| Electronics > Projects, Designs, and Technical Stuff |
| Evaluation of a 40V/5A DIY lab power supply |
| << < (3/5) > >> |
| David Hess:
--- Quote from: NANDBlog on August 21, 2019, 03:02:46 pm ---The IRFP150 has about 1 K/W Rthjc, meaning that with a 25 degrees heatsink, you have a 175 degrees junction with 150W. So 40V/5A is pushing the limit a bit. --- End quote --- Even if you can get a single transistor to handle the power dissipation, it can be advantageous to instead use several smaller ones just to spread the heat load out over a larger surface area rather than concentrating it in one spot. |
| xavier60:
The floating type design makes possible the best performance such as very low regulation dropout voltage and fast control of the output stage. Having everything referenced to the + output can be a little confusing I have found. Although it did result in easy off-line SMPS pre-regulation control with one of my projects. The link shows a - output referenced design. The BJT output stage is controlled by a voltage to current level shifter transistor. It has very stable control and is reasonably fast. It has 1V dropout. https://www.eevblog.com/forum/projects/linear-lab-power-supply/?action=dlattach;attach=751023 |
| tszaboo:
--- Quote from: David Hess on August 21, 2019, 03:55:35 pm --- --- Quote from: NANDBlog on August 21, 2019, 03:02:46 pm ---The IRFP150 has about 1 K/W Rthjc, meaning that with a 25 degrees heatsink, you have a 175 degrees junction with 150W. So 40V/5A is pushing the limit a bit. --- End quote --- Even if you can get a single transistor to handle the power dissipation, it can be advantageous to instead use several smaller ones just to spread the heat load out over a larger surface area rather than concentrating it in one spot. --- End quote --- What I wrote there, you have to read it this way: Not possible, not gonna work. |
| TheJC:
--- Quote from: David Hess on August 21, 2019, 02:49:31 pm --- --- Quote from: TheJC on August 21, 2019, 12:39:18 pm ---Do you guys have any more good papers about linear PSU design or at least some general rules? --- End quote --- Take a look at power supplies produced up until the 1990s by HP, Tektronix, Power Designs, Lambda, etc. They are fully documented and reverse engineering them from their service manuals can teach a lot. --- End quote --- Very good idea, I already took a look at some old HP/Agilent ones. Keysight has a good page with a lot of old HP/Agilent supplies and also their Service manuals. Unfortunately, the easy ones are only for lower voltages |O. But still, they are very good to learn. --- Quote from: Kleinstein on August 21, 2019, 02:58:58 pm ---To check the design it is a good idea to use a simulation, before actually solder. --- End quote --- Oh yeah I do that all the time, but sometimes you can't go around soldering or at least building it up on a breadboard. |
| TheJC:
--- Quote from: xavier60 on August 22, 2019, 12:28:49 am ---The floating type design makes possible the best performance such as very low regulation dropout voltage and fast control of the output stage. Having everything referenced to the + output can be a little confusing I have found. Although it did result in easy off-line SMPS pre-regulation control with one of my projects. The link shows a - output referenced design. The BJT output stage is controlled by a voltage to current level shifter transistor. It has very stable control and is reasonably fast. It has 1V dropout. https://www.eevblog.com/forum/projects/linear-lab-power-supply/?action=dlattach;attach=751023 --- End quote --- Would be definitely a better idea. Also, it would be then possible to implement David Hess idea of having the current feedback at the output level. I just tried to simulate the above circuit (only the CV part) and the output ripple seems to be massive. Idk if I build up wrong or its just the LTSpice simulation, I will have to take look tonight. |
| Navigation |
| Message Index |
| Next page |
| Previous page |