esp_now latency (AEGHB-603)

[programmeddeath1]

I am sending multiple values to a slave esp using esp_now broadcast. I am sending points at a frequency of 1 every 30ms and above. For the first few points the slave receives the message within 5 to 7ms. but for certain subsequent points it goes upto to 50ms and even 100ms

esp32 speedtests.xlsx

Point	SentTime	ReceivedTime	ExecutedTime	SlaveReceivedTime	SlaveExecutedTime	CommunicationTime	TotalExecutionTime	SlaveCommunicationTime	SlaveTotalExecutionTime
5	1711798622.65379	1711798622.66174	1711798622.66286	1711798622.67827	1711798622.67946	7.95	9.07	15.41	16.6
6	1711798622.69424	1711798622.70244	1711798622.70363	1711798622.72855	1711798622.7297	8.2	9.39	24.93	26.07
7	1711798622.73481	1711798622.74294	1711798622.74411	1711798622.78723	1711798622.78854	8.13	9.3	43.13	44.43
8	1711798622.77513	1711798622.78358	1711798622.78472	1711798622.83769	1711798622.83887	8.46	9.59	52.97	54.16
9	1711798622.81556	1711798622.82401	1711798622.82515	1711798622.88931	1711798622.8906	8.45	9.59	64.15	65.45
10	1711798622.85593	1711798622.86431	1711798622.86551	1711798622.94036	1711798622.94153	8.38	9.57	74.86	76.03
11	1711798622.89636	1711798622.90404	1711798622.90528	1711798622.99113	1711798622.99247	7.69	8.92	85.86	87.19
12	1711798622.9367	1711798622.945	1711798622.94616	1711798623.04216	1711798623.04331	8.31	9.46	96.0	97.16
13	1711798622.97704	1711798622.98467	1711798622.98596	1711798623.09191	1711798623.09301	7.63	8.92	105.96	107.06
14	1711798623.01738	1711798623.02585	1711798623.02706	1711798623.14438	1711798623.14561	8.47	9.68	117.32	118.55
15	1711798623.05784	1711798623.06586	1711798623.06703	1711798623.19467	1711798623.1958	8.02	9.19	127.64	128.77

Above is a sample of the values The max latency is gong upto 200ms and avg of 50 ms over 37 points sent. The data packets are trajectory points that need to be input to motors Count : 32 POINT: positions: [0.1438207544378186, 0.1438207544378186, 0.22379780136537714] velocities: [0.25470507451224855, 0.25470507451224855, 0.2014506547443208] accelerations: [-1.907522980622658, -1.907522980622658, -0.8210376281913943] effort: [1.0] time_from_start: secs: 2 nsecs: 775355899 Their size is less than 100bytes per single message. I am broadcasting them through - esp_err_t result = esp_now_send(0, (uint8_t *)point, sizeof(trajPoint)); where trajpoint is the message - typedef struct trajPoint { char mastercmd[4]; float pos[3]; float velo[3]; float accel[3]; float count; } single_point;

What could be causing this latency in between? Is there a threshold for difference between intermediate points?

EDIT2 I figured out another thing, if the time difference between successive points is less than 20ms, thats when this delay/buffer overflow starts occuring. Whenever time difference between two points is more than 20ms, it sends each point at 5 to 6 ms range -

points sent @0.03 secs 

Point	SentTime	ReceivedTime	ExecutedTime	SlaveReceivedTime	SlaveExecutedTime	CommunicationTime	TotalExecutionTime	SlaveCommunicationTime	SlaveTotalExecutionTime
0	1711955168.43399	1711955168.44746	1711955168.44866	1711955168.4542	1711955168.4553	13.47	14.67	5.53	6.64
1	1711955168.46448	1711955168.47901	1711955168.48011	1711955168.48454	1711955168.48568	14.52	15.62	4.43	5.57
2	1711955168.49504	1711955168.50925	1711955168.51047	1711955168.51594	1711955168.51707	14.21	15.43	5.46	6.59
3	1711955168.52538	1711955168.53109	1711955168.53229	1711955168.54092	1711955168.542	5.71	6.91	8.63	9.71
4	1711955168.55576	1711955168.56162	1711955168.56282	1711955168.57009	1711955168.57123	5.86	7.05	7.28	8.41

 points sent @0.02 secs -

Point	SentTime	ReceivedTime	ExecutedTime	SlaveReceivedTime	SlaveExecutedTime	CommunicationTime	TotalExecutionTime	SlaveCommunicationTime	SlaveTotalExecutionTime
0	1711955028.20987	1711955028.22388	1711955028.22511	1711955028.23112	1711955028.23227	14	15.24	6	7.16
1	1711955028.23026	1711955028.23497	1711955028.23617	1711955028.25614	1711955028.25728	4.7	5.9	19.97	21.11
2	1711955028.2507	1711955028.2558	1711955028.25693	1711955028.2778	1711955028.27893	5.1	6.23	20.87	22
3	1711955028.27102	1711955028.27754	1711955028.27866	1711955028.30291	1711955028.30414	6.52	7.64	24.26	25.48
4	1711955028.29132	1711955028.29831	1711955028.29945	1711955028.32478	1711955028.32589	6.98	8.12	25.33	26.44
5	1711955028.31172	1711955028.31905	1711955028.32019	1711955028.34978	1711955028.35095	7.33	8.48	29.59	30.76

[lhespress]

@programmeddeath1 Please see the test report as the attachment. ESP-NOW Test Report.pdf

[nanshenwei]

Here is my setup. Hope this helps. https://github.com/nanshenwei/espnowSpeedTest.git

[lhespress]

@programmeddeath1 Closing this issue since there has been no update on this. Please feel free to reopen if required.