| Electronics > Repair |
| 18V/3A HY1503 PSU - weird CC/CV behavior while charging battery cells |
| << < (3/15) > >> |
| iMo:
FYI - I've tried to do a stability analysis - the FRA in LTspice24. I've been using 2 different transistors as I have not imported the old libraries yet (I added AB's lib only). Also a green led instead the 2.4V zener. The phase margin of CV is from 60 to 100deg for lowest to highest output voltages with 10ohm load (and CC not acting) and C11 (the C4 in the sim) 10n to 1u. |
| xavier60:
Load transient tests are meaningful to me. I don't think R19 and C2 are proportioned properly. C2 should be made larger or R19 should be made larger. Actually, making C2 large only affects the settling time, not high frequency stability and transient response. |
| iMo:
--- Quote from: xavier60 on September 18, 2024, 08:37:12 am ---..C2 should be made larger or R19 should be made larger. Actually, making C2 large only affects the settling time, not high frequency stability and transient response. --- End quote --- Yep, I can see in the sim C2 affects the settling with pulsing load. What do you mean by the "..C2 should be made larger or R19 should be made larger.." ? Could you point me to a formula or an approach how to design/balance it? PS: I know it has been discussed for years already here, but perhaps you have the info handy.. |
| xavier60:
--- Quote from: iMo on September 18, 2024, 09:03:51 am --- --- Quote from: xavier60 on September 18, 2024, 08:37:12 am ---..C2 should be made larger or R19 should be made larger. Actually, making C2 large only affects the settling time, not high frequency stability and transient response. --- End quote --- Yep, I can see in the sim C2 affects the settling with pulsing load. What do you mean by the "..C2 should be made larger or R19 should be made larger.." ? Could you point me to a formula or an approach how to design/balance it? PS: I know it has been discussed for years already here, but perhaps you have the info handy.. --- End quote --- I dont understand the theory too well, just experience from experimenting. The value of C2 seems too small for R19 being 1K. Make C2 0.1uF. Then confirm that the CV loop is stable but likely with slow transient response. Increase R19 to improve the transient response while maintaining good stability. Then decrease C2 to improve settling time without causing instability. |
| iMo:
What I remember from somewhere the RC should introduce a zero well below gain BW of the opamp. Like GBW=1MHz with the 741. The DC gain is set to 1 by the 10k/10k (so I had to use R=10k let say, with my 1k the gain will be 0.1 when talking inverting amp). The transistors have got none gain there. The capacitor is calculated from f_zero=1/(2*pi*R*C), where the f_zero should be say 10% of the gain BW, for example. Thus 100kHz for example. So the C = 1/(6.3*100k*10k)= aprox 150pF. Or something like that? |
| Navigation |
| Message Index |
| Next page |
| Previous page |