Evaluation board with STM32MP157 MPU

This thermal analysis has been done on an evaluation board with an STM32MP157 device, without casing (see the figure below).

Figure 15. STM32MP157x-EV1. This picture is not contractual.

While running commands, the script monitors the following:

The CPU load with mpstat
The board temperature via / sys / class / hwmon / hwmon0 / temp1_input
The memory used via the free command

STM32MP157 power dissipation

As stated in STM32 thermal parameters, the dissipated power varies with T_J. The power consumption in this example is measured for different T_J values (see the table and the figure below).

Table 7. STM32MP157 power dissipation versus T_J
T_J (°C)	STM32MP157x-EV1 board (supplied with 5 V)		STM32MP157 power dissipation (mW)
T_J (°C)	Current consumption (A)	Power consumption (mW)	STM32MP157 power dissipation (mW)
39	0.4	2000	970.2
46	0.4	2000	973.8
65	0.405	2025	1006.4
81	0.41	2050	1053.1
95	0.42	2100	1104.7
106	0.432	2160	1195.5
116	0.445	2225	1280.6
125	0.456	2280	1366.6

Figure 16. Junction-temperature-dependent dissipated power for STM32MP157

STM32MP157 thermal measurements at T_A = 25 °C

The figures and table below detail the thermal measurements performed at T_A = 25 °C.

Table 8. STM32MP157 thermal measurements at T_A = 25 °C
T_J (°C)	Thermal camera measurement (°C)			STM32MP157 power consumption (mW)	Total power consumption (mW)
T_J (°C)	Avg	Min	Max	STM32MP157 power consumption (mW)	Total power consumption (mW)
46	46.2	40.52	51.4	974	2000

Figure 18. HDC at 25 °C and junction-temperature-dependent dissipated power for STM32MP157

Once the HDC of the design is defined (by measurement at 25 °C), and assuming that this HDC is constant at 25 °C and 85 °C, the operating point of the design at 85 °C is given by translating the HDC at 85 °C as shown in the figure and table below.

Figure 19. HDC at 85 °C and junction-temperature-dependent dissipated power for STM32MP157

Table 9. Measurement interpolation at T_A = 85 °C for STM32MP157
Parameter	Value
T_J	110 °C
P_TOT	2200 mW

In conclusion, the STM32MP157 device is safe, with no thermal runaway, and T_J remaining < 125 °C.

Internal validation board with STM32MP135 MPU

This thermal analysis has been done on an internal validation board with an STM32MP135 device without casing.

This analysis is based on the use cases detailed in the table below. They correspond to CPU frequencies of 900 MHz and 1 GHz:

Table 10. Use cases for STM32MP135 MPU
Name	Description	CPU load	DDR load
cpuburn	-	100%	R = 3 Mbyte/s, W = 0 Mbyte/s, R&W = 4 Mbyte/s
dd	Idle (weston launcher, netdata) + dual eth + USB hard drive + dd on hard drive (ETH cables unplugged)	100%	R = 129 Mbyte/s, W = 40 Mbyte/s, R&W = 169 Mbyte/s

STM32MP135 power dissipation

Use case about cpuburn

As stated in STM32 thermal parameters, the dissipated power varies with T_J. The power consumption in this example is measured for different T_J values (see the following tables and figure).

Table 11. STM32MP135 power dissipation versus T_J at 900 MHz (cpuburn)
T_J (°C)	STM32MP135 power dissipation (mW) at 900 MHz
20.00	711.44
31.00	721.04
43.00	740.00
47.00	748.83
52.00	757.93
56.00	763.62
66.00	791.42
71.00	804.79
78.00	823.36
86.00	857.92
91.00	881.62
96.00	903.01
98.00	914.23
101.00	926.44
106.00	951.83
111.00	973.00
116.00	994.98
119.00	1014.19
129.00	1110.12
139.00	1238.89

Table 12. STM32MP135 power dissipation versus T_J at 1 GHz (cpuburn)
T_J (°C)	STM32MP135 power dissipation (mW) at 1 GHz
21.60	795.45
32.00	800.12
37.00	806.26
44.00	825.79
49.00	833.00
56.00	841.39
61.00	861.53
68.00	890.82
76.00	912.60
83.00	936.60
89.00	967.97
100.00	1014.80
110.00	1061.20

Table 13. STM32MP135 thermal measurements at T_A = 25 °C (cpuburn)
Parameter	900 MHz	1 GHz	Unit
T_J	47	49	[°C]
V_CORE	1.32	1.332	[V]
I_CORE	180	195	[mA]
P_CORE	237	260	[mW]
V_DD	3.33	3.321	[V]
I_DD	13	19	[mA]
P_DD	43	63	[mW]
V_CPU	1.415	1.414	[V]
I_CPU	306	331	[mA]
P_CPU	432	468	[mW]
V_DDR2	1.2	1.206	[V]
I_DDR2	30	35.5	[mA]
P_DDR2	36	43	[mW]
STM32MP135 power dissipation	748	834	[mW]
STM32MP135 power dissipation	+11%		[mW]

Figure 20. Junction temperature-dependent dissipated power for STM32MP135 (cpuburn)

Use case about dd

As stated in STM32 thermal parameters, the dissipated power varies with the T_J. The power consumption in this example is measured for different T_J values (see the following tables and figure).

Table 14. STM32MP135 power dissipation versus T_J at 900 MHz (dd)
T_J (°C)	STM32MP135 power dissipation (mW) at 900 MHz
20	586.44
31	596.04
43	615.00
47	623.83
52	632.93
56	638.62
66	666.42
71	679.79
78	698.36
86	732.92
91	756.62
96	778.01
98	789.23
101	801.44
106	826.83
111	848.00
116	869.98
119	889.19
129	985.12

Table 15. STM32MP135 power dissipation versus T_J at 1 GHz (dd)
T_J (°C)	STM32MP135 power dissipation (mW) at 1 GHz
21.6	670.45
32	675.12
37	681.26
46	700
49	708.00
56	716.39
61	736.53
68	765.82
76	787.60
83	811.60
89	842.97
100	889.80
110	936.20

Table 16. STM32MP135 thermal measurements at T_A = 25 °C (dd)
Parameter	900 MHz	1 GHz	Unit
T_J	43	46	[°C]
V_CORE	1.32	1.32	[V]
I_CORE	177	198	[mA]
P_CORE	233	261	[mW]
V_DD	3.33	3.33	[V]
I_DD	13	19	[mA]
P_DD	43	63	[mW]
V_CPU	1.415	1.398	[V]
I_CPU	173	199	[mA]
P_CPU	244	278	[mW]
V_DDR2	1.268	1.268	[V]
I_DDR2	75	77	[mA]
P_DDR2	95	97	[mW]
STM32MP135	615	700	[mW]
STM32MP135	+ 13%		[mW]

Figure 21. Junction temperature-dependent dissipated power for STM32MP135 (dd)

HDC at 25 °C and junction temperature

Once the design HDC is defined (by measurement at 25 °C), and assuming that this HDC is constant at 25 °C and 85 °C, the operating point of the design at 85 °C is given by translating the HDC at 85 °C as shown in the figure below.

Maximum ambient temperature (STM32MP135 MPU)

The maximum ambient temperature for the cpuburn and dd use cases are given in the figures below.

Figure 23. Maximum ambient temperature(Cpuburn)

Figure 24. Maximum ambient temperature (dd)

Conclusion

The measurement details given in the previous sections give:

An 11% and 13% increase of the power consumption between 900 MHz and 1 GHz. These are respectively for the cpuburn and dd use cases.
A 2 and 3 °C increase of T_J between 900 MHz and 1 GHz (for cpuburn and dd use case respectively)
No risk of thermal runaway

Table 17 Maximum ambient temperature (STM32MP135 MPU) summarizes the maximum ambient temperature for various configurations.

Table 17. Maximum ambient temperature (STM32MP135 MPU)
Frequency	Maximum ambient temperature (°C)
	T_J = 105 °C		T_J = 125 °C
	cpuburn	dd	cpuburn	dd
900 MHz	77.5	81	93.5	97.5
1 GHz	74	78	91	94

Evaluation board with STM32MP157 MPU