The simplistic Salas low voltage shunt regulator

Status
Not open for further replies.
Jump to Latest
#2,101
80VA is more than enough and if you mount vertically against face or back of box you can contain everything in one box. No hum issues.
Mind you that the French charge high cost freight. But you got to focus on making your phono working properly first and foremost, right? You don't shop for the cherry before making sure you got the cake recipe sorted out.
 
#2,108
Salas

First thanks for your design. I just built the symmetric +5,-5 board for the dac-end. I have a 12.5R Mills resistor on both sides for R1. After I dial in 5v with a 1K load I only get a drop of 1.44 volts (115mA) across R1 on the +5 side, lower than expected. On the negative side I get the closer to expected 2.2V drop across R1 (176 mA). On the positive side I am using IRFP9240's, on the negative side IRFP140's. I was pretty careful checking all of the transistors and leds, they should almost be matched. Any ideas?

Thanks
JimS
 
#2,109
jims said:
I have a 12.5R Mills resistor on both sides for R1. After I dial in 5v with a 1K load I only get a drop of 1.44 volts (115mA) across R1 on the +5 side, lower than expected. On the negative side I get the closer to expected 2.2V drop across R1 (176 mA).
Click to expand...
if equal or near equal current symmetry is required then one must calculate the resistor value after testing and measuring the completed regulator.

The variations in components, particularly the Vgs of the various FETs, causes voltage differences around the circuit that far outweigh the tolerance of +-10% resistors.
 
#2,110
Thanks Andrew

I do not need equal current symmetry. I was just worried that the big difference between the current across R1 meant the either I did something wrong, have a bad transistor, and/or my shunt was not going to perform well. If the differences are OK and are due to Vgs differences in the fets, thats OK.

Thanks again

JimS
 
#2,111
Just change your weak side R1 according to the drop across it by I equals V divided by R so to match your rails current potential. Not because you may need more than available but to have comparable Zo per rail. At so low Vout the Vgs differences play a bigger role. You can parallel some other resistor over your R1 and check it out until it picks up the circa 176mA to match. Looks like you need 8R2 total. But the Vgs can change a little as the Mosfet will run more current. You will see by measuring the drop again.
 
#2,116
Thanks, I think I will measure all of the voltages in the circuit when I get home from work, just to double check that I did not make a mistake. Then I will try changing resistors, although I might just seek out and replace the offending transistor.


BTW, I just had the Mills 5 W resistors leftover from another project.

THanks again

JimS
 
#2,117
Just fix the resistor value. Matching Vgs between PMos & NMos and between batches isn't easy luck, there are jigs for that when producing amps with multiples, it will not change anything here, take the easy way, tweak the associated resistor. And we don't know if it is just due to less Vf CCS Leds.
 
Status
Not open for further replies.
Top Bottom