With schematic in the post #1092 the load regulation is lost when compared to the positive regulator. From this schematic load regulation is better by moving the lower resistor of the resistor divider to the anode of the TL431 ( instead of the PSU output ). This is proved in the posts where I give the equations for the output voltages in these different solutions, post #1093 #1094. Why it is better is clear in the equations.
Moving this resistor gives no loss about the benefits from the denoiser.
The current mirror has the potential to give a negative regulator that has exactly the same behavior as the positive regulator. This is true assuming an ideal current mirror. This result is confirmed in simulation using a 2 transistors current mirror, where I get very near the same results about load and voltage regulation, PSRR, Zout, noise.
Stability from simulations is highly questionable, whatever solution. Fiddling with component values to get stability can be misleading.
Last edited:
Had some PCBs made using tombo56's gerbers from post #939:
https://www.diyaudio.com/forums/pow...fit-upgrade-317-based-reg-19.html#post6255032
Will show progress soon.
https://www.diyaudio.com/forums/pow...fit-upgrade-317-based-reg-19.html#post6255032
Will show progress soon.
Last edited:
Please give me an image of the corrected circuit. Presently with my smartphone, I don't have the simulator to load the .asc.
Ok. I like my old notebook so much with WinXP 
The good about XP is no upgrade. Finally a stable debugged Microsoft OS.
You are using the TL431 as a zener in both regulators.
Then load and voltage regulation is governed by the Vbe + 10 ohm of the driver transistor. Why you added this 10 ohm ? ( re is proportional to Ic, 26 ohm at Ic = 1mA).
Have you looked at the max current and power at the TL431 that occurs in the case of: No load on the PSU and max mains voltage ( +10% ). Life is easier for the TL431 at max load that gives a max of unregulated voltage droop.
You added an inductor at the transistor of the denoiser, a bead I presume, isn't redundant with the compensation at the collector, do they add ?
The good about XP is no upgrade. Finally a stable debugged Microsoft OS.
You are using the TL431 as a zener in both regulators.
Then load and voltage regulation is governed by the Vbe + 10 ohm of the driver transistor. Why you added this 10 ohm ? ( re is proportional to Ic, 26 ohm at Ic = 1mA).
Have you looked at the max current and power at the TL431 that occurs in the case of: No load on the PSU and max mains voltage ( +10% ). Life is easier for the TL431 at max load that gives a max of unregulated voltage droop.
You added an inductor at the transistor of the denoiser, a bead I presume, isn't redundant with the compensation at the collector, do they add ?
If you change the "LM" to "LT" it becomes easy. LT3042/45 or LT3094 with external BJT achieves (or surpasses) all the above goals
LT3094 is -20V 500mA.If you change the "LM" to "LT" it becomes easy. LT3042/45 or LT3094 with external BJT achieves (or surpasses) all the above goals
How do you fit a BJT for more output current.
I did not find how in the datasheet.
How about voltage ? I need 32V output.
LT3094 needs an output capacitor > 10uF with ESR< 0.03 Ohm and ESL<1.5nH.
Last edited:
Thanks for thorough response bohrok.
Of course to no one's surprise my first attempts didn't fly. 24VIN only seeing 1.29VOUT. I did throw in a fresh LM317 in case I ESD'd the first one during construction.
I take it this is the LM317 keeping VOUT above the ADJ pin which does show 0V.
Is this what it looks like when Q1 is dead? I might pop in a fresh MPSA06 see if that works quick and easy before I dive into the mess on the bottom side of the PCB.
Only other thing I can see (of course I don't see very well) is the 5K trimpot at R2 which I connected wiper to ADJ, one of the other pins to GND and the other NC. Not sure if this is correct, pots are my archnemeses.
Last edited by a moderator:
You could check the datasheet of LT3045/LT3042 and also conclude that nothing is written about using pass transistors.
But you can read it here:
Increasing Output Current of the Ultralow Noise, Ultrahigh PSRR LT3042 200mA Linear Regulator | Analog Devices
It seems there is hardly any competition left for regulators with a maximum output voltage of 20V with regards to noise. The slightly less high performance with pass transistor still is way better than any of the old dinosaur regs.
I have built several PSU's for DACs and streamers with single LT3045, 3 in parallel and some with pass transistor and there really is no valid reason to not use LT3045 (or TPS7A4700). If you want performance that is. The circuit of Elvee is meant to upgrade existing designs. In new designs LM317 should be avoided. It is probably easy of use and nostalgia that it is carved in the minds of audio guys. Make the switch as it is an easy upgrade to better audio performance!
IMO the largest drawback of the pass transistor is the dropout voltage now being 1.5V. Last week this got me into trouble with a LT3045 5V PSU powered by a 6V transformer.
Last edited:
Jean Paul - yes you’re correct that the ‘competition’ in the 20 v or lower area has been pretty much run to completion. And we have good articles how to boost output currents too.
The question is if there is (similarly) good performance available for higher voltage (say 36v or 48v) ? Anyone any suggestions ?
The question is if there is (similarly) good performance available for higher voltage (say 36v or 48v) ? Anyone any suggestions ?
Maybe you should debug by first setting up a regular LM317 PS without denoiser. Just connect C3 negative terminal to ground.
If you change the "LM" to "LT" it becomes easy. LT3042/45 or LT3094 with external BJT achieves (or surpasses) all the above goals
Almost, but not quite: 40V input and I don’t think it would play nice with several 10.000 uF low ESR capacitors at the output, as is usually found in the regular CRC power supply of a class A amplifier.
I find this as interesting development challenge and entertainment (popcorn not included), that eventually could be used in my amplifier builds.
Finally, how would LT regulator answer the concerns expressed in the neighboring thread?
Are you really fine with IC voltage regulators ?
It helps prevent oscillation and ringging if regulator is "smacked" with high and fast transient load.
It works together with RC compensation in the same direction achieving slightly better result.