You could certainly design an EVA suit that has a low enough thermal conductivity (or effective thermal conductivity) such that it has a loss equal to the typical thermal output from a human body.
We can estimate heat output from the human body based on the typical (suggested) calorific intake.
$$2,000\text{ cal} = 8.37\times10^6\text{ J}$$
If we assume all this is turned into heat energy, then we can design a suit to match this energy input. According to Wikipedia the average surface area of the human body is $1.73\ m^2$.
So we would want to design a suit with a thermal flux of $4.84\times10^6\text{ J}/m^2/\text{day}$.
Joules/day is kind of a messy unit system. We can tidy that up and turn into joules / second, or watts.
$$4.84\times10^6\text{ J}/m^2/\text{day} = 4.84\times10^6\times(1/(60\times60\times24))\mathrm{\ W}/m^2$$
That leaves us with a value of almost exactly $56\mathrm{\ W}/m^2$.
So to design for the worst case scenario (coldest surroundings) you have both conduction from the wall to the atmosphere, and radiation.
The convection from the outside wall of the suit is (derived from Fourier's law):
$$\mathrm{Q} = κA(T_{\mathrm{hot}}-T_{\mathrm{cold}})/d$$
Where:
- $\mathrm{Q}$ = rate of heat transfer (in watts)
- $κ$ = thermal conductivity of the wall (in watts per meter kelvin)
- $A$ = heat transfer surface area (in square meters)
- $T_{\mathrm{hot}}$ = hot temperature (in kelvins)
- $T_{\mathrm{cold}}$ = cold temperature (in kelvins)
- $d$ = distance to ambient temperature (in meters)
The radiation can be calculated with the following equation (derived from the Stefan-Boltzmann law):
$$q = \sigma \times A \times \Delta T^4$$
Where:
- $q$ = heat transfer per unit time (in watts)
- $\sigma$ = Stefan–Boltzmann constant ($5.670373 \times 10^{-8}\, \mathrm{J\, s^{-1}m^{-2}K^{-4}}$)
- $A$ = total surface area of the emitting body (in square meters)
- $\Delta T$ = $T_{\mathrm{hot}}-T_{\mathrm{cold}}$ (in kelvins)
Depending on the design of your suit the internal heat transfer is going to happen in one of two ways, either conduction though material or radiation, if the design is MLI based.
So is it possible to build an EVA suit that doesn't require an external heat source? Yes. It's a question of the mass of the resulting suit and whether the human inside could move around in the potentially heavy suit.