 200 turns on the core now so I'll pass some current thru and scope the output. Yes know can't open circuit a CT. I've seen the damage that can occur on powerline equipment (800 to 5/138KV CT). Burned the terminals at the shorting block before the equipment was finally de-energized. Voltage must have been in the range of a car ignition system.

16. ### KnRele New Member

Jan 7, 2013
20
8
With 200 turns, you will probably find that the full-scale secondary current will be in the vicinity of 125 mA for a 25A primary current -- this would give you 125V across a 1kOhm resistor, so it will probably need attenuation instead of amplification. The more turns on the secondary, the smaller the current there and the smaller the voltage drop across a resistor connected to it.

Current-transformers can be scary -- even 10 mA with «nowhere to go», that is some large resistance corresponding to an open circuit, sets up some serious voltages. Consider Ohm's law: 10 mA times 1 MOhm becomes 10 kV in theory, though in practice, insulation breaks down and there is arcing and fires...

For amplification, op-amps (operational amplifiers) are relatively easy to design with if you want some sort of linear operation for metering or a reasonably accurate display. If it is a matter of determining if a current is above or below some threshold, a comparator or transistor may be able to do the job. It all depends.

17. ### mbohuntr Active Member

Apr 6, 2009
413
32
Agreed, please make sure you understand the safety issues with an open circuit! We always ground them and 2nd person verify before working on the circuit. Very dangerous!

18. ### shortbus AAC Fanatic!

Sep 30, 2009
4,295
2,221
debe likes this.
19. ### tinkerman Thread Starter Member

Jul 22, 2012
136
34
Not impressed so far with results on my doughnut CT. 200 turns. 22 amps I get less than a volt into 1K resistor. The wave form has a lot of harmonics in it too. I don't know where that's coming from. Even at much lower amps I get much the same waveform but of course lower ac volts across the resistor. 1400 more turns? Ouch.

Need a new approach.

hgmjr likes this.

Oct 14, 2008
1,669
241