Analog signal linearity
- Methier
- Offline
- Junior Member
- 
				  
		Less
		More
		
			
	
		- Posts: 22
- Thank you received: 0
			
	
						17 Dec 2017 18:06		 -  17 Dec 2017 18:13		#103228
		by Methier
	
	
		
			
	
	
			 		
													
	
				Analog signal linearity was created by Methier			
			
				Hi, this is my first post
I'm working with mesa 7i77+5i25, i have everything connected and working, I'm in the tunning part,
i decided to test if the analog signal was at the correct scale, by opening a opened a halscope, and tested the commanded speed and encoder feedback, all the PID gains to zero just because i wanted to see the output without any PID correction, then i noticed that at low speeds i get a matching speed signal between commanded speed and encoder feedback, but when i increased the signal to get higher speed i got a noticeable difference.
Based on that i measured the voltage at the output,
Commanded - Measured
Voltage(VDC) - Voltage(VDC)
10 - - - - - - - - - - - - - 9.55
9 - - - - - - - - - - - - - - 8.5
8 - - - - - - - - - - - - - - 7.6
7 - - - - - - - - - - - - - - 6.7
6 - - - - - - - - - - - - - - 5.9
5 - - - - - - - - - - - - - - 5.2
4 - - - - - - - - - - - - - - 4.3
3 - - - - - - - - - - - - - - 3.00
2 - - - - - - - - - - - - - - 2.02
1 - - - - - - - - - - - - - - 1.02
0 - - - - - - - - - - - - - - 0.05
Measurements where made with a 24VDC power supply connected to field voltage input, i checked if there was some noise with a oscilloscope and everything was ok, i made the tests inside my house so i don't think that it was external noise from the shop, ground connection is ok.
i know that is not a big deal, perhaps PID can compensate that difference, but i think it will improve the performance of the system.
Is there a function to linearize analog output?
i can program my own, that's why i like linuxcnc but if it is already made i can save some time and go to the next step and see those chips flying around.
Best Regards
					I'm working with mesa 7i77+5i25, i have everything connected and working, I'm in the tunning part,
i decided to test if the analog signal was at the correct scale, by opening a opened a halscope, and tested the commanded speed and encoder feedback, all the PID gains to zero just because i wanted to see the output without any PID correction, then i noticed that at low speeds i get a matching speed signal between commanded speed and encoder feedback, but when i increased the signal to get higher speed i got a noticeable difference.
Based on that i measured the voltage at the output,
Commanded - Measured
Voltage(VDC) - Voltage(VDC)
10 - - - - - - - - - - - - - 9.55
9 - - - - - - - - - - - - - - 8.5
8 - - - - - - - - - - - - - - 7.6
7 - - - - - - - - - - - - - - 6.7
6 - - - - - - - - - - - - - - 5.9
5 - - - - - - - - - - - - - - 5.2
4 - - - - - - - - - - - - - - 4.3
3 - - - - - - - - - - - - - - 3.00
2 - - - - - - - - - - - - - - 2.02
1 - - - - - - - - - - - - - - 1.02
0 - - - - - - - - - - - - - - 0.05
Measurements where made with a 24VDC power supply connected to field voltage input, i checked if there was some noise with a oscilloscope and everything was ok, i made the tests inside my house so i don't think that it was external noise from the shop, ground connection is ok.
i know that is not a big deal, perhaps PID can compensate that difference, but i think it will improve the performance of the system.
Is there a function to linearize analog output?
i can program my own, that's why i like linuxcnc but if it is already made i can save some time and go to the next step and see those chips flying around.
Best Regards
		Last edit: 17 Dec 2017 18:13  by Methier.		Reason: never used a table	
			Please Log in or Create an account to join the conversation.
- PCW
- 
				  
- Away
- Moderator
- 
				  
		Less
		More
		
			
	
		- Posts: 17366
- Thank you received: 5062
			
	
						17 Dec 2017 18:11		 -  17 Dec 2017 18:14		#103231
		by PCW
	
	
		
			
	
	
			 		
													
	
				Replied by PCW on topic Analog signal linearity			
			
				7I77 analog linearity should be better than .1 % and .05V offset at 0 is way too high
How are you measuring the output?
Also check the 7I77s 5V source. If this is low, you can have analog output issues
					How are you measuring the output?
Also check the 7I77s 5V source. If this is low, you can have analog output issues
		Last edit: 17 Dec 2017 18:14  by PCW.			
			Please Log in or Create an account to join the conversation.
- Methier
- Offline
- Junior Member
- 
				  
		Less
		More
		
			
	
		- Posts: 22
- Thank you received: 0
			
	
						17 Dec 2017 18:15				#103232
		by Methier
	
	
		
			
	
			
			 		
													
	
				Replied by Methier on topic Analog signal linearity			
			
				negative lead to GND (from the same terminal block ofc)
positive to AOUT0
i checked with multimeter and oscilloscope, both gave me same readings.
Regards
					positive to AOUT0
i checked with multimeter and oscilloscope, both gave me same readings.
Regards
Please Log in or Create an account to join the conversation.
- PCW
- 
				  
- Away
- Moderator
- 
				  
		Less
		More
		
			
	
		- Posts: 17366
- Thank you received: 5062
			
	
						17 Dec 2017 18:21				#103233
		by PCW
	
	
		
			
	
			
			 		
													
	
				Replied by PCW on topic Analog signal linearity			
			
				I would check the 5V, If that is OK and you are sure about you multimeter accuracy,
I would get an RMA for the card as it clearly broken.
					I would get an RMA for the card as it clearly broken.
Please Log in or Create an account to join the conversation.
- andypugh
- 
				  
- Offline
- Moderator
- 
				  
		Less
		More
		
			
	
		- Posts: 19676
- Thank you received: 4552
			
	
						18 Dec 2017 12:29				#103283
		by andypugh
	
	
		
			
				
My multimeter gets inaccurate when the battery is low. (eventually it over-reads b a factor of two!)
Can you test with two multimeters and see if they agree?
The "lincurve" HAL component could correct for this, but maybe not reliably, because PCW says that something is wrong, so the error might not remain constant over time.
					
	
			
			 		
													
	
				Replied by andypugh on topic Analog signal linearity			
			are sure about you multimeter accuracy.
My multimeter gets inaccurate when the battery is low. (eventually it over-reads b a factor of two!)
Can you test with two multimeters and see if they agree?
The "lincurve" HAL component could correct for this, but maybe not reliably, because PCW says that something is wrong, so the error might not remain constant over time.
Please Log in or Create an account to join the conversation.
- Methier
- Offline
- Junior Member
- 
				  
		Less
		More
		
			
	
		- Posts: 22
- Thank you received: 0
			
	
						18 Dec 2017 16:15				#103295
		by Methier
	
	
		
			
	
			
			 		
													
	
				Replied by Methier on topic Analog signal linearity			
			
				Hi, thanks for your reply
I'm waiting for another multimeter from a friend and I was using the 5v power supply from 5i25 board (readings gave me 4.85v), I bought an independient adjustable 5v power supply to see if there are any improvement.
I will do tests today at noon.
Best Regards
					I'm waiting for another multimeter from a friend and I was using the 5v power supply from 5i25 board (readings gave me 4.85v), I bought an independient adjustable 5v power supply to see if there are any improvement.
I will do tests today at noon.
Best Regards
Please Log in or Create an account to join the conversation.
- Methier
- Offline
- Junior Member
- 
				  
		Less
		More
		
			
	
		- Posts: 22
- Thank you received: 0
			
	
						15 Apr 2018 02:13				#109039
		by Methier
	
	
		
			
	
			
			 		
													
	
				Replied by Methier on topic Analog signal linearity			
			
				Hi, sry for the delay with my reply, the independent 5v psu made a good improvement with the analog out, perhaps because the PC is an old one the PSU is about to die, i did some tests with good results so i will keep the external one.
The problem now are the servo motors, quite old, and no maintenance from the last owner damaged im some way the feedback devices.
Anyway thanks for your help.
					The problem now are the servo motors, quite old, and no maintenance from the last owner damaged im some way the feedback devices.
Anyway thanks for your help.
Please Log in or Create an account to join the conversation.
		Moderators: PCW, jmelson	
		Time to create page: 0.110 seconds	
