Pairing up Mesa 7i76e with VFD Delta C2000

09 Jan 2019 12:45 #123825 by andypugh
A line driver into a high-impedence oscilloscope might give a higher than specified voltage. You could try loading it with a resistor. It is rated 50mA, so I would first try a 1k resistor (5mA) and see if that squares-up and reduces the peak and noise in the signal.

To figure out the resolution, just watch the counts in halmeter, rotate the spindle 10x by hand, and divide the change in counts by 10.
09 Jan 2019 19:02 #123852 by vmihalca
I did some more research being helped by a friend that's more into electronics.
We have discovered that the board outputs differential outputs, and there is no common ground between the phases.
Also we discovered that the voltage is from -7.5 to +7.5 and not from 0 to 15
We created a voltage divider using two 500ohm resistors, and filtered the noise with a 1nF capacitor. We ended up with a 4.8v peak to peak.
But this voltage is from -2.4 to +2.4 and not from 0-4.8v.
Can I feed this signal into the mesa board? Or do I need a 0-5v signal.
09 Jan 2019 19:47 #123861 by andypugh
The Mesa inputs are selectable TTL or Differential, IIRC.
What you have is differential. (which is a good thing, it is more noise-resistant)


Page 2 W4 W5 W6
09 Jan 2019 20:21 #123867 by vmihalca
So you're saying its ok to feed into the mesa board a signal that' from - to + and not from 0 to +?
I would not like to get my board fried.
I thought I need to supply a signal between 0 and 5V
09 Jan 2019 20:34 - 09 Jan 2019 20:35 #123868 by PCW
7I76E differential encoder inputs have a common mode range of -7 to +12V
and a differential sensitivity of 200 mV, and a maximum differential voltage of 5V,
So the will take positive and negative inputs as long as they are in that range

You must make sure that no input exceeds +13V or -9V or you may damage the input receiver chip. If you have large signals and large noise spikes, you should use your input divider. Note that the 7I76E encoder inputs are terminated with a 120 Ohm resistor across the differential inputs pairs, so you may need to adjust your divider accordingly. As long as you get about 2V across the inputs its good.
Last edit: 09 Jan 2019 20:35 by PCW.
10 Jan 2019 21:52 - 10 Jan 2019 21:54 #123924 by vmihalca
I have made today some measurements and some tests. I have discovered that the encoder board has a SG pin that I assumed it comes from 'signal ground'.
I have connected the ground of the probes to the ground, and then connected the probes to the A and A' with a 110ohm resistor between them to simulate the internal resistor that mesa has.
See the measurements in the pics below.
I am wondering if the noise that appears in the signal can get out of the -7/+12v range and if this could affect the mesa board.
Given how the signal looks now, do you think its safe to connect the wires to the Mesa's encoder input?
Also, do I need to connect the ground from the encoder to the GND encoder pin on the Mesa?
Last edit: 10 Jan 2019 21:54 by vmihalca.
10 Jan 2019 22:30 #123926 by tommylight
I would not connect that. You are using non shielded cable for signalling, shielding is a must in such machines, to many devices making to much electrical noise.
BTW, cable shielding should be connected only on one side, not both. Usually on the electronics side.
10 Jan 2019 23:08 #123931 by vmihalca
Do you think a double shielded FTP cable would do it?
You said about connecting only one side; here the sides I have are the Mesa board and this board that converts from the resolver signal.
To which of them should I connect the shielding? Where should I connect the shielding? To ground?
Sorry if I'm asking stupid questions, I'm a very beginner with the electronics. :)
11 Jan 2019 14:33 #123975 by tommylight
Electronics side.
Encoders or resolvers should be grounded at the Mesa input side ( or other board ), motors should be grounded at the driver side, sensors and switches also at the board side, etc
Time to create page: 0.090 seconds
Powered by Kunena Forum