# Wheatstone bridge: Modifications for different temperature sensor resistance

I am trying to build a Wheatstone circuit to detect temperature changes of a low resistance (about 10 ohm) wire.

For a first try I would like to use a circuit proposed in this Maxim Application Note 6144 How to Linearize Wheatstone-Bridge Circuit for Better Performance, Figure 7:

I am a beginner in using and understanding schematics. So in general I have two questions for the beginning:

1. As stated, the circuit in figure 7 is an example to be used with a PT1000. I guess that's why R1 and R3 were chosen to be 1k Ohm? What needs to be changed if R3 is only 10 ohm for example?

2. In the example, some capacitors don't have values. Some are names C_FLY, C_FELT, C_HOLD or don't have a name. Can you guys explain to me why they are named like that and what values would make sense in this circuit?

• You don't need a wheatestone bridge to detect resistance. You can just use a voltage divider circuit. Oct 26, 2020 at 12:27
• I don't want to detect the resistance directly, but get a voltage output depending on the resistance change due to the temperature change. This will be read by a 16 bit ADC. My google research showed that a wheatstone bridge does exactly that, but needs to be linearized and amplified. Thats why I chose to start with the linked scheme. Oct 26, 2020 at 12:41
• @Se1fie but for precise temperature measurements, the Wheatstone bridge is the right choice. You'll have a hard time making a high-resolution measurement on a linear scale with just a voltage divider! The Wheatstone Bridge was invented to do exactly what Kj Ell wants: high-accuracy measurements of resistances by balancing two voltage dividers. I recommend you read the wikipedia article on Wheatstone bridges – it's quite insightful, Se1fie. Oct 26, 2020 at 12:41
• for those capacitors, you'd best look at the 44267 data sheet, they are intimately related to its operation. From its single supply rail, and the fact that it's handling voltages above and below ground, I would guess it has an internal charge pump which uses those components externally, because sizeable capcitors don't integrate with ICs well. Cfly sounds like the 'flying capacitor' of the charge pump. See if you can guess what Chold does. I can't guess what Cfelt does! Oct 26, 2020 at 12:53
• Marcus Muller I know what a Wheatstone bridge is just I don't use it unless it is homework. Oct 26, 2020 at 13:00

As stated, the circuit in figure 7 is an example to be used with a PT1000. I guess thats why R1 and R3 chosen to be 1k Ohm?

Exactly!

And what needs to be changed if R3 is only 10 Ohm for example?

In a wheatstone bridge, you want the resistors on opposite sites to have the same product. So, in this circuit,

$$R_1 \cdot R_4 \overset!= R_2 \cdot R_3.$$

So, if R3 is 10 Ω, you could keep R2 and R4 as 10 kΩ, and set R1 to 10 Ω.

However, it's usually a good idea to keep the relative orders of magnitudes, so, R3=R1=10 Ω, R2=R4=100 Ω would theoretically work better.

You might be getting into current sourcing ability problems of your 2.5V source and your opamps with a 100Ω load only, but:
I've never seen a 10 Ω temperature resistor (not a good idea to make the impedance of sensor so low), so I presume this remains a theoretical problem.

In the example some capacitors don't have values. Some are names C_FLY, C_FELT, C_HOLD or don't have a name. Can you guys explain me why they are named like that and what values would make sense in this circuit?

Yes, but it makes no sense for us to explain them: they're explained in the MAX44267's datasheet quite explicitly, and better than we could do it here ad hoc in an answer.

Watch out, though: the labels on the schematic in the Application Note are um, badly done.

C_FELT is connected to pin 6, which is Vss according to the datasheet. So that's the "low-ESR bypass capacitor C_FILT" described in the table at datasheet page 11.

• That helps alot! I actually checked the datasheets of the components but did not find the information. After a second look, you are right the informations are there with exact numbers. I will try to build the schematic as it is above in KiCad. And yeah, the resistance is low. It is the filament of a really small lamp. Oct 26, 2020 at 12:58
• One more question if you don't mind, as I am trying to learn some stuff here as well. :) Can you explain what you mean by: You might be getting into current sourcing ability problems of your 2.5V source and your opamps with a 100Ω load only Oct 26, 2020 at 13:50
• If you look at the schematic, the left opamp is in a negative feedback configuration: it will try to set its output such that its inverting input is at the same potential as its non-inverting input. Its non-inverting input is at GND, and its inverting input is directly connected to "A". Thus, A will be at GND. "V" is at some positive voltage: That means that "V"/10 Ω or something flow through your bulb, at least. That can be quite a lot for a poor measurement opamp or a poor precision voltage source. Oct 26, 2020 at 13:56
• Okay, thank you. I think I got that. First I'll go ahead and give it a try in designing and testing the setup and see what happens. We have a fast temperature sensor I can test it against to see if the voltages are at least in a correct region. Oct 26, 2020 at 14:06
• Also note : a drive voltage that is successfully used with a 1k resistance will cause 100x the dissipation in a 10R resistor : this will increase its temperature and probably make quite a mess of the measurement. Oct 26, 2020 at 15:01