Mark -- will look at it on the bench this afternoon. Once you start getting into the weedy part of PSRR layout seems to become a critical issue.
Mark,
You are speaking of spice results. If the spice results are 20% better than tested results then the spice circuit may have some merit.
While you are at it please run a simulation of the bootstrapped 317 and 337 pre-regulated Jung/Didden regulators.
My suspicion is that the simulation results will be somewhere near the 20% delta compared to the published Jung/Didden regulators results.
DT
You are speaking of spice results. If the spice results are 20% better than tested results then the spice circuit may have some merit.
While you are at it please run a simulation of the bootstrapped 317 and 337 pre-regulated Jung/Didden regulators.
My suspicion is that the simulation results will be somewhere near the 20% delta compared to the published Jung/Didden regulators results.
DT
Naaw. If it's important, you'll do it yourself. If it's not important, you won't.
If it's too difficult or too time consuming, that means: it's not important. If you can't be bothered to learn how to create and run simulations, that means: it's not important. If you are waaay too "over subscribed", if your normal day requires "multi tasking" far too deeply to even consider creating a simulation, that means: it's not important.
How important is it?
If it's too difficult or too time consuming, that means: it's not important. If you can't be bothered to learn how to create and run simulations, that means: it's not important. If you are waaay too "over subscribed", if your normal day requires "multi tasking" far too deeply to even consider creating a simulation, that means: it's not important.
How important is it?
How important is it? Not so much.
Mark,
Thanks for the clever reply.
I thought that you could be persuaded to help make the point, you are a serious fellow.
Oversubscribed, not so much, I have been selling the peaks.
I am getting the idea that simulation can/does miss a lot. I recently put a new analyzer on the bench and have been testing the homemade creations from over the years.
It is amazing how the transformer/rectifier grunge and IM products get into everything.
If we were to specify minimum regulator performance, I am not sure what high frequency rejection would look like. In terms of the analyzer what weighting curve will we end up using? I am thinking about it.
I have not put LTspice on the new Intel coffee lake computer yet.
What is next; finish the Borbely FET line stage and roll in a few different regulators including the Silent switcher and see what the graphs look like.
DT
Mark,
Thanks for the clever reply.
I thought that you could be persuaded to help make the point, you are a serious fellow.
Oversubscribed, not so much, I have been selling the peaks.
I am getting the idea that simulation can/does miss a lot. I recently put a new analyzer on the bench and have been testing the homemade creations from over the years.
It is amazing how the transformer/rectifier grunge and IM products get into everything.
If we were to specify minimum regulator performance, I am not sure what high frequency rejection would look like. In terms of the analyzer what weighting curve will we end up using? I am thinking about it.
I have not put LTspice on the new Intel coffee lake computer yet.
What is next; finish the Borbely FET line stage and roll in a few different regulators including the Silent switcher and see what the graphs look like.
DT
Good. Maybe you might want to investigate Omicron's "Bode 100" (web link) test equipment. A couple DIYA members have one, and use it to extract the loop gain and phase plots of real hardware (amplifiers, power supplies, test gear, etc). Occasionally they upload the Bode 100 plots here. Looks like slick gear, and sidesteps many of the doubts & worries about simulations.
I found an old omicron application note on how to use it as an S-Parameter analyzer.
Attachments
Last edited:
I've bought me a Red Pitaya for XMAS!
14 Bit, 125 MSPS, 2 ADCs, 2 DACs, network, USB, Linux, open source, cheap.
< Red Pitaya >
14 Bit, 125 MSPS, 2 ADCs, 2 DACs, network, USB, Linux, open source, cheap.
< Red Pitaya >
Gerhard, I have been ogling that also, Elektor had a deal on it recently.
Did you check out the measurement software - is it ready to use?
Jan
Did you check out the measurement software - is it ready to use?
Jan
I have tried the scope, that seems to work; otherwise I'm more
interested to use it as a software define radio. But I have put
it on hold, I must finish those low noise preamps first. I tend to
have more projects in parallel than I can handle.
There is a picture in the thread "No RF Gear here?" in the lounge.
interested to use it as a software define radio. But I have put
it on hold, I must finish those low noise preamps first. I tend to
have more projects in parallel than I can handle.
There is a picture in the thread "No RF Gear here?" in the lounge.
Pass ba-3 line stage
is there a way to use this board for the ba3 as line stage power supply ie, with mods to bring the output to 24/-24v ? Thanks.l
is there a way to use this board for the ba3 as line stage power supply ie, with mods to bring the output to 24/-24v ? Thanks.l
Jan, I noticed in one of your replies in another PS thread, you stated that the 120uF caps in the "Super Regulator" should not have low ESR.
First...120uF caps aren't as common as 100uF caps.
Second...after doing some searches at Mouser, I came up with these 120uF caps which have a ESR of 514mOhms:
860020675017 Wurth Electronics | Mouser
I've never heard of or purchased Wurth caps before, but do you think they will work okay in the "Super Regulator"?
Thanks.
First...120uF caps aren't as common as 100uF caps.
Second...after doing some searches at Mouser, I came up with these 120uF caps which have a ESR of 514mOhms:
860020675017 Wurth Electronics | Mouser
I've never heard of or purchased Wurth caps before, but do you think they will work okay in the "Super Regulator"?
Thanks.
I have a bandelier of Panasonic FC -- 120uF/25V -- 200 pieces @ $0.40 -- how many do you need. NOS from Mouser.
The first electrolytics specified by WJ in the 1995 articles were quickly obsoleted. The Panasonic FC has also gone unobtanium in 25V, but generally available in 35V at the usual suspects.
The first electrolytics specified by WJ in the 1995 articles were quickly obsoleted. The Panasonic FC has also gone unobtanium in 25V, but generally available in 35V at the usual suspects.
I certainly don't need 200 pieces, but it's good to know that you have a good supply.
My understanding is that the Panasonic FC, FM and FR caps are all low ESR caps.
Are you saying that low ESR caps will work just as well as higher ESR caps in the SR?
The low ESR comment was for the cap on the output. This cap plays a role in the stability and should not be low ESR like some boutique film caps. This is most easily fulfilled by using any cheap electrolytic at that position.
The other electrolytics can be good quality and low ESR if preferred, but will have very little if any impact on performance.
Jan
I'm about to order parts for a +15v super regulator. Looking at the schematic at the Diyaudio store I calculated R6 from the formula R6 = (Vout*R7/Vref)-(R7+R16). So If R7 is 1k ohm as in the schematic and Vref is 6.9v then R6 = (15*1/6.9)-1.1 or a 1.07k ohm resistor. Does this look about right?