[TEST, PIC HEAVY] Cheap eBay 5A TTL driver

Hey !
I recently bought one of those cheap laser drivers: 5A 12VDC Power Supply Driver for 445nm 450nm 3.5w NDB7A75 Blue Laser LD with TTL
Not to use it as an actual driver but to perform tests on it.
But hey, if it works fine I might use it with my NDB7A75.

So I started by making a test load with 1N4007 diodes and a 5W 1ohm resistor:


(the Arduino is just used to "switch on or off" the driver using the TTL port on it)
When the driver is powered, it takes about 5s for its "ready LED" to light up, after that it starts giving juice.
First thing first, the output voltage:


I measured that after about 30s of run-time, since the voltage seems to drop a bit over time (sometimes I measured 5.2V without any voltage drop but most of the time the voltage starts at 4.8V and drops to around 4.7V ).
Then I measured the current with my oscilloscope (by measuring the voltage across the resistor; 100mV = 100mA) to see if there was nasty ripples.
Start:


(don't forget to look at the bottom left and bottom right for the scales)

A bit closer:


During operation:


The driver is outputting about 1.2A at average here.
Which means that it outputs about 5.64W while the power supply was giving 6.24W (so the driver have an average efficiency of 90%).

A closer look on the ripples:


So yeah I'm not a pro about laser drivers but those ripples of about 300mA peak to peak seems a lot to me.
What do you guys think ?

Will comment more once I'm done dealing with these network outages.

May want to resize your images, they fit fine on my screen, but others may complain that they're too large. 800px horizontal is usually fine, and whatever that results in vertically.

diachi said:

Will comment more once I'm done dealing with these network outages.

May want to resize your images, they fit fine on my screen, but others may complain that they're too large. 800px horizontal is usually fine, and whatever that results in vertically.

Click to expand...

Ok no problem !
Doesn't the forum resize the images automatically accordingly to the screen size ?
I'll change the resolution when I'm back home.

Sceeker said:

Ok no problem !
Doesn't the forum resize the images automatically accordingly to the screen size ?
I'll change the resolution when I'm back home.

Click to expand...

It does resize them for me, doesn't seem to do it for everyone though, seeing as every time someone posts larger images someone complains about them being too big! :thinking:

great info and testing! im not so pro on electronics but i would only accept a 300ma ripple difference on the high powered nubmxx series diodes that are very robust.dont use it on a ndb7a75 diode,i wouldnt risk it as its not that cheap diode and especially if you plan to drive it above the rated current

deadshadow said:

great info and testing! im not so pro on electronics but i would only accept a 300ma ripple difference on the high powered nubmxx series diodes that are very robust.dont use it on a ndb7a75 diode,i wouldnt risk it as its not that cheap diode and especially if you plan to drive it above the rated current

Click to expand...

Thanks!
Well it's more like 200-230mA average (the very big peaks are not always here) but yes that still is enormous!
I was planning to use it with my NDB7A75 before the tests, I changed my mind afterward

Have you checked the input signal to the driver? Maybe that'd explain away some of your noise? Although I wouldn't be surprised if it's the driver causing it. Those China drivers are cheap for a reason.

Voltage change you're seeing could be a result of the diodes heating up as you run the test. Keep in mind Vf of a diode changes depending on current too.

My very first thought was your diodes in your dummy load aren't robust enough to measure anything higher than a watt. You need to rebuild it using at least 10 amp 1000 PRV rectifiers with a 0.1 ohm 1% 5 watt resistor. That will mean that your current will be 10X the measured voltage drop across your resistor. If I am reading your scope measurement correctly, you are going from off to on and we are seeing that transition. The small amount of ripple you have is occurring as pulse trains. Could it be introduced by something external to the driver? At any rate, it isn't large enough to be of concern to me.

diachi said:

Have you checked the input signal to the driver? Maybe that'd explain away some of your noise? Although I wouldn't be surprised if it's the driver causing it. Those China drivers are cheap for a reason.

Voltage change you're seeing could be a result of the diodes heating up as you run the test. Keep in mind Vf of a diode changes depending on current too.

Click to expand...

The input signal to the driver comes from my laboratory power supply, which has an extremely clean output so it doesn't come from that :/

That's what I thought for the voltage change so I tried with some 1N5408 and even with some MUR1520 (with heatsinks!) to eliminate the heat parameter but it still happens.
As for the current, it was always the same so I don't think that it comes from that.

Sceeker said:

The input signal to the driver comes from my laboratory power supply, which has an extremely clean output so it doesn't come from that :/

That's what I thought for the voltage change so I tried with some 1N5408 and even with some MUR1520 (with heatsinks!) to eliminate the heat parameter but it still happens.
As for the current, it was always the same so I don't think that it comes from that.

Click to expand...

No, I mean the input modulation signal, not the input power.

If the current is remaining constant and the voltage is changing that suggests something is changing with your load. Seeing as it's a constant current driver, it'll adjust the output voltage to keep a constant current based on changes at the load.

paul1598419 said:

My very first thought was your diodes in your dummy load aren't robust enough to measure anything higher than a watt. You need to rebuild it using at least 10 amp 1000 PRV rectifiers with a 0.1 ohm 1% 5 watt resistor. That will mean that your current will be 10X the measured voltage drop across your resistor. If I am reading your scope measurement correctly, you are going from off to on and we are seeing that transition. The small amount of ripple you have is occurring as pulse trains. Could it be introduced by something external to the driver? At any rate, it isn't large enough to be of concern to me.

Click to expand...

Like I said on my last post I tried with different diodes (I just didn't makes any photos) but the results were pretty much the same.
Apart from this, I don't have such a resistor but thanks, I'll buy one soon!

Yes I'm going from off to on.
Are you talking about the ripples of 300mA peak to peak ? You don't think it is a problem ??
Or are you talking about the ripples that are measured before the driver is powered up (or starts outputting something, because it is already powered up, I only send 5V to the TLL port to make it start outputting) ?
If yes then I'm not sure I know where they're coming from.
Could be "noise" from the Arduino being plugged in/powering up.

I don't know if I explained myself well, English is not my native language
I hope I didn't repeat myself too much.

diachi said:

No, I mean the input modulation signal, not the input power.

If the current is remaining constant and the voltage is changing that suggests something is changing with your load. Seeing as it's a constant current driver, it'll adjust the output voltage to keep a constant current based on changes at the load.

Click to expand...

Oh ok ! The input signal is not modulated it's plain 5V, the driver does not output anything if nothing is sent to the TTL port.
Sending plain 5V is supposed to make it work as a "normal driver", with a constant current and no modulation.

The voltage only changes the first few seconds, after that it stays quite stable.
They both stay constant (since the load doesn't change).

I tried 3 different browsers. I'm not seeing any images. Just me?

sometimes I measured 5.2V without any voltage drop but most of the time the voltage starts at 4.8V and drops to around 4.7V

Click to expand...

That's because the forward voltage of a PN junction drops as the temperature rises. You have several PN junctions in your test load. This is normal.

paul1598419 said:

You need to rebuild it using at least 10 amp 1000 PRV

Click to expand...

Why would the reverse bias voltage rating matter AT ALL, if you never reverse bias the diode, Mr. Paul?

Cyparagon said:

That's because the forward voltage of a PN junction drops as the temperature rises. You have several PN junctions in your test load. This is normal.

Click to expand...

Yes but like I said a bit earlier I tried with different diodes and even TO-220s with heatsinks to try to get rid as much temperature as possible.

Sceeker said:

Oh ok ! The input signal is not modulated it's plain 5V, the driver does not output anything if nothing is sent to the TTL port.
Sending plain 5V is supposed to make it work as a "normal driver", with a constant current and no modulation.

The voltage only changes the first few seconds, after that it stays quite stable.
They both stay constant (since the load doesn't change).

Click to expand...

The input signal can still be a source of noise, even if you aren't dong anything with it other than holding it at 5V. It's supposed to just make it turn on over a certain voltage but that doesn't mean it's actually doing that. Check your 5V modulation signal. It's probably fine, but check it anyway.

To reiterate on what Cyp said, Vf decreases as temperature of the PN junction increases. Even if you use diodes in a TO-220 package bolted to a heatsink, the junction temperature will start to increase as soon as you start to pass larger amounts of current through it.

The datasheet for your diode should specify a thermal resistance for the PN junction to case, measured in °C/W (Temperature difference per watt of heat dissipation) most likely.

Those ripples are a switchmode power suppply keeping a constant output - the occur when the output cap voltage is low, and disappear when it is restored to sufficient voltage. The chip is taking an average here it seems.

You can improve this by adding more output capacitance. Since there is a cap already there this will not make things worse: if you power up the driver and connect the laser afterwards you're likely to destroy it.

As for the test load: that's really no good for anything near 5 amps. This 1n4007 diodes take 1 amp continously, at 5 they will overheat to the point of going short or open circuit, i've even seem them desolder themselves from circuit boards (and even surviving that).

As in the 1 ohm, 5 watt resistor the problem is a lot bigger: if you ran 5 amps through that, it would dissipate 25 watts.

At this point this is a 1 amp testload, with the resistor being potentially good up until about 2.2 amps.

By the scope screenshots i guess you are currently running/testing it at 1 amp which is okay, but if you ramp it up to any more current, you will need to build a better test load.


Oh and a PS: i can't tell from the picture, but make sure the resistor is a film type, not a (more common) wire would one since high the inducance of those can cause serious problems for drivers.

diachi said:

The input signal can still be a source of noise, even if you aren't dong anything with it other than holding it at 5V. It's supposed to just make it turn on over a certain voltage but that doesn't mean it's actually doing that. Check your 5V modulation signal. It's probably fine, but check it anyway.

Click to expand...

Just checked it, and like you said, it's fine.

diachi said:

To reiterate on what Cyp said, Vf decreases as temperature of the PN junction increases. Even if you use diodes in a TO-220 package bolted to a heatsink, the junction temperature will start to increase as soon as you start to pass larger amounts of current through it.

Click to expand...

Yes but I should have at least have different voltage change, right ?
But it's the exact same one, 4.9V to 4.7V.
And the 1N4007 diodes were getting actually pretty hot after a long time, but the voltage stayed at 4.7V
And some times (rarely) I even measured a very constant voltage at 5.2V (without changing anything in the setup).
I'm not saying you're wrong at all, I just want to understand.

diachi said:

The datasheet for your diode should specify a thermal resistance for the PN junction to case, measured in °C/W (Temperature difference per watt of heat dissipation) most likely.

Click to expand...

Yep, I just never really paid any attention to it, now I will :thanks:

Benm said:

As for the test load: that's really no good for anything near 5 amps. This 1n4007 diodes take 1 amp continously, at 5 they will overheat to the point of going short or open circuit, i've even seem them desolder themselves from circuit boards (and even surviving that).

As in the 1 ohm, 5 watt resistor the problem is a lot bigger: if you ran 5 amps through that, it would dissipate 25 watts.

Click to expand...

Like I said earlier, I tried with others diodes, I just didn't take any photos. And for the resistor, I know that but it's the lowest one I have that can handle at least 5W.
And I'm running my tests at about 1.2A.
I ordered some 0.1ohm resistors.

Benm said:

Oh and a PS: i can't tell from the picture, but make sure the resistor is a film type, not a (more common) wire would one since high the inducance of those can cause serious problems for drivers.

Click to expand...

Oh I'm so dumb! How couldn't I have thought of that?!
It's a wire wound one