This sounds quite bizarre, I'm fairly sure I would have put some minimum and maximum limits on values to prevent "issues" so it's quite odd that negative values can go in. Could you please let me have the following:

• Reply to the email with a screen dump, or the .psu file so I can have a look
• Details of the build number from Help | Abpout
• Details of the OS you're using

In the interim, I'll have a look at this end to run some checks. Thanks very much.

Regards,
Duncan

I had a sim would not resolve. "Timestamp has fallen..."

On re-entering I realized I had entered C1 as "-0.5u"... negative!

The font used on my installation makes this very hard to see.

While a negative capacitance has some meaning in theory, in context of PSDU users this is almost certainly a finger-fumble.

I dunno if it is worth adding a sanity/censor-check to refuse negative caps. I have been using PSDU for a decade+ and I think this is the first time I made this screw-up.

Hmmm, negative resistor---- ah, it accepts the input then blows-up at 5,102V. Which is probably correct: the neg-resistor is powering the circuit, just as neg-R repeaters power-over telephone line loss.

Thank you Duncan and Paul for the guidance. I now downloaded the latest build 76 from the io Files folder.

Regards
Luis

On 11/3/2022 4:08 PM, Greenpitu@... wrote:

I am getting this error when trying to download the official windows executable from

Isn't that an old version?

The current versions seem to be on groups.io under the 'Files' tab.

/g/duncanampspsud/files/psud2_windows

There appear to be some issues with the server, I've just been on chat to support.?

The files stored at??were just legacy files anyway and are way out of date so I've updated the page to reflect this.

All downloads can now be found at the Files area on this groups.io page which is on the left hand bare if you are accessing this via the web, or at the following link if you are reading this by email:?/g/duncanampspsud/files

Regards,
Duncan

Thanks for the help so much.

So the question I am unsure of is what makes a good transient response compared to another, I sort of felt like the speed (tightness of the curve) to return to the steady state is desirable to be as rapid as possible; within the context of acceptable ripple and the lack of ringing / misshape of the profile curve or the output voltage. Is this correct and should I also look for ringing in other areas of the circuits?

This is the PS circuit I am trying to model, with approximately the design I am moving towards btw. The 6c45 takes voltage from halfway up the doubler of 2 x 5u4GB rectifiers, and I have struggled a bit to have confidence this is modelled correctly.

Hello!

I am getting this error when trying to download the official windows executable from?

Anyone seen this one before?

Thanks!

Forbidden

You don't have permission to access this resource.

Additionally, a 403 Forbidden error was encountered while trying to use an ErrorDocument to handle the request.

You're doing the right thing by using the step change to model the transient behaviour.

If ringing at the transition point is a problem, I've found it helpful to make the first C of a CLC section a significantly lower value than the second C. This has the advantage of presenting a greater damping from the load on what effectively has become a tuned circuit.

The nice thing is you can try some values and check out the effect without having to assemble the components.

Regards,
Duncan

So I am considering updating some old capacitors in an existing power supply B+ on a driver valve 6c45 and B+ on an output valve 211 in a single ended power amplifier. The existing design uses CLCLC for 6c45 valve and CLCLCLC for the 211.?

I am considering changing both of these for CLC to simplify the potential resonances/ringing in the Power supply, based upon some advice from another DIY builder, and to make them a little more responsive.

I have had a number of conversations on DIY AUDIO and the question I wanted to post here is I hope quite simple, as I have a question related to PSUD2 in particular.

Apart from Voltage and ripple of the stabilised supply for a steady state design objective, how can the transient response or any other additional design objective be modelled, or simulated.

So far I have used the profile of the initial start up curve and put in a step change in a constant current load, and tried to make the response as rapid and oscillation free as possible?

Any guidance gratefully received.

On 7/29/2022 12:26 PM, Rulg wrote:

Spice LT ... i have to enter V * 1.414

Duncan PSUD and power transformer makers "assume" AC power is sinusoidal. (For several reasons, it always is, nearly.) And because the original load on AC power systems was light-bulbs and heaters (and motors, a special case), which hot-up and put-out according to the RMS of the wave, RMS is the conventional spec.

SPICE is an all-purpose equation solver and does not make this assumption for you. A full SPICE pack has square, triangle, and other waves as user-defined. The convention is to specify the "peak" because that's always clearly defined no matter how wonky the wave shape is.

the Duncan spice model literature for the 6SL7 and 6SN7 also give a maximum plate beyond the spec sheet figures.

Yes, the "6SN7GTB, ECC32, 6FQ7/6CG7, 5692" model:

--says 0..450V. I think some of these tubes are rated for 450V steady. And the 330V DC tubes will swing their plates above 330V in transformer-loaded circuits, with complete safety (because they spend the other half of the time below 330V). We need to model transient conditions. No, a tube does not blow-up the instant it sees more than the steady rating.

i can't afford to blow stuff up

I believe smoke is the best teacher. However too much smoke is bad; and tubes don't smoke when they die. But tubes do not die easy. Running tube plates 40% over rating is not uncommon. They make more trouble that way, and in 1950 RCA/GE would prefer to sell you a higher-priced tube with bigger numbers, but if you watch your voltmeter at power-up you can easily shut-down long before you kill a tube.

Electrolytic capacitors at over-voltage can die in a minute or so, AND they spit damp paper at the ceiling (or your face).

On PSUD2 you would enter the off-load RMS voltage of one leg, so for 800V CT that would be 400 (probably a bit more than that as PSUD expects the off-load voltage). It multiplies by 1.414 automatically for you to get the peak. You're right, SPICE based programs don't do this and you have to take the additional step of multiplying by SQRT(2).

You'd probably want the edit transformer properties to look something like this:



Note the [...] buttons at the right of the two fields above.

The first button at the end of the Value field brings up the off-load voltage calculator. This takes a guess at the off-load voltage given the transformer nominal RMS voltage, rated current, and regulation in %. Small transformers would be about 10%, bigger ones around 5% but the manufacturers data sheets should give some help with this. You could get something like the following for your 800V CT transformer:



The second button after the Source res field uses resistance measurements and real-world voltage measurements to dial in the transformer values. For secondary resistance, just measure one leg of the secondary - CT to outer. For example, if you measure a primary of 5 ohms and a secondary of 68 ohms on one leg, followed by voltage measurements of 121V AC on the primary and 418V RMS on one leg of the secondary you could end up with the following:



In this case, you would use a transformer voltage of 418V in PSUD with a source impedance of 127.7 ohms. It's an estimate, and gets messed up if you have windings for filaments etc., but it should put you in the right ball park.

Hope this helps.

Regards,
Duncan

Hello

Thanks to the help received from people (as an ex chemist I have some experience but zero training in electronics) I have been able to resolve a number of very basic MacBook-related PSUD2 issues - the double tap on the track pad still initiates annoyance though and hopefully there may be a way to remove the need in PSUD3?

As to PSU design... it's now a relatively easy task for me to "lay out a design" with nanovolts of variation at say 10 seconds after 9 second reporting delay ~ but in truth I find I have scant real innerstanding of what that really means.? Moreover those nanovolts are achieved with what I assume are huge amounts of inductance (20/40H) and large, expensive and normally non-stocked link capacitors of the sort that, if they are available, necessitate buying boxes of them from the US of A.? However, when I try to model published PSU designs (e.g. Jac van de Walle/JAC Music, Menno van der Veen, Claus Byrith etc) not only do they achieve their results using, in comparison to my efforts, fewer, simpler arrangements in effect more elegant, "rational" and "economical."?

If that were not enough I would appreciate a steer on whether incorporating Llundahl - type double winding chokes ~ is anything "mystical."?

Yep there is much to learn.? Thanks for the forum's continued understanding and guidance.? Is a rational guide possible if not AI optimisation?!

Thanks, Now I get it

On 6/6/2022 7:35 AM, dadumda@... wrote:

I don't see any?graf.

On 6/6/2022 1:29 PM, Duncan Munro wrote:

Have you checked any of the boxes in the results area (like V(C1)) to
enable the chart line?

Like this (attached):

Hi Morten,

Have you checked any of the boxes in the results area (like V(C1)) to enable the chart line?

Regards,
Duncan

开云体育

Totally awesome. Great work Duncan!

开云体育

Outstanding! Nice work, Duncan.

JL

About time for an update on PSUD3, it's been a bit stop/start however the schematic editor and simulator is now done. The only major component to complete is the charting capabilities.

Rather than put a text narrative on here, I've done a quick video to show what's been happening so you can actually see the Development Alpha version being put through its paces:?

Regards,
Duncan

Virus-free.

Hi
I am trying to learn how to use the?PSUD2 for Mac.

very very nice, thank you so much for this great tool.