I have a voltage output from a piece of kit that ranges from 0 to 7Vdc and I want to interface this to an ADC that requires 0 to 5Vdc. Normally I'd think - simple - voltage divider - BUT the 0 to 7Vdc is already the result of the kit using a voltage divider scaling down from 0 to 300Vdc. I'm thinking you can't just stick a voltage divider on the output of a voltage divider so any ideas?

Thanks

 

opamp buffer followed by a divider?

 

Why not change the voltage divider on the O/P of the original cct. If it is reducing it from 0 -300V down to 0 - 7V changing it to 0 - 5V shouldn't be very hard at all. Having said that if the ADC is expensive I like the idea of the OP Amp buffer.

 

Why not change the voltage divider on the O/P of the original cct. If it is reducing it from 0 -300V down to 0 - 7V changing it to 0 - 5V shouldn't be very hard at all. Having said that if the ADC is expensive I like the idea of the OP Amp buffer.

Thanks for your reply. Reasonable - except that at the very least that would void the warranty on the expensive original piece of equipment.

opamp buffer followed by a divider?

Thanks for your reply. Yes I'm thinking this is probably the way to go. I was trying to avoid yet another PS as I guess I'll need a 7Vdc one for the buffer as well as the 5Vdc one for the ADC etc. But I guess you can't have everything.

 

For the PS: yes, two supplies will probably be necessary. But, you could do a two-stage PS: the first stage supplies 8-10V which drives your opamp. Then you can use a cheap 7805 linear regulator to generate the 5V from the 8-10V supply.

 

Do you have any idea what the values of the voltage divider in the equpment is. There is probably no reason why you couldn't voltage divide the O/P again.

 

For the PS: yes, two supplies will probably be necessary. But, you could do a two-stage PS: the first stage supplies 8-10V which drives your opamp. Then you can use a cheap 7805 linear regulator to generate the 5V from the 8-10V supply.

At the moment the "main" PS in my kit is 25Vdc. I use a 5V switching reg (RS 532-8050) to get my 5V for the TTL circuitry (MCU, ADC etc) so I guess I can add a 9V version (RS 532-8066) to power the buffer.

Do you have any idea what the values of the voltage divider in the equpment is. There is probably no reason why you couldn't voltage divide the O/P again.

In short no. The only info I have is this from the manual:

The factory-installed 50:1 internal voltage divider provides a maximum of
7 V under open circuit conditions. This is an impedance-protected
functional extra low voltage (ELV) output to prevent shock, energy,
and fire under normal conditions at the machine interface
receptacle and under single fault conditions with the machine
interface wiring. The voltage divider is not fault tolerant and ELV
outputs do not comply with safety extra low voltage (SELV)
requirements for direct connection to computer products.

 

You should be able to run the opamp off of your 25V supply. There are a large number of opamps that are perfectly happy with 32V single supply. LM324 is not a bad device to start with.

 

Sounding good. Thanks for the help - I'm not really an electronics person - but I can kind of muddle through - once I get it into the ADC and I can get to the software bit I'm in my zone

 

With regards to the statement in the manual about not being SELV and not safe to conect to a computer I would also consider an optocouple somewhere as well.

Just a thought

 

You should be able to run the opamp off of your 25V supply. There are a large number of opamps that are perfectly happy with 32V single supply. LM324 is not a bad device to start with.

Had a second thought on this: The 25Vdc is unregulated so I'm thinking the switching reg may be an idea anyway.

With regards to the statement in the manual about not being SELV and not safe to conect to a computer I would also consider an optocouple somewhere as well.

Just a thought

Reasonable. The chain will be: external kit -> opamp buffer -> voltage divider (7V..5V) -> ADC (on board an Atmega) -> digital out to computer parallel port.

I guess the opamp will really do all the protection I need particularly if I clamp the output between 0 and 5 but I could do optos between the atmega and the parallel port I suppose.

Really appreciate you guys help on this - thanks.

 

Hi Stirling,

You could try using the unregulated 25V. How much unregulated is it? What is the source? If it comes from a full bridge rectifier with a reasonable ripple, then you could try this trick:

Instead of connecting the 25V directly to V+ of the opamp, pass it through a RC low pass filter first. Say 1 to 5 Ohm resistor and as big a cap as you can find (100uF or larger).

Opamps don't use much current and also they have a fairly good Power Supply Rejection Ratio (PSRR) which means ripple on the power does not get through to the output. Because of this, a simple RC circuit may be suitable for the opamp.

Optos: Yes, use this between the ATmega and the parallel port. Maybe someone else can suggest a circuit idea that you can use - I'm not that experienced with opto's in this application.

I would also recommend that you build the entire circuit inside a housing which is either strong plastic, or metal connected to earth. Just another level of safety.

 

The PS is a toroidal -> FWBR -> 11,000 uF cap giving 10A @ 25Vdc. Primary purpose is to drive stepper motors but I already tap off through a switched DCDC converter to give me 5V for my logic (stepper drives and atmega etc.) which all works fine. Just need to add in this extra bit for the 0..7 to 0..5 and I'm done really.