O.T. Convert 0-150 VDC to 0-50VDC range to allow data logging.

I am helping a company who would like use their Graphtec GL220 data recorder, to record the voltage the KB DC drives are putting out to the magnetic brakes. The GL220 recorder only goes up to a max of 50VDC.

I am hoping to find help with something we can purchase, or build, to allow this. Any help would be trememdously appreciated.

Try a voltage splitter (voltage divider). In simplest terms, it is two load resistors in series. The resistor sizes are selected so that the voltage across one is 50 volts, instead of the full 150 volt supply.

http://en.wikipedia.org/wiki/Voltage_divider

R2 = R1 / (Vin/Vout - 1) = R1/(150/50 - 1) = R1/2
R1 = (R2 X Vin)/Vout - R2 = 3R2-R2 = 2 R2

To set current across R1 load to 0.020 Amp, R1 = E/I = 50/.020 = 2500 ohms, then R2 = 2500/2 = 1250 Ohms.

Your recorder output scale should be multiplied by 3 to convert the reading scale back to a range of 0 to 150 volts DC.

Lancie, thanks so much. I think it's been too long since I re-visited my electrical books.

That will work. Thanks again.

We need to know the impedance of the input to accurately select the proper resistor values. 1% to 2% precision. Lowest values would be 10K and 5K, 1 watt. That puts us at 10 mA. Add a 51 volt zener (1N4757A) across the 5K for protection.
Edit: There will be EMF from the brake coil. A TVS across the 5K might be better. Littlefuse 5PK51A. Check the maximum input spike that the analog module can handle.

keithkyll said:

We need to know the impedance of the input to accurately select the proper resistor values. 1% to 2% precision. Lowest values would be 10K and 5K, 1 watt. That puts us at 10 mA. Add a 51 volt zener (1N4757A) across the 1K for protection.

Click to expand...

I'll have to check the manual for the GL220 when I go in, and check impedence. Thank you for the help.

R2 = R1 / (Vin/Vout - 1) = R1/(150/50 - 1) = R1/2
R1 = (R2 X Vin)/Vout - R2 = 3R2-R2 = 2 R2

EDIT: To set current through R2 load to 0.020 Amp (20 mA), Rtotal = 150V/0.020A = 7500 Ohms, R2 = E/I = 50/.020 = 2500 ohms, then R1 = 100/.020 = 5000 Ohms.

Lancie1 said:

R2 = R1 / (Vin/Vout - 1) = R1/(150/50 - 1) = R1/2
R1 = (R2 X Vin)/Vout - R2 = 3R2-R2 = 2 R2

EDIT: To set current through R2 load to 0.020 Amp (20 mA), Rtotal = 150V/0.020A = 7500 Ohms, R2 = E/I = 50/.020 = 2500 ohms, then R1 = 100/.020 = 5000 Ohms.

Click to expand...

Lancie, thanks for the correction. This will work perfectly. The impedance of the recorder is high, so I think I will be ok with these numbers.

Thanks for the 'back to basics' lesson. I needed it.

I hate to be a pain. For the 5K resistor at 20 mA, it will draw 2 watts at full voltage. 100 * .02 = 2. That's a bit hot.
I chose larger values to cut the current in half. The goal is to use the largest value resistors possible. Too high, and the input impedance will add an offset to the divider.

I'd buy an isolator signal conditioner.

If the signal ground for the AI is different than the signal ground for the 150V source, the common mode could easily fry the AI.

keithkyll said:

We need to know the impedance of the input to accurately select the proper resistor values. 1% to 2% precision. Lowest values would be 10K and 5K, 1 watt. That puts us at 10 mA. Add a 51 volt zener (1N4757A) across the 5K for protection.
Edit: There will be EMF from the brake coil. A TVS across the 5K might be better. Littlefuse 5PK51A. Check the maximum input spike that the analog module can handle.

Click to expand...

keithkyll said:

I hate to be a pain. For the 5K resistor at 20 mA, it will draw 2 watts at full voltage. 100 * .02 = 2. That's a bit hot.
I chose larger values to cut the current in half. The goal is to use the largest value resistors possible. Too high, and the input impedance will add an offset to the divider.

Click to expand...

Keith. This is a very good point. I'm not too worried about heat, as the brake voltage only peaks up to 90V or so, for a period of 8 seconds max. This voltage divider will be hooked up to a $1000 Graphtec GL220 data recorder. As I will probably have to order resistors anyway, I would like to go with the most accurate, and safest to the recorder, option.

If I get the channel impedence of the recorder, can we come up with the safest way to protect the recorder. The brake coil this is energizing is small. It is only keeping tesion on thread.

Hold voltage (output while line is stopped) = 30V

Run voltage (output while line is running) = 9V

Deguess voltage (reverse output to remove residual) = -30V I think. If it is more, I may have to make my range 0-200V instead.

Stop voltage (output while line is stopping) = peak to 90someV for no longer then 8 seconds.

The system does deguess the coils every so often, so I have to keep that in mind, as I figure this out.

The TVS is the most important. If the impedance is high enough, we can go to 2 mA. 50K and 24.9K. When the TVS clamps, all of the surge will get dumped across the 50K. TVS's are great, but if we overload it, the typical failure mode is shorted. Keeping the resistors high in value helps minimize TVS failure, by limiting current.
The TVS will fully conduct, just like a diode, when the degauss voltage goes negative. Ideally, we would like to disconnect during degauss.

danw said:

I'd buy an isolator signal conditioner.

If the signal ground for the AI is different than the signal ground for the 150V source, the common mode could easily fry the AI.

Click to expand...

signal ground will be the same for the brakes

Dan, I searched around, and couldn't find a signal conditioner capable of this. Is there any that you know of?

Bob

O-200Vdc input, very high isolation.

http://www.api-usa.com/pdf/api/api_4380ghv3.pdf

I wonder if you may be better off sensing current. That way you would know if you got an open coil or loose connection. Also current proportional to field which is proportional to force thus braking action.

Could do with a shunt but I would give serious thought to a hall effect current sensor with either 50 mV or 0 to 10 output. The hall effect would also isolate you from the 150 line to solendoid - added safety for personnel and for instrument.

Dan Bentler

sure please use an isolator, and do not use voltage but current. take it off from the ground side of the brake so in the return.
You will have to filter it as the output will be PWM or alike.

btw when resistors are used a standard one can only have 100 volts, so please use some in series.