For voltages below 20V this does not make much sense. LT30xx with external BJT costs less than this and the performance is better.
That is something we don't know yet as it was only simulated till now. Try to get 5A or even just 3A out of LT304x with pass transistor. While LT304x are extremely good there is no equivalent for higher current (yet). I use LT304x with passion and I would like to see a higher rated version but alas.
BTW some new LT30xx products were launched recently.
BTW some new LT30xx products were launched recently.
Last edited:
Maybe, but in simulation LT30xx with external BJT works just as fine (or better) as this
Anyhow I have some doubts about the stability of this OPA551 denoised regulator. It definately requires a low-ESR capacitance at output and quite a lot of it. LT30xx is proven to be stable with 10uF or even lower.
Not maybe. You won't be able to get the desired higher current nor can you go higher than 22V input and 15V output and OP needs 24V as we can see. I posted the link to LT3042 application with pass transistors a while ago. Only way is to have many LT3045 in parallel. I have such a device with 6 x LT3045 but that is an expensive affair and cooling is an issue. You won't probably not believe it but I am measuring a LT3042 with pass transistor PSU today and I can not make it work to deliver 2.2A. Can attach a picture if you like.
What is the problem using a low ESR cap ? Any regulator has its peculiarities and most caps are low ESR anyway. Try LT3080 without even a power LED at the output for instance It is refreshing to see a newly developed design and a guy trying to make it work. I am sure he will also like to prove it is stable. LT304x are extremely good but that is not a reason not to design a discrete high power LDO for specific purposes. We should applaud any endeavour in this direction!
What is the problem using a low ESR cap ? Any regulator has its peculiarities and most caps are low ESR anyway. Try LT3080 without even a power LED at the output for instance
It is refreshing to see a newly developed design and a guy trying to make it work. I am sure he will also like to prove it is stable. LT304x are extremely good but that is not a reason not to design a discrete high power LDO for specific purposes. We should applaud any endeavour in this direction!
Last edited:
Ehm, have you beat me to build a prototype and check real conditions?
Simulation works stable with 100 uF only being smacked with 1,5 A constant load modulated with +-1 A load at 1 kHz or 10 kHz.
Last edited:
I was just referring to your question regarding 5V & 15V regulators. Not what OP was trying to achieve.
Ehm, have you beat me to build a prototype and check real conditions?
Simulation works stable with 100 uF only being smacked with 1,5 A constant load modulated with +-1 A load at 1 kHz or 10 kHz.
Probably depends on the simulation models. In my simulation it oscillates with e.g. 100uf/200mohm ESR at output. Adding series inductance as well seems to help but I am not sure how realistically that is modeled in Spice.
I have no plans to build this since I don't have a suitable use case. And I just recently ordered some LT3045s
I fully agree that having 100mOhm ESR or less at output is not a problem. I just wanted to inform other potential users of this. And also that ESR increases quite much during the life time of electrolytic capacitors. Typically specified end-of-life allows up to 300% increase.
BTW Even with 10mF/20mOhm at output it oscillates if you mimic "audio gurus" and add a smaller low ESR bypass (e.g. 1uF)
Ok, for this to happen you also need to add series inductance (e.g. 50nH) to the larger cap.
Hello,
I need clean 5V for AVCC for a DAC. Needs about 50-70mA. I am planning on doing +-15V with LM317 + denoiser for opamps, and I wondering on the best way to make 5V for AVCC of DAC. Would another lm317 + denoiser fed by +15V from the lm317+ denoiser output be a good way? If so, do I just use a 100nF input cap for the 5V LM317 on its input? Or would a 7805 + dienoiser be sufficient?
I need clean 5V for AVCC for a DAC. Needs about 50-70mA. I am planning on doing +-15V with LM317 + denoiser for opamps, and I wondering on the best way to make 5V for AVCC of DAC. Would another lm317 + denoiser fed by +15V from the lm317+ denoiser output be a good way? If so, do I just use a 100nF input cap for the 5V LM317 on its input? Or would a 7805 + dienoiser be sufficient?
78xx and LM317 are not recommended for new designs. For your low current purposes you could use LT3045/3094 for +/- 15V and a single LT3045 for 5V. Simpler, smaller, ultra low noise, LDO etc. Smaller is important as you don't want the stray in of HF/RF. All this if you want best results. If you want nostalgia you could use the ECC83 and 300B of regulators.
De denoizer circuit is meant to enhance existing LM317 designs.
De denoizer circuit is meant to enhance existing LM317 designs.
Last edited:
I agree that LT regulators should be used for DACs supply or AVCC. LM317 is “old horse”.
LM317 with denoiser or dienoiser is very good regarding noise, output impedance and PSRR. However, it’s an unanswered question is it good for the DAC supply. Apparently, many were built (some for use with DACs) but there is not a peep anywhere how do they perform. I have given for free a dozen of PCBs.
While by measurements, denoised regulator is extremely good, there is no warranty it will perform sonically as good as expected for the specific task.
In the DAC section, some knowledgeable members suggest that NRJ 78xx regulators are sonically suitable for the AVCC supply. Maybe, this would be a good route.
LM317 with denoiser or dienoiser is very good regarding noise, output impedance and PSRR. However, it’s an unanswered question is it good for the DAC supply. Apparently, many were built (some for use with DACs) but there is not a peep anywhere how do they perform. I have given for free a dozen of PCBs.
While by measurements, denoised regulator is extremely good, there is no warranty it will perform sonically as good as expected for the specific task.
In the DAC section, some knowledgeable members suggest that NRJ 78xx regulators are sonically suitable for the AVCC supply. Maybe, this would be a good route.
Even NJR 7805 are no competition for LT304x. Not even close. It is the regulator version of opamp rolling. So replacing an existing 7805 for a Onsemi or NJR 7805. Convenience versus results.
Strange thing is that in my experiments with TPS7A4700 it gives slightly better results than LT3045. Purely sonical.
Strange thing is that in my experiments with TPS7A4700 it gives slightly better results than LT3045. Purely sonical.
Last edited:
Is what you are saying valid for the analog or digital supplies?
I'm from Europe so I'm tied to Farnell or TME and neither have the lt3042/5 in stock. I already have a bunch of lm3x7/78xx and BJTs
Depends of course on the DAC chip. Datasheets may help. In some cases the supply name can be misleading: e.g. AVDD in some AKM DACs should be connected to digital supply.
I do not quite understand the need for PSU high performance above 100KHz ( even above 1MHz ), this is normally taken care by good 0.1μF at IC supply pins. Those are the really important ones.
I read a lot about low ESR, but no mention at what frequency.
Can one enlight me about best bypass caps to effectively perform above 100KHz.