295 replies to this topic

[firemachine69]

I agree with the statement that these batteries can't be charged while cold. Your statement "discharge them below zero it damages them and they'll fail on you fairly quickly."

 

I use my Celestron Evolution regularly at -20-30 degrees Fahrenheit (below 0) as that is when we get our clearest skies here in the winter. My Batteries are still going strong after many recharges. Of course, I always let my Evolution mount warm up after bringing it inside before charging. When the telescope is that cold moisture condenses on it quickly. So, I let it warm up and dry before recharging it. 

 

 

I have a simpler solution, I don't even use it in the winter unless it's above freezing. 

[roelb]

I agree with the statement that these batteries can't be charged while cold. Your statement "discharge them below zero it damages them and they'll fail on you fairly quickly."

 

I use my Celestron Evolution regularly at -20-30 degrees Fahrenheit (below 0) as that is when we get our clearest skies here in the winter. My Batteries are still going strong after many recharges. Of course, I always let my Evolution mount warm up after bringing it inside before charging. When the telescope is that cold moisture condenses on it quickly. So, I let it warm up and dry before recharging it. 

Their is a temperature sensor on the battery to avoid charging at minus 0 °C.

[archer1960]

I have a simpler solution, I don't even use it in the winter unless it's above freezing. 

I'd lose an awful lot of imaging time if I followed that rule!  :-)

[micgo]

hey guy. i m having a humming noise when turning left and right. feel like it's a problem related to the clutch. i have found a video online and it's same exact problem. any idea ?

 

https://www.youtube....h?v=vs_bQuuqSEo

 

micgo

[mlord]

My Evo-8 decided to start behaving exactly as described in the first post here, not powering on and blinking the various LEDs in an unhealthy fashion.  I've pulled the circuit board, and am gradually fixing things as I find them.

 

Most of the connectors on this board have tiny little fuses on most signals, to protect against Bad Things™.  These little grey coloured devices are all numbered with the prefix "FB".  So first I went around checking every one of those, on both sides of the board, with my ohm-meter.  Each of them should read zero-ohms, unless the fuse has blown.  All of mine were fine.

 

Next thing to check, if the mount is not powering up, is the +5V regulator which powers much of the on-board stuff.   Here it is, component U5 circled in RED:

 

 

The one on my board was labelled as a GH18H, which is a version of the AMS1117-5.0 regulator commonly used on all sorts of stuff.  You can test it by powering the board from a +12VDC power supply through the DC jack, and turning the switch on.  For this to work, the small 2-pin connector from the switch, and the 3-pin connector from the DC jack need to be plugged in.  Nothing else has to be plugged in.

 

The pin-out of the regulator is as shown:

 

 

Turn the power switch on.  Use a volt meter and measure the voltage between Ground and the +12V pin -- it should be around +12V, but anything from 7.5V and up is fine for now.  Mine read about 12V exactly.

 

Then measure the voltage between Ground and the +5V pin (or the tab), which should read at least 4.8V and no more than 5.1V.  If not, it needs to be replaced.  Mine read 0V, so I knew it was dead, or there was a short on the output side somewhere.  Not seeing any smoke, I figured it was dead. 

 

I have a bin full of that kind of stuff here, so I removed the original regulator and swapped in a regular AMS1117-5.0 in place of it.  The mount now powers up, shows the battery charging briefly (it was already fully charged), and the hand-controller mostly works on my mount. 

 

Next, I tried slewing the mount from the hand-controller, and discovered the AZM motor at the base of the mount is not working.  So I swapped connectors between it and the ALT motor, and then the ALT motor was not working.  So the two motors and their cabling are actually fine, and the problem is on the main circuit board.

 

The obvious place to look is the L293DD motor driver chip.  I could probe the inputs to see if it is being given the correct signals from the onboard micro-controller.  If so, then the L293DD will need replacing.  Or I could just replace the chip anyway, and then see what happens..  haven't decided which yet.

 

Cheers

Edited by mlord, 17 October 2022 - 02:19 PM.

[mlord]

Note that the Evolution's circuit board uses modern Surface Mount Devices (SMD) for nearly everything.  This makes it more difficult to work on than older through-hole style boards.  Removing a multi-pin SMD, without damaging the circuit board itself, can be quite challenging.

 

I use a hot-air rework station (https://www.amazon.c.../dp/B07FBGFT3K/) rather than a soldering iron for stuff like this.  Eg. to remove that voltage regulator, I just set my hot-air gun to 330C and pointed its air nozzle directly down on the regulator, moving it around in small circles for 10-15 seconds until the chip could be lifted with no resistance using tweezers.

 

Installing the new regulator was done similarly:  I appled lots of flux on the four pads, placed the new chip onto them, and again heated it all up until I could see the solder melt and flow slightly.  This could instead have been done with a regular soldering station.

 

If/when I replace the L293DD chip, I'll probably use the hot-air station again, though that chip can be challenging because the middle eight pins are all connected to Ground as a heat-sink.. I may have to just snip those pins off in place, and remove them after getting the main chip body off.

 

Another way to remove SMDs would be with special ChipQuik low-temperature solder, which enables complex stuff to be done without use of a hot-air station.  I have that here too. 

Edited by mlord, 17 October 2022 - 07:11 PM.

[mlord]

If/when I replace the L293DD chip, I'll probably use the hot-air station again,

..

Another way to remove SMDs would be with special ChipQuik low-temperature solder

I figure it'll be easier and safer to test the chip off of the board than on, so time to remove it!

 

After looking at the SMD devices directly underneath the L293DD chip, heating up such a large area was bound to cause something else to drop off the underside in the process.  So instead..  30 seconds of ChipQuik later, plus clean-up, and there we have it! 

 

 

I can now test the chip with test-leads and a volt-meter.  If it all looks good logic-wise, I'll reattach it, but with the three input pins (enable2, input3, input4) for the AZM motor "lifted"(*).  Then I can just apply +5V and/or GND to those pins to see if the AZM motor actually moves for real.  If either set of tests fail, a new chip gets soldered on once I get hold of one (it's in the mail).

 

Otherwise.. bad news, this would mean the main micro-controller has likely failed on one or more of the GPIO pins associated with those three signals.

 

(*) Lifting the pins is a lot easier and less risky with the chip removed than trying to do it in-situ, especially with those nearby large capacitors getting in the way.

Edited by mlord, 17 October 2022 - 07:46 PM.

[1979mgb]

mlord,

I like your contribution to this topic. Nice to know that these boards can be fixed. Many would never attempt it. However, with this kind of input the brave can press on.

After all, if the board is going to be replaced some inexpensive components are worth a shot.

 

I thank N2KEN for starting this topic and to all who have contributed to our understanding of this mount.

[mlord]

I like your contribution to this topic. ..

Thanks.  My philosophy here is quite simple:  It's already broken, and there's nobody else who can/will repair it any more cheaply than me just buying a replacement board.  So.. Why not give it go? 

 

The worst I can do is.. still end up with a broken board.

But I'll hoping for much better than that!

 

Cheers!

[mlord]

I accidently fried the L293DD chip while hooking it up for testing, so further progress here will have to wait until the replacement chip(s) arrive, likely next week.

 

Otherwise.. bad news, this would mean the main micro-controller has likely failed on one or more of the GPIO pins associated with those three signals.

If it later turns out that the main micro-controller is suspect, things get a bit more complex.  I _think_ those chips can still be sourced, but they also likely need to have some bootloader code pre-programmed into them before they will work (and accept a firmware upgrade) on this board.

 

 

But.. in the photo above, one can see a bare pad of 10-contacts nearby.  My first guess, without looking at the pin-out of the controller, is that this is intended for a JTAG-style connector, used by Celestron to directly download code (including bootloaders) to the main chip.  And if they can do it, then so can we! 

 

Most likely it's actually for a Serial Peripheral Interface (SPI), with which I happen to have considerable prior familiarity.  These are commonly used as a direct interface for programming internal flash memory.  Maybe I'll look up that chip now and see.. just in case.

Edited by mlord, 18 October 2022 - 03:54 PM.

[mlord]

If it later turns out that the main micro-controller is suspect, things get a bit more complex.  I _think_ those chips can still be sourced,

The micro-controller is a MK10DX64VLH7, with 64KB of built-in flash memory.  The empty pads near the chip do indeed connect to its JTAG pins, for programming the flash memory with a bootloader or whatever.

 

The chip appears to be unavailable world-wide -- not even on eBay !! 

[1979mgb]

Try this link.

 

https://www.hkkzd.co...circuits-ics-16

[mlord]

Try this link.
...

Good find!  I'll see if they actually will sell me a couple, just in case it turns out to be needed.

 

Northern NY..  where abouts? We're probably not that far apart.  I'm quite close to Ogdensburg.

[1979mgb]

Massena, Cornwall bridge crossing. I have some used boards if you like. My test rig is available too. See post #138.

[mlord]

Thanks to the generosity of user 1979mgb, I now have a sample set of dead and partially-working "donor" main boards (Rev.F) for the Evolution mount.  Thanks Charles!

 

One surprise, is that the ALT/AZM motors in the Evolution appear to be powered at about 14.5VDC, instead of the expected 12VDC.  I initially saw this with my own board, but wasn't sure if the higher voltage was due to something else wrong with it.  But I see the same voltage on the best of the new "donor" boards, so it's probably intentional.

 

A common point of failure seems to be the L293DD motor driver chips.  My board had a partially dead chip there, and so did the first "donor" board I tested.  Two more to go.  My UK-based electronics wiz buddy tells me that a common failure mode for H-bridge chips (eg. L293DD) is where they develop a dead short internally in at least one mode of operation.  This would explain why the 14.5VDC power supply circuit on my own Rev.K board was making click-click noises before I removed the L293DD chip -- and now it does not make those noises!

 

The L293DD chips are rated for up to 36V, so 14.5VDC isn't an issue for them.  But maybe there is a start-up voltage spike at power-on that is shortening their lifespans -- just a guess for now.  If I can later prove this to be the case, then perhaps a zener diode might get added to limit such spikes.

 

More when/as I learn more.  But thus far, very encouraging!

 

Anyone else here have a surplus dead Evolution main board?

Edited by mlord, 22 October 2022 - 08:22 AM.

[mlord]

One of the other "donor" boards turned out to have a fully functional L293DD chip.  I removed it using ChipQuik, and tested it on a breadboard first.  It appeared to be fine, so I soldered it onto my own Rev.K board using the hot-air re-work station, and hooked everything up again.

 

IT WORKS! 

 

No complaints from the mount now -- runs fine from battery, runs fine with mains connected.  And there's hope for at least one other of the donor boards here now too -- I just need to receive the batch of L293DD chips ordered from China.  They're probably still about 7-8 days away.

 

Meanwhile, time to reassemble the mount.  Apparently Halley's comet is chucking debris at our atmosphere tonight (meteors)!

 

Cheers

Edited by mlord, 21 October 2022 - 09:56 PM.

[1979mgb]

Mark,

Glad I could help. I could tell that you were the component level person to benefit from the spare "DEAD" boards. Obviously there had to be good components.

Nice to know that components from older revisions can be used to repair newer revisions. 

 

So, everyone please don't throw those "DEAD" boards away make them available.

 

Thanks, Mark for your additions to this topic. It will be a help to those willing to tackle this level of repair. After all what is there to lose if it's "DEAD" anyway.

Keep up the good work.

Charles

Edited by 1979mgb, 21 October 2022 - 10:03 PM.

[mlord]

A common point of failure seems to be the L293DD motor driver chips.  My board had a partially dead chip there, and so did the first "donor" board I tested.

 

The L293DD is a motor driver chip.  It has four output pins, two for each motor (ALT, AZM).  What makes the chip special, is that it can switch either GND or V+ (14.5VDC here) onto any of the output pins, under software control.  So by driving a pair of those pins as V+,GND the motor will spin clockwise.  But it can also be made to spin in reverse, counter-clockwise, by driving that same pair of pins as GND,V+ instead.

 

This means that internally, the L293DD has to have two driving transistors (one for V+, one for GND) for each output pin.

 

The failure mode for the L293DD is when one of those driving transistors dies, effectively turning it into a straight piece of always-connected wire inside the chip.  So, if say that "wire" is the one for V+, and software then tries to drive that pin as a GND pin, we end up with V+ connected straight to GND inside the chip.  A hard short-circuit, causing the power supply to rapidly cycle on/off trying to protect itself, as the chip turns into a space-heater.

 

It is possible under that scenario for the chip to also back-feed onto the +5V control lines, damaging the +5V regulator and possibly other things in the mount.  This is what I think must have happened on my own Rev.K board.

 

The Evo main board has built-in fuses on all of the outputs, but not on the inputs to the L293DD chip.  A way to protect the rest of the board from a failing L293DD chip would be to also have fuses on the two power lines to the chip, Vs and Vss.  In the event of a short-circuit failure per above, the fuse(s) would blow, cutting the chip off from the rest of the main board.

 

There are actually (large) fuses in the 14.5VDC power supply section, but they don't get time to blow because the power supply processor detects the short and shuts down, then turns on, then shuts down, then.. and so on.
 

I don't see an easy way to retrofit fuses, other than perhaps lifting the Vs and Vss pins, and soldering SMD fuses between each of those and its corresponding pad on the circuit board.

[mlord]

Here's a link to a more detailed overview of the L293D/DD chips:  https://diyodemag.co...motor_driver_ic

 

The L293DD in the Evo is different from the L293D (single-D) used in some other Celestron mounts.  I wonder if the designers at Celestron took the differences into account though.  The L293D (single-D) has a lot more tolerance for spikes in current draw as motors start up or change direction.  The L293DD is much more fragile in that regard.

 

One thing I will do when my supply of fresh L293DD chips arrives, is to lift the Vs pin of one chip so that an amperage meter can be inserted between it and the 14.5VDC supply.  Then I should be able to measure actual and peak current draw through the chip as it operates both motors simultaneously.

 

The chip limit is a mere 0.6A, and I suspect the mount exceeds that limit, which would explain why so many of these seem to fail.  If this turns out to be true, then a solution might be to use two L293DD chips, and use one for each motor, rather than powering both motors from a single chip.

 

Maybe I should order more chips...

Edited by mlord, 22 October 2022 - 09:36 AM.

[mlord]

Many have missed one point that leads to problems and damage to the board. You cannot charge the mount and use it at the same time using a 12V 2A power supply. To do this, you need to buy a 12V 5A type meenwell for $ 18 or a Celestron and set 4.5A in the settings. This is stated in the instructions (to charge and use the mount, a 5A power supply is installed for this).

 

The Celestron Manual for the Evolution says:

External Power - Sets the maximum potential current draw from the power supply Default is 2 0A for the included power supply. Any setting higher than 2.0 requires a higher capacity power supply, which is not included with the telescope. The telescope has built-in fail safes if the External Power setting is incorrectly set, but we recommend always using a suitable power supply for the given setting

 

When used with a higher capacity power supply, this setting allows you to charge the internal battery at the fastest speed while using the telescope, and while optionally charging your smart device from the USB charger.

 

As one can read, there is no prohibition against charging and using the mount while powered from the included factory 2-amp power supply.  It will simply charge more slowly when the mount is drawing power, than it would if a higher-capacity power supply were connected and configured through the menus.

Edited by mlord, 22 October 2022 - 03:50 PM.

[mlord]

Two questions for those here who have replaced the main board in their Evo:

• Did you keep the original dead board?
• If so, can I have it?

Since few outside of Celestron seem to be able to repair these things, and lead times there are measured in months.. I'd like to collect a few more dead boards for parts, which might eventually enable me to fix future dead boards for others here.  Not as a business or anything, but as a hobby service to get mounts back in business sooner.  But for that, I need your dead boards to learn from and use for parts.

[GFisher2001]

Two questions for those here who have replaced the main board in their Evo:

• Did you keep the original dead board?
• If so, can I have it?

Since few outside of Celestron seem to be able to repair these things, and lead times there are measured in months.. I'd like to collect a few more dead boards for parts, which might eventually enable me to fix future dead boards for others here.  Not as a business or anything, but as a hobby service to get mounts back in business sooner.  But for that, I need your dead boards to learn from and use for parts.

 

1. Dropping in my mlord repaid board now. I marked all of the cables fairly well, BUT, I'm not 100% positive I got the Power SW wire right (although it really has to be right) and I didn't do a good job distinigshing the ALT motor Wire from the AZM motor wire. 

 

Assuming I was bone headed and swapped the ALT/AZM motor wire, things won't fry, just operate differently right? 

How damaging would getting the Power SW and say the Wifi SW or Wifi LED wire be? 

 

2.  Does anyone have a photograph or skematic of what color wires attach to which poles on the power connector? I want to triple check before connecting power to my board that I rewired it properly. 

 

Note: I stole this image from earlier. 

 

Edited by GFisher2001, 12 November 2022 - 01:00 PM.

[mlord]

The labels for all of the connectors are on the BOTTOM of the board.  Just take a photo of the underside before reassembly.  

 

Here's a photo of the underside of a Rev.F board, which has exactly the same connector layout as Rev.J and Rev.K, and shows the labels for each connector:

 

[GFisher2001]

Awesome, how can i test the Red, Green and Blue wiring to make sure my replacement 12V was wired correctly this time?

 

The bottom just says "Power". 

 

I have a multimeter but I don't know what color should read what. 

[mlord]

Give me a few minutes to fetch my allen keys and I'll open up my mount and photograph it for you.