# DC-DC Isolator

Discussion in 'Power Electronics' started by gillesfan, Nov 5, 2017.

1. ### gillesfan

17
1
Nov 5, 2017
Hello Guys

I was wondering if someone could help me with a circuit?

I am trying to read the inputs from a slot car track to a computer ADC. My attempts to directly connect the track have resulted in dead pins on the computer.

My initial attempt being
Code:
```0-22V DC ------------
|
10k Ohm
|
|---------------------150 Ohm-------------- Computer
|
2K2 Ohm
|
GND----------------------------------------------------------- Computer GND
```
I suspect the attempt to connect Computer and slot track GNDs was the root of my problem.

How can i introduce a DC-DC converter which will isolate the grounds?

It is important the output follows the input proportionally.
I know this would be fairly trivial if the input was AC, unfortunately its not and i really need to isolate the output.

Last edited: Nov 5, 2017
2. ### duke37

5,364
769
Jan 9, 2011
The motors in the track will generate voltage spikes which may be very high. The reduction in voltage due to the input resistors will help but may not be adequate.

Put a zener across the ADC input to limit the positive and negative voltages.

3. ### Harald KappModeratorModerator

10,768
2,425
Nov 17, 2011
Having the track isolated from the computer is a good idea. But this is not a dead simple thing. Several options come to mind. The most simple is this circuit:

By splitting the10 k resistor of the voltage divider into 2 * 4.75k resistors, you provide a high impedance part in bot ground and power rail. This is not complete isolation but protects bthe computer's input from high currents. Additionally the zener diode as suggested by duke protects the input from voltage spikes while the small capacitor helps to smooth the DC input for better stability when doing the A/D conversion.

A more sophisticated scheme can be built using a linear optocoupler, e.g. the IL300.

4. ### gillesfan

17
1
Nov 5, 2017
Thank you guys.
This information is a big help.

I am currently using simple optocouplers for the timing system, the linear Opto couplers sound ideal. In my ignorance i was unaware of their existence.

I shall start experimenting with them.

All the Best

5. ### gillesfan

17
1
Nov 5, 2017
Well i seem to have a system which tracks the changes well but the computer ADC is not seeing it.
I'll let you know how it goes

6. ### Alec_t

2,909
791
Jul 7, 2015
Exactly what 'inputs' are you interested in ? Just the track supply voltage; or some track-superimposed signals ?

7. ### Harald KappModeratorModerator

10,768
2,425
Nov 17, 2011
Upload a circuit diagram. It's hard to check what you're doing without one.

8. ### gillesfan

17
1
Nov 5, 2017
HI

I think i have killed my IL300s

It was working like this yesterday but now it is not.

In essence, what is happening is that the IL300 seems to be deadening the input of the LM741

The input goes into the LM741 just fine and comes out with the waveform i expected. A DC offset with half sine on top.

This is all fine until i connect it to the IL300, then instead of transferring the signal across as it did yesterday it deadens it to a slightly lower DC.

The same think happens with all my IL300s and did not happen yesterday.
I have checked my circuit several times but the problem remains.

Last edited: Nov 13, 2017
9. ### Harald KappModeratorModerator

10,768
2,425
Nov 17, 2011
The 741 is completely unsuitable for this application. You need a rail-to-rail opamp like e.g. the OPA341.
Also Track - should be connected to DC gnd.

Plus you need an output amplifier to generate a useable signal for the ADC from the weak opto-output. See this application note.

10. ### gillesfan

17
1
Nov 5, 2017
Is it likely i have killed my IL300s? Is there any way to check them?

Diode Checks
IL300
Pins 1,2 -- Infinity and 1880
Pins 3,4 -- Infinity and 770
Pins 5,6 -- Infinity and 770

OPA341s seem hard to source are there alternatives?

To clarify the arrangement
THROTTLE MONITOR (for calculating fuel consumption and controlling sound)

My electronics knowledge unfortunately is limited (more familiar with coding) i chose the 741 as a direct replacement for 241s in an example circuit i saw.

Ideally an output of 0- 3V or 2.5-4.5V should result i'm not bothered.

Any help with what i am doing is hugely appreciated

The DC polarity was used for the timing system which is as follows

TIMING CIRCUIT

Where
A is 470 Ohm
B is 1K8 Ohm
C is 1K8 Ohm

THis has been stable, but i admit i am struggling a bit. I may have drawn my diode teh wrong way round. It's the right way around in the circuit.

Any help with these circuits is really useful.

Last edited: Nov 13, 2017
11. ### gillesfan

17
1
Nov 5, 2017
I am not interested in anything superimposed its only superimposed because of the timing system

What i want to do is use the computer to monitor throttle position .
Its currently sampling the timing every 1/1000 sec, and also two ADCs for which i was going to sample throttle position, average these every 1/20 sec or so and then use that to approximate fuel usage, engine RPM (for playing digitised sound) and detect skid/accidents (when the car leaves the slot, the Voltage increases from the 0-15V range up to about.

Initially i connected the comouter directlt to the track (with resistors to reduce the voltage to avoid over voltage, and diodes to avoid reverse voltage and capacitors to avoid spikes.
Unfortunately i still killed a couple of pins on the computer which i suspect must have been to floating GND between track and Computer

Thus i introduced the opto couplers.

12. ### gillesfan

17
1
Nov 5, 2017
My test case is using a rectified sine wave source but the actual slot track source will essentially be a varying DC from 0-22V

13. ### gillesfan

17
1
Nov 5, 2017
OK well with a different circuit i can confirm my IL300s are not dead.
Thus 1880, 770 and 770 seem to be sensible numbers.

I will push on with the new circuit and see what happens

14. ### BobK

7,682
1,686
Jan 5, 2010
If those are millivolts on the diode test range then they are quite reasonable. 1.8V for the LED and .77V for the photodiodes.

Bob

15. ### gillesfan

17
1
Nov 5, 2017
Hi i assume that is what they are, it is the diode setting on my DVM, and yes 1880 for the LED and 770 for the photodiodes yes.

I have a system working now with the 741s.

It is not ideal and my ADC is exhibiting very odd results when the IL300 is connected that is, when the circuit is not connected to the ADC i get a steady voltage.

Yet when the ADC IS connected, i get a sine wave at about 500Hz. Both on the ADC resonse AND on a scope connected.

Any idea why my ADC is overlaying a sine wave? ANy idea how to eliminate it?

16. ### gillesfan

17
1
Nov 5, 2017
Well i am not disatisfied wth the signal until i connect it to my ADC, which seems to turn it into quite a different signal.

17. ### gillesfan

17
1
Nov 5, 2017
Revised circuit associated with the traces above

I have replacement rail-rail op amps on their way.

18. ### Alec_t

2,909
791
Jul 7, 2015
Why is the 741 getting its +ve supply via R2?

19. ### BobK

7,682
1,686
Jan 5, 2010
And where is the required buffer ampl after the IL300?

Bob

20. ### gillesfan

17
1
Nov 5, 2017
Ahhh there has to be a buffer amp? i do not have one in there will that make a difference? I have seen circuits with and without i was trying to do it step by step and i have not put a buffer amp in after the IL300 so far

Well the output is essentially the same except at the ADC where the out put is different but still not what i want.

Last edited: Nov 16, 2017