EGIS motor

[Channel Hopper]

the motor run to half speed !!!!
pulse is 2.5 ms long instead of 1.2ms .
i put a at89lp51 in place of 89s52 an now i manage pulse from encoder at real speed (pulse every 1.2ms).
i have adapted firmware to show 5 digits number (1 to 65535) .
on AZ turn i receive about 50000 pulses (for 180°) on EL movement it is 30000 pulses!!!!!!

I was hoping the overrun might be somewhat less than half of the original 127 pulses at 18V.
If so you could also try at 9V and 27V to work out the best slow start/stop voltage from the power supply.

The above is assuming you can tap off additional secondary points from the transformer.

 

[dreambox1959]

I'm not sure we understand each other!
I have a switching power supply followed by a PWM mosfet stage, I can modulate the voltage and the average value on a DC motor as if the voltage was lowered.

 

[dreambox1959]

I observe a defect that FEMI had described to me
on the EGIS motor which works with a big screw, there is a small difference in position depending on whether you come from the left or the right : it is the mechanical play of the gear.
a trick could be to pass the destination point and then come back to the same side but if it is windy it will not work !!!!
in a perfect system there should be no mechanical play !!!!



maybe less but my pen is large

 

[Channel Hopper]

Yes , mechanical play is unacceptable unless it is within the tolerances of the original design.
The over run however needs to be corrected as a priority since you want accuracy in your own modifications.

 

[Channel Hopper]

I'm not sure we understand each other!
I have a switching power supply followed by a PWM mosfet stage, I can modulate the voltage and the average value on a DC motor as if the voltage was lowered.

In that case you could plot a graph that calculates the overrun vs angular velocity.

 

[dreambox1959]

I was inspired by the documentation of the controller egis 103.
I added correction parameters for the movements of the four directions.
I stop the motor supply a few pulses before the arrival to stop at the right index.
These numbers are configurable on the screen.
It works !

for the mechanical play, TRUST will see this problem, this motor has already been dismantled several times.

 

[Channel Hopper]

I was inspired by the documentation of the controller egis 103.
I added correction parameters for the movements of the four directions.
I stop the motor supply a few pulses before the arrival to stop at the right index.
These numbers are configurable on the screen.
It works !

It won't work for all positions though with just one setting (for example single satellite hops where there is just a few pulses).

As an example, what is the current timing and pulse count to jump between 14W and 15W ?

 

[Trust]

Play is a common issue on Egis and a problem when you have a dish with a small openings angle
JP do you noticed some AZ play in the Egis ?
Because I made some modifications on that motor and other ones that been in my hands by making the main worm bearing adjustable so it’s more tight to worm wheel

 

[dreambox1959]

the mechanical play corresponds to my photos, if I arrive at the desired position from the left after a previous movement to the right on my model with the pencil, I have a 3mm gap at the end of my arm, I can push the turret by hand and you can feel this little game!
I did not remove the main worm but there is this little play.

Channel Hopper​

in use, my program causes the motor speed to decrease as the dish approaches the final destination so we always arrive at reduced speed and the inertia is identical at each position.
anyway the final position is programmed by the user !!
my program makes sure to arrive at the setpoint defined by the user.

 

[dreambox1959]

I will return the EGIS engine to TRUST tomorrow.
I have worked 2 weeks on adapting my positioner to the EGIS, I have achieved a satisfactory operation!
The main difficulty was that the motors of the egis provide a lot of pulses.
36000 pulses for 180° I said to myself that 100 pulses for 1° are sufficient so by counting only the high states of the pulse I have a correspondence between the strong weight and the degrees of the mount:
150000 = 150° , 9854 = 98°54 .
I added a function to compensate for the inertia of the motors and stop at the right place.
There is just a small card to be added on the motor to adapt the level from 24V to 5V.
Those who are interested will contact me .

 

[Channel Hopper]

the mechanical play corresponds to my photos, if I arrive at the desired position from the left after a previous movement to the right on my model with the pencil, I have a 3mm gap at the end of my arm, I can push the turret by hand and you can feel this little game!
I did not remove the main worm but there is this little play.

Channel Hopper​

in use, my program causes the motor speed to decrease as the dish approaches the final destination so we always arrive at reduced speed and the inertia is identical at each position.
anyway the final position is programmed by the user !!
my program makes sure to arrive at the setpoint defined by the user.

Then you have partly overcome the mechanical flaws with a non mechanical solution.

An active tracking piece of software to ensure the dish stays centred on the satellite next ?

 

[dreambox1959]

Yes it s also the case in original egis box !
see post #29
on egis i have 100 pulses per degree , on a classic actuator it s about 17 pulses /degree

 

[Hugocz]

Greetings to all egis fans. I'm back to work and hope to post news soon. I redesigned my website in the new Joomla 4 editorial system. That's why some links may no longer work. But the base address does not change. News You can probably find everything you need from here.

 

[dreambox1959]

hello hugo
I am in the process of adapting my positioner (initially designed for classic motors or HH motors) for the EGIS.
With the help of TRUST (which lent me a 15-day EGIS), I adapted my firmware (which is written in MCS51 assembler).
I achieved a very good result, I was surprised by the number of pulses, it's very surprising especially when it's managed by an old 8085 type microprocessor!!
I use an AT89LP51RD (12 times faster than an 8751).
I already have a reliable software base and I am ready to exchange information with you.

see below a nice idea !

 

[Hugocz]

Hi.
I don't like placing the EGIS diagonally at all. EGIS was not designed to operate in such a position. I assume an excessive load on some bearings. Aren't you worried about the lifespan of EGIS? What weight of parabola do you have attached?

I agree that stopping Egis exactly at a set position is a problem.
My relay control will overshoot the correct spot by about 50 pulses. This is already known for weak satellites. :-)
I expect a lot from the speed control of the motors. Here I will see the result only after the prototype is put into operation.
At the request of a few fans, I started researching the software in the original EPS-103 unit (8085 assembler).
I don't expect much from it.
But maybe I'll see how professionals from the Egis company stop the engines exactly at the position.
Maybe the OKI M60233 circuit (inputs from sensors?) is important.
Now I would mainly like to finish my prototype. But I'm definitely interested in your work. :-)
Hugocz

 

[dreambox1959]

this is not my EGIS montage, it is a photo found on the net.
for the stop in the right place I solved this problem by stopping the power supply a little before the desired point, it is in the egis documentation (post #29) it works very well but the values are to be determined on site.
the pwm is not very proportional: when I go from 50% to 70% the engine is already at full throttle.
I opened a motor (post23), there are chokes in the brush holder, they partially cancel the PWM effect...

 

[Hugocz]

I've also seen EGIS positioned like this on the net. I thought it was an inspiration for you. Again, I don't think it's a good idea.

Regulating the motor rotation speed is only part of the solution. The second part is program support. Setting a constant that stops the motor before the desired position is the easiest way. But the accuracy for the elevation depends on the weight of the parabola and its current inclination.
I'm thinking of slowing down in steps. According to my experiments, the dependence of speed on PWM is approximately logarithmic. But I still have no idea how complex a program I will be able to design.
Functional hardware is a priority.

Chokes in the PWM signal path can be a problem. But it depends on the frequency. I believe that in the range of 50HZ to perhaps 500 Hz, the correct value will be found, at which the influence of the chokes will be minimal.

 

[dreambox1959]

for me i choose pwm at 5khz (cycle is 200µs)
and it slow down with a calculation of pulses that remains from final goal .
the more dish approach the less speed is high .

 

[Trust]

This is how I travelled with the Egis Rotor on my way to South France .
No room inside the car by all the dresses and shoes from my lovely wife .
Delivered to a friend of JP in Valence and yesterday picked it up from JP at a spot near Macon .
Had a (too)short chat with JP , nice to meet him .

 

[Hugocz]

Interesting thing.
During yesterday, many interested people downloaded the original documentation for the EGIS positioner from my website. I don't know if this is related, but my website is down today. It seems that hackers have become interested in them. I just want to let you know that my website is currently down. I am aware of this and will restore from backup. But there is a possibility that the situation will repeat itself. We'll see.