For your convenience:
https://oshpark.com/shared_projects/uWxyBEL1
http://www.zeilhofer.co.at/wiki/doku.php?id=uln-power-supply&s
Comparison of LT3042 chip PSRR compared to a former "next-big-thing" chip
:
http://www.diyaudio.com/forums/digi...-noise-regulator-dac-clock-2.html#post4996352
https://oshpark.com/shared_projects/uWxyBEL1
http://www.zeilhofer.co.at/wiki/doku.php?id=uln-power-supply&s
Comparison of LT3042 chip PSRR compared to a former "next-big-thing" chip
:http://www.diyaudio.com/forums/digi...-noise-regulator-dac-clock-2.html#post4996352
Last edited:
No, why would I ? Finished project. This LT3042 design possibly is a better PSU than the old but quite excellent performing LM723 design. Noise specs were not too shabby with that one but it was (FWIW) far from an LDO design. Most LDO's had worse specs than LM723 then. I encourage anyone willing to build one to try it out with a Subbu DAC. I was not even thinking about Subbu DAC to be honest as it has 4 onboard regs and the PSU was more like a well specced rectifier/prereg. Strange that you mention this as you were one of the very few that was not satisfied with V3. The only one to be precise
I don't see 0.8 uV noise LDO regs too often ! Seems like a worthwhile and affordable ultra low noise PSU for various purposes so I ordered some PCB's. Difference in specs with TPS7A4700 is spectacular IMHO. Here is a chance to build a true DIY non commercial ultra low noise board.
I don't see 0.8 uV noise LDO regs too often ! Seems like a worthwhile and affordable ultra low noise PSU for various purposes so I ordered some PCB's. Difference in specs with TPS7A4700 is spectacular IMHO. Here is a chance to build a true DIY non commercial ultra low noise board.
Last edited:
I wish the designer had pumped a dozen milliamps through the 2-diode "ground offset" mechanism in the preregulator. Get those diodes into a regime where their dynamic impedance is LOW. Here I show a JFET CCS doing the dirty work. You'll note I eschew the Walt Jung connection where the regulator pulls itself up by its bootstraps.
_
_
Attachments
Or just leave out the prereg and use lower input voltages by choosing a right transformer like a 5V AC type. With a carefully chosen transformer losses can be reasonable. Only when using let's say a 12 V transformer for 5V output the prereg can be necessary depending on current. In my book choosing a 12V transformer with an ultra low drop 5V LDO is somewhat off From the Vi to the Vo pins a coil can be soldered instead of the 7805.
Or use jumpers for the diodes and use a 7806.
The prereg only seems to be there for versatility purposes. In the words of the author: 78xx as a pre-regulator, to reduce power loss on the LT3042 and to increase the max. input voltage.
From the Vi to the Vo pins a coil can be soldered instead of the 7805. Or use jumpers for the diodes and use a 7806.
The prereg only seems to be there for versatility purposes. In the words of the author: 78xx as a pre-regulator, to reduce power loss on the LT3042 and to increase the max. input voltage.
Last edited:
J.Paul.... Subbu was mentioned because i know hot plays. And if you will described difference even as preregulator will show me (other members that are interested in lt3042) some picture of reg. Like subbu other dacs also have more regs near dac ic. Because of the size of this pcb i have not thought to use it in that position. If you tell us that you didn"t hear difference it is ok. That mean that use of lt3042 at that place will not bring significant improvement. I know that you were not satisfied with my opinion about Subbu. I wrote a few times that subbu is ok but nothing special compared to serious dacs. It is wery good price performance dac. Let's go with lt3042...
That is something I haven't said and can not say as I haven't tried this board yet. I just gave a link to affordable PCB's for a reg superior to TPS7A4700 (on paper). It may perform way better than commercial offerings so it may be a serious contender. This is a board for those that like to try stuff out and not specifically for those that need to hear from others which boards are good.
Anyone willing to try this specific board with Subbu can do so. The original board already was very good performing. I am encasing my very last V3 and maybe I will try the LT3042 too. A serious PSU for a serious DAC
It does not matter. It would have mattered if I had several complaints but I haven't. Dust in the wind.
Last edited:
I'd be in if the boards came with the regulator already soldered in....with some 'easy' way to adjust the op voltage. Here's one that i'm thinking of ordering but there's no info on how they implement the op voltage selection: https://www.aliexpress.com/item/Dua...lgo_pvid=efcb5161-89a8-4f75-93d5-0a1b9c2307d3
I built a few regs with two parallel Lt3042 as mentioned in the application information of the lt3042 datasheet. It reduces the noise to 0.8/sqrt(2) uV and doubles the max. output current to 400mA.
You can scale it up easily by using more devices in parallel.
http://cds.linear.com/docs/en/datasheet/3042fa.pdf
You can scale it up easily by using more devices in parallel.
http://cds.linear.com/docs/en/datasheet/3042fa.pdf
Attachments
Ah I got the boards and they look nice but solder pads are small (rectifier and caps). I thought of using LT3045 (500 mA) as it is pin compatible. With such low noise levels I aks myself if it really matters to make noise even lower
Point is that I haven't found the chips at my local distributors yet.
Point is that I haven't found the chips at my local distributors yet.
Ah I got the boards and they look nice but solder pads are small (rectifier and caps). I thought of using LT3045 (500 mA) as it is pin compatible. With such low noise levels I ask myself if it really matters to make noise even lower
Point is that I haven't found the chips at my local distributors yet.
Yes just lowering the "general"or integrated noise doesn't really make sense. But the noise and psrr is also current depending, see figures G43 and G50 in the datasheet (eg psrr >20db increase@10khz)... And maybe it is just for my "gut feeling" but I want to keep noise and psrr in the hf region very low if I m feeding a DAC with it.
Like you I'm very pleased by the specs of this device and compared to the TPS7A4700 it is clearly superior....edit: and the datasheet with all informations is really well made. It covers all questions and it is worth reading it completely - especially when it comes to stability C_set and C_out parameters - ESR and ESL(!) must be considered (I had difficulties to find the a cap which fulfils the requirements).
I had the same problem as you - couldn't find them. So I asked LT for samples.
Last edited:
I find this site to be really useful when it comes to part availability: https://octopart.com/lt3042emse%23pbf-linear+technology-57149843
As you're in the Netherlands, it might be easiest to order from Farnell or Digikey.
How did you lay out the Cset/Rset on the PCB to minimize parasitics/noise? I find that there's a lot of layout guidance w.r.t. using a single regulator, but for multiple ones much less so.
Since a few years I'm using "deadbugstyle" to build my devices. Not very beautiful - looks like a birdnest. For wiring I used thin copper strands from litzwire or coppertape insulated with kaptontape.
Of course short traces close as possible to the IC. I tried to build it symmetrically according to Cset/Rset.
I will post a picture next time.
is Farnell providing parts to hobbyists again? A few years ago they told me that I have to order it from an official distribution shop https://hbe-shop.de/. But the shop has only a restricted range of parts - usually this is the reason why you want to order parts at farnell, mouser etc...I was very disappointed by their politics so that I ve never ordered parts.
edit https://hbe-shop.de/Art-2475651-LINEAR-TECHNOLOGY-LT3042EDDPBF-LDO-REGLER-EINST-02A-0-15V-DFN-10 6,46€ dont know how much LT charges for the order but maybe it would be cheaper to order directly from LT lt3042 around $3.22 ...sure without duty, tax etc.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.