Power supply output current and laser power

Even if the board operates at 3.6v, the fact we can’t get more than 2.0v out of that DA1 concerns me. Should it, or is there a valid reason it could be set up like that?

We have of course no record of what voltage was being sent to the OG PSU when it was working and no way to test it now…we put it back in the other day and the LCD panel just produced error codes, so it’s obviously finally given up the ghost rather than limping along. However we were assured our new one was a direct replacement.

I think we’re convinced the laser tube is not part of the problem, but other than that we still have too many variables. We MAY have replaced the old PSU with a model that is sold as compatible, but isn’t truly. The controller card MAY have failed to deliver the DA1 analog voltage for some reason. It MAY be operating on a 3.6v system instead of 5.0v.

Actually one thing I forgot…I stuck the multimeter into a lot of points on the controller card today…including the clearly labelled DC5v points for the X, Y and Z-axis motors. These DID all read exactly 5.0v.

In your experience, is it possible that a controller card could have 5v and 3.6v? My simple brain (as in I try to simplify things wherever possible) would assume “5v for all” to be a better option.

What we DO know now (thanks to you mostly) is that our new PSU expects a 0 to 5v input and translates that into a 0 to 100% strength value in terms of the maximum milliamp setting controlled by the potentiometer; (a little recessed screw that is the very devil to find and when you do, offers no assistance in which way to turn it or how much effect a half turn or full turn has on the output. Guess. Twiddle. Check for effect. This is not my idea of fun).

But I’ve made at least one decision. We are changing our “test” steps from 0%, 25%, 50%, 75% and 100% power to 0%, 20%, 40%, 60%, 80% and 100%.

Why?

Because if we are working from 0 to 5v, they should equate to single volt steps; 0, 1, 2, etc.

See? Simpler. :grin:. I do like simple.

If we configure the replacement card correctly, there are three options for the output when 100% is selected as follows

2.0v
Then this is exactly the same as our current card. Regardless of whether it’s operating on 3.6v or 5.0v, this would seem “wrong”. This for me would be the worst result.

3.6v
Means your OG theory is correct and we need to find an OG PSU to match. Second worst option.

5.0v
Means the board failed (perhaps as a result of the PSU failing and taking some component or other on the card with it…or vice versa). Best result. Once we’ve installed the card it should work with our new PSU and we can start cutting (and posting dancing kittens here).

Here’s hoping.

Jack; thanks for stepping into the fray!

Good to confirm that L=Logic, thanks for the Wiki link. I don’t understand it all in detail, but it provides enough info to be of some help. I’m familiar with 3.3v logic and 5.0v logic from Arduino systems, and the occasional need to translate between the two. I’ll admit I hadn’t encountered 3.6v logic until Ed mentioned it.

Still a bit confused by the “active high” and “active low” and including the water cooling confuses me further…I don’t know our machine well enough (disclosure; my original role for the owner was “VPM” - Volunteer Packing Monkey, but I get involved in all the stuff that gets in the way of him designing, which is his unique strength and in my opinion the companies . So things like chasing filter companies whose activated charcoal filters went from having a lifetime of 6 months to barely making it to four weeks).

Sorry I digress…I thought the “active low” on our machine just referred to the laser; I was of the opinion that the cooling system was separate, but I may be wrong. One more thing for me to check!

Unless you work with this, it’s unlikely that a single article will educate you to a masters in electronics. Just pick up what you can and it will eventually snap for you.

Water cooling has nothing to do with the active state of a signal, I used that as an example, appears not to be a good example…


Signals, in the digital world are either off or on. What the computer (controller) is looking for may be a high signal or low signal.

A pwm output signal is defined as it being on (or active high) for a percentage of it’s period.

Inputs to the Ruida are generally active low, meaning that a valid door protect signal will be active low when indicating the door is shut. Inputs are generally this way because many manufacturers have an optical isolator in circuit, when pulled low, will complete the circuit to ground turning the isolation on or active.

Hope this make sense?

The output driving the lps needs to have full 0 to 5V swing to be able to control the normal lps.

:smile_cat:

Clear as mud! No just joking I get most of it. My confusion is mostly around the fact that this isn’t a ruida driven machine, it’s a Leetro card (MPC6565), that we use LaserCut not Lightburn (for now), and that a lot of the settings I see make me think we should be active low not active high…but we are connected to the H line, not L on the PSU.

However if we got that wrong, surely the laser would cut when we think it should move and move when we think it should cut?

The optical sensor I get. If it sees light, it’s on (“high”) so covering it up with the door turns it off (“low”)?

Skipping over a whole bunch of stuff, it is possible the controller is producing differential signals, so LAS+ and LAS- signals move in opposition: one is high while the other is low. Then the controller has a separate polarity configuration that determines whether the LAS+ output is high-when-active or low-when-active, with the LAS- signal being the exact opposite. Ditto for the PWM outputs.

Differential signalling has some advantages in noisy environments, none of which seem to be taken advantage of in this machine. Perhaps they’re ignoring all that, using just the + outputs, and setting the controller config to make that output active-when-low.

It’s obvious they never intended for civilians to get this far under the hood. :grin:

The Sparkfun level shifter from contemporary 3.3 V logic to OG TTL levels might get the job done. The one caveat will be the frequency response: the 10 kΩ pullup resistors can’t drive cable capacitance very well.

This comes into play with the 20 kHz PWM carrier frequency I assume the controller is producing. The frequency may be a configuration option as well, but you can’t shouldn’t set it much lower than 10 kHz to keep it far above the PWM demodulating filter in the HV power supply.

Yick & similar remarks.

Yes I did. Symptoms of Parkinson’s include lowered cognition when the pills wear off. I’ll read it later. I’m just smart enough to know I’m not smart enough right now.

Yick, indeed.

1 Like

Right. Pills active, brain on.

Working backwards…I’ve definitely seen an option to set the PWM frequency in LaserCut machine settings and from memory 20 kHz was the default figure.

Sparkfun logic translator was just a “something like this” example; I’m open to other options (whilst at the same time hoping they turn out not to be necessary).

Differential signalling; wouldn’t this require both LAS+ AND LAS- to be connected? They’re not, so if that’s required, this isn’t a thing.

There’s a rule of thumb in programming: never write the most clever code you can, because you’re not quite smart enough to debug it.

One benefit of experience is expecting Future Me won’t be quite as bright as Present Me, because Past Me could obviously think rings around both of me. :grin:

Got it in one!

Because both outputs have the same information, however, you can use either one if you don’t need the noise-reduction advantages of differential signalling. The controller’s active-when-low setting makes the + output go low when it’s active, so the overall effect is the same as the single (non-differential) output on a Ruida controller.

Poking around a bit produced an overview sheet showing the same wiring as in your machine, so that’s what they intended to do.

An MPC6575 manual (which I assume is the same as a 6565, only more better-ish) shows the same differential pins, but also says “All inputs and outputs are single-ended”. Which may be a mild mistranslation, as each individual terminal pin is single-ended with respect to the circuit Common voltage, but the two pins for each signal are differential with respect to each other.

Conspicuous by its absence: electrical specs like voltage / current / pullups / whatever.

Most likely, I have no idea what they’re doing.

The only thing that’s differential used on these that I know of is the USB port and I doubt they really handle it that way or they would be more dependable.


N is neutral and L is line, or the mains inputs. From @ednisley link.


These are marked similar, are inputs, but the +/- doesn’t related to differential control.

@Grey_Area1 Active is relative to the something else. Many systems use a pull up, where the input is actively pulled to supply voltage or a ttl type high signal. This controller may use positive voltage as active. Yours is wired to L (which, I assume, would be the same as LAS).

Is there any markings, on the lps itself, where these wires go? A photo would be great.

:smile_cat:

I’m not religious, but this produced an involuntary “Amen!”

1 Like

Scroll up, I think I posted pics of both the inputs on the card and power supply already?

Edit: post #7 in this thread…

Yes I found the photo. Thanks…

It looks like the normal wiring for these.


I think we’re making this more complicated than it needs to be. If there is a problem, the wires wouldn’t change… I’d look at a configuration issue or hardware issue. Kind of where you started…

To isolate this, we can look at your measured voltages to the lps, which are wrong coming from the controller itself. If the controller is sending bad signals, then it’s a controller/software issue and that would eliminate the lps and tube. Which you have replaced.

You can measure these with the lps unplugged… remove the signal line (biggest, on right) and measure your signal at the controller

All of these should be measured with respect to ground.


Usually when these fail, it’s total failure, not it isn’t working quite right… Computers have a tendency to do that, especially if we’re going to assume a hardware failure of the controller.

:smile_cat:

I think we did that and still got the “not quite working properly” result of 2v output when asking for 100%.

Really hoping the replacement card will enlighten us, but I’ve thought that at every stage, and so far each time we’ve just taken more steps into the dark…

Tomorrow is another day!

What replacement card?


If you controller is broken, it needs to be fixed.

:smile_cat:

We have a contact who is going to provide us with an MPC6565 that he just happens to have lying around. We’ll be open-hearting (or braining?) as soon as it arrives. Really hoping for some actionable info sometime soon; I’m sick of scratching my head.

That is rapidly becoming a very popular club…

Further poking around suggests other cheap & readily available level shifters / translators aren’t quite up to the task, either. The problem is they’re all intended to work with contemporary “TTL” voltage levels, not the voltages you’ve measured in your system, on one side and lower signal voltages on the other side. They’re not intended to convert between two different “TTL” voltage levels.

For example, the TXS0108E level shifter found all over Amazon requires a low-side power supply under 3.6 V and can’t handle inputs higher than that, which is too low for the Leetro’s output.

The 74LVC245 requires no more than a 3.6 V supply, but can withstand inputs up to 5 V, so it would handle the Leetro’s outputs. They’re only available as bare chips, at least on Amazon, which turns it into a circuit-building / debugging exercise.

The dirt-cheap “IIC” level shifters come with no specs at all and images with missing / illegible topmarks, so I assume they’re junk from the QC reject bin.

AFAICT, grafting a new power supply onto that old-school controller in the absence of any useful doc is doomed to failure.

Ok, well thanks for taking the time to check, really appreciate it. Snow day today in the UK, so no progress. :disappointed:

@jkwilborn (not related to #37 post) I have the UTG932E maybe the output can be checked with a Function Generator set to 10KHz square wave and varying the amplitude [0,5]V?

Now there is a good idea!

To mimic the PWM output, you’d set:

  • Square wave output
  • Frequency = 20 kHz (to match the controller setting)
  • Amplitude = 5 V peak-to-peak (to match the supply input spec)
  • Offset = 2.5 V (so it’s 0 to +5 V, not ±2.5 V)

Then as you vary the duty cycle from nearly 0% to nearly 100%, the tube current should rise from nearly zero to nearly its maximum value.

Dump the beam into a big mug of water, because the laser tube will be running continuously as you twiddle the duty cycle and watch the current.

For science! :grin:

1 Like