New laser driver attempt
Added 2018-06-09 18:24:25 +0000 UTCMy last laser driver consisted of a high efficiency switching DC-DC converter 'to get close' and classic precise linear regulation to get the current precisely right. That works well and is safe, but not very energy efficient.
This bloody battlefield is the result of me trying to make the DC-DC converter behave like a stable constant current source to drive the laser much more efficiently. Happy to announce that the battle was won and the idea works!
It only needs a shunt resistor and a dual op amp: 1 op amp acts as a high side shunt amplifier, the other one compares the amplified shunt voltage with an analog input voltage and drives the DC-DC converters 'trim' pin accordingly.
In some cases the whole thing starts drawing tens of watts while putting out 0 ampere into the load, so some evaluation and fine tuning is still necessary ...
Comments
Cool chip, awesome website, will visit again!
Marco Reps
2018-06-09 22:17:39 +0000 UTCI used the 78S40 a while ago to make a constant-current switch regulator - it contains an additional uncommitted opamp. My application was step-up, but the chip works also for step-down. <a href="https://www.sciencetronics.com/greenphotons/?p=717" rel="nofollow noopener" target="_blank">https://www.sciencetronics.com/greenphotons/?p=717</a>
Uwe Zimmermann
2018-06-09 22:07:06 +0000 UTCNope, nothing measured yet, just the first sign of life! I also like the linear regulation better, but the laser guy mentioned something like battery powered, so I wanted to try and improve the efficiency. And as long as there are no transients or overshoots a laser for material cutting / engraving etc. can take a bit of ripple, right? The shunt amp output goes to the input voltage 'comparator' so I'd call that a large feedback loop involving everything: DC-DC, laser, shunt, ... I am not exactly sure about the terminology, bi-stable would be just on/off wouldn't it?
Marco Reps
2018-06-09 19:14:07 +0000 UTCHave you measured the ripple current in your constant current? A linear method would be alot cleaner. And, is your comparator without negative feedback and just bi-stable?
Faraz Sadrzadeh
2018-06-09 18:58:59 +0000 UTC
