Amiko DM3800 current overload?

[deeptho]

I have an old amiko dm3800 rotor, carrying a laminas 1.2 M dish.
This mostly worked fine.

Now yesterday I suddenly noticed a very strange problem: frequently the dish did
not want to move and the LNB did not even produce a signal. After a lot of trouble shooting
it became clear that the LNB and the switch behind the positioner did not receive any power,
which leads me to believe that the positioner activates its over-current protection.

I think this over-current activations occurs even when the dish does not receive any commands
to move. The LNB is fine. I think my Amiko has a hardware problem. Did anyone experience something
similar?

Below is the full detective story.... and a work around. :->

First some background: As the dish started to develop some slack, I tightened the slack adjustment a few
months ago. At that time, I also measured the current, which sometimes peaked
between 300 and 350 mA when moving the dish. A bit on the high side, but it worked.

So my first thought was that the motor draws too much current, perhaps due to the hot weather which might
cause extra tightening. This was not the problem although the currents I measured were still a bit on the high side.

Why led me to believe that it is an overload is that this problem only occurs when using a V-polarised mux
(so 18 Volt) and never on an H-mux. Now here is the strange thing: after some further experimentation,
and measuring voltages and currents, it became clear that this problem occurs even when not sending
any dish-move commands. So the motor did not even have to run. Simply tuning to an H-transponder
made the Amiko power off the LNB quite consistently.

I did have a swicth installed between the Amiko and the LNB, but removing it did not make a difference.
So it is strange: the Amiko cannot even handle this normal use case of an LNB directly connected to
the positioner anymore.

Then I had another little breakthrough: recently I made some changes to the blindscan drivers to fix a
very old issue: most TBS cards never power off the diseqc equipment and LNBs (Perhaps this is the
reason I have accumulated a bog box of faulty LNBs over the years?). This consumes needless power.
It also prevents the use of priority swicthes and such.

I had fixed this recently and it turned out this was actually why I started seeing problems with the Amiko.
The old "keep all equipment always powered on" seems to have camouflaged the problem
with the Amiko perhaps for a very long time. Fixing this bug exposed the problem.

To test my theory of an overload, I made a software workaround: whenever the LNB needs to go from 0 to 18 Volt,
I now do that in two steps: first I power up to 12 Volt. I wait a little while and then switch to 18 Volt. The
idea is that the problem is that the Amiko does not handle well starting all devices (including itself) at the
same time and thus goes into overload mode. Starting with 12 volt power limits startup currents and
keeps the Amiko happy.

This actually works, but it should not be necessary, should it?

 

[a33]

What you describe seems to be the so-called inrush-current problem?
I've seen it described on a german forum, I believe; where especially Unicable LNBs seem to cause this problem.

The problem of the inrush current can be seen as two-fold: The receiver is too quick to respond with the protect mode, to a (short-moment) high inrush current; and the LNB (and other equipment) have too high capacity capacitors with no inrush current limiter.
I didn't know that a motor could also contribute to this problem.

From what I recall, the only solution is what you did: bypass the problem, somehow. Or buy other equipment (but it might be hard to determine which equipment contributes most to the problem)....

So, this is what your description did make me think of.

Greetz,
A33

 

[deeptho]

A33,

sounds plausible. At least in this case the receiver was not at fault because the voltages were still there.
The LNB is inverto, the switch Spaun. So, nothing exotic.

I should use scope to inspect the current. The voltages looked like this (at the receiver side). You can clearly see the various 22khz bursts
ending in a continuous 22khz signal. Perhaps due to the long cable and/or the large capacitors you meantion,
it takes about 400ms for the DC voltage to stabilize. The diseqc commands are sent while the voltage is still settling, but that works.



The motor specs state that it can go up to 350mA, and at least according to a multimeter, it does not
exceed that much (the highest was 380mA). The only thing the specs say about the current limit is
that there is one.

 

[Terryl]

I've seen something like this before, what it could be are your location is a momentary drop in DC voltage out at the motor, this would cause a spike in DC current, and thusly the over current protection (in what ever device) would cut in.

How I fixed it was to upgrade the coax to one with a pure copper centre conductor, and quad shielding, this gives you less DC voltage drop at the far end, and you should be using +13 volts DC for your tests not +12 as +13 and +18 are the normal DC switching voltages from a receiver.

Unless you're using that on purpose.

 

[deeptho]

My cable is already like you descibe. The voltage was I think slightly below 13 volts.
The high voltage was sowemwhat above 18 Volts (20? from memory) which could contribute to the problem.
The official voltages are indeed 13 and 18

 

[Terryl]

I think some LNB's have a Plus or Minus 1/2 a volt range, others that I have seen have a 1 volt range, +20 would be a bit too much.

The big problem is trying to measure a DC voltage or current on an active RF circuit as the probe used could/would cause problems on the circuit.

What I used in the past was a modified receiver, I found the LNB transponder voltage switching circuit and monitored the DC voltage and current from that point.

 

[deeptho]

I was thinking of using a small resistor in series. It will perhaps interfere with the RF signal, but that is ok for that test

 

[Terryl]

I don't have the modified receiver anymore, but I do remember a bit on what I did.

After a bit of poking around I found the variable voltage chip and circuit used to switch between the two voltages, I cut the trace for the main output voltage and did use a .1 ohm resistor like you mentioned in series, this allowed me to monitor the DC voltage going into the resistor, (voltage going to the LNB and the voltage drop across it, (current measurement) the circuit involved had RF suppression components to keep the circuit from interfering with the RF signal from the LNB.

It worked fine till one day the main tuner took a nose dive.

 

[a33]

I was thinking of using a small resistor in series. It will perhaps interfere with the RF signal, but that is ok for that test

I found the post, where an LNB inrush current of as much as 2700mA is reported:

Neuling hat LNB / TV Kuriosität

Nee, war mein Fehler. Im Eingangsbeitrag hattest Du ja geschrieben, dass der Dyon an einem der Legacy-Ports hängt. Damit sehe ich, außer auf Verdacht...

forum.digitalfernsehen.de

A resistor could work, and you would have a choice to put that at the receiver end (involving all current-receivers), or at an individual current-consumer ("the culprit").
I've also thought about using a coil (with a diode anti-parallel, for protection?) and a capacitor. No idea if that would work better.

Ideally, the receiver power supply should not be able to supply such a peak current, I would say. So that the coil (+ diode) should ideally be at the power supply?

Not sure what effect to expect from a .1 ohm resistor, as @Terryl described. Seems a very low resistance value, compared to a possible 2700mA...?

Greetz,
A33




Edit: I also found a post by @raceroad where he tested the difference between using an electrolytic capacitor, or a ceramic capacitor in an LNB:

Instabiler Empfang bei LG OLED65CX8LB und Unicable, Satellit (DVB-S) - HIFI-FORUM

Und vielen Dank für das lesen meiner Anfrage! Grundproblem: Mit meinem neuen LG TV bekomme ich unter Nutzung einer Unicable Installation keinen stabi

www.hifi-forum.de

It seems the latter is better.
Don't know if he exchanged them himself, or just compared different LNB designs. It seems to be the latter...

 

[Terryl]

A .1 ohm resistor is a common value for an in-line current measurement, you measure the DC voltage drop across it with a milli volt meter, or in most current measuring meters (amp meters) the voltage drop is what moves the meter. (analog type)

I found this digital meter that will measure up to 5 amps with an inline resistor of .01 ohms, (10 milli ohms) this so the small value resistor does not affect the circuit under test or measurement.

https://www.laurels.com/datasheets/dcmeter.pdf

 

[a33]

OK.
I thought it was about a resistor to limit the (inrush) current, but you are talking about a measuring resistor....

Confusion...

greetz,
A33

 

[deeptho]

For measuring I already tried simply a multimeter in the circuit. So I can measure current (up 50 380mA but very brieflly). More interesting would be the time behaviour. Peak currents and such. That is why I would use a resistor. The most time consuming part with all these experiments
is screwing/unscrewing these nasty F connectors.

 

[Terryl]

For testing they make push on connectors, you screw them onto the coax F connectors, then just push them on.

Google would be your best bet for these hummers.

 

[Terryl]

Like these puppy's... You would have to look for them at your location.

https://www.amazon.com/CIMPLE-CO-Coaxial-Connector-Corners/dp/B06XGWWXPW

 

[a33]

For testing they make push on connectors, you screw them onto the coax F connectors, then just push them on.

To add: I also know such plugs as named "F-quick-connector".
So the name isn't standardized.
For 'push-on F-connectors', also compression-type F-connectors are found (e.g.

) ...

And indeed: they are easy to use! So I always like to have a few lying around; same as barrel connectors!

Greetz,
A33

 

[deeptho]

The main problem I have with the screw connectors is that you have to insert them exactly right, or the the threads will not "catch"
each other. You keep turning the connector and nothing happens. Perhaps it is just an incompatibility between German and Chines models.

 

[a33]

Perhaps it is just an incompatibility between German and Chines models.

I had also ordered quite a few F-connectors from a Chinese website, some years ago. They fitted very badly to 'normal' female F-connectors, when I tried them.
Also their barrel connectors were junk: very bad thread.
So I've put them somewhere far away.

Still, with 'normal' connectors, I sometimes have the same experience as you: that you really have to push the F-connector a bit, for the thread to catch. I guess the center pin meets resistance, while being pushed in, so that the thread doesn't catch easily.

Therefore I find: those F-quick connectors are really handy, for test-setups.


BTW I also had bought 4 very cheap chinese 8/1 diseqc 1.1 switches (2 for direct use, 2 for spare). Their F-connectors were 'normal', and they switch normally. Just one has become defective, after a year or two. So I'm still content with those...


Greetz,
A33

 

[deeptho]

I have the opposite problems: The male crimpon connectors I ordered from Germany interact poorly with Chinese male connectors. So the only solution will be for some couples to get a divorce and then marry another connector :->

Especially my satellite meter is very stubborn.