Re: How to run PSUD on macOS or Linux/ARM?
As Duncan posted here recently- get cheap Windows machine. It may not be the answer you wanted, but it may be expedient. me> I am seeing minimal-spec Windows machines under $300 I've bought 5 mini PCs since August, two cost real money, however three of them have been bargain basement "Celeron" type machines going for GBP ?149 with 500GB solid state drive and 16GB of DDR memory. At today's rates, that's USD $187 a piece pre-loaded with Windows 11 and they run great out of the box. For single-threaded stuff like PSUD they are still over 50% of the speed of a Core i9 machine costing 6 times the price. Computing has never been cheaper in real terms :)
Regards,
Duncan
|
Re: How to run PSUD on macOS or Linux/ARM?
Hi,
?
I'm afraid PSUD2 won't currently work on the M1 processor (or any other ARM derivative). I was looking at exploring ARM support for PSUD3, however it would be subject to some performance testing to see if ARM-HF is viable - PSUD is very math intensive, so it may end up being a non-starter.
?
Here's the current offering for PSUD2:
?
| OS |
Target |
Support |
| Windows |
32 bit or? 64 bit supporting 32 bit |
Supported |
| MacOS |
x86 based |
Deprecated |
?
And future plans for PSUD3:
?
| OS |
Target |
Support |
| Windows |
i386 or x86-64 |
Future planned |
| Linux |
i386 or x86-64 |
Future under investigation |
| Linux |
ARM-HF |
Possible depending on performance |
?
Regards,
Duncan
?
|
Re: How to run PSUD on macOS or Linux/ARM?
Parallels should work.? I used it for this purpose several years ago, though not since.?
Hello
?
Looking for ways to make PSUD work on my Apple computer (M1 chip). I am running macOS or Linux (in a virtual ARM machine).
?
Thanks for any hints or ideas!
|
How to run PSUD on macOS or Linux/ARM?
Hello
?
Looking for ways to make PSUD work on my Apple computer (M1 chip). I am running macOS or Linux (in a virtual ARM machine).
?
Thanks for any hints or ideas!
|
Re: Hello everyone. I'm looking to build a super smooth 5v linear supply
Hi Joe,
?
PSUD2 can be downloaded from the files area, however I don't think it will be much help to you as it does not assist with the design of regulated supplies.
?
You might get more help from a posting in the forum where I'm sure there will be people who have done this kind of thing before. Hope that helps.
?
Regards,
Duncan
?
?
?
|
Hello everyone. I'm looking to build a super smooth 5v linear supply
Hello,
Trying to dersign a circuit for a super regulated and smoothes 5v linear power supply for a DAC. Does anyone know where the PSUD2 download is located?
Thanks
Joe
|
Re: PSUD chart colours (was Adding up all the currents! I'm lost!)
I lost my lists of upgrades and possible improvements? (wish list)? for? "PSUD-3".....and that was years? ago.?
The colours were one item.
?
Here is some additional ideas.?
?
(A)? 3 phase rectification...with either 50 hz? or 60hz.? Typ, 6 x rectifier legs are used. The big advantage with 3 phase rectification is... the raw DC? is just 5.2% ripple....so most of the filtering is already done.? It makes the subsequent filtering an easier? task.? ?With a FWB / FWCT / FWD setup....on 60 hz, the raw DC has 67% ripple.? Ripple freq? is 2F? ?with single phase (100/120 hz)? ?vs? 6F? with 3 phase (300/360 hz).? With a FWB, on 50/60 hz single phase, caps get charged up...or....'topped up'? 100/120 times per second.? ?With 3 phase, and a FWB, caps get topped? 300/360 times per second.? ? The static and dynamic regulation is vastly improved with 3 phase power.
(B)? Resonant choke input.? This was used by a lot of RF amplifier manufacturers, and also commercial broadcast in the past.? From the output of the rectifier assy, B+ goes through the series choke, then the usual shunt filter cap.....the difference being..... there is an extra cap? ?wired in parallel? with the choke.? ?The choke is parallel resonated? at the ripple freq,? typ 100/120 hz....(when single phase 50/60 hz used).? ?Choke resonated at 300/360 hz? when 3 phase used.?
(C)? When a? C-L-C? filter is used...... it would be nice to have the option to parallel resonate the choke at the ripple freq...sorta like an elliptical filter.?
?
###? ?The drawback to resonating the choke in (B) and also (C) is.... it only provides additional attenuation at the ripple freq...and not the harmonics of the ripple freq.? ?I saw this on small commercial? 5 kw FM broadcast PA's.... (single phase 60 hz).? A resonant choke was used to kill ripple at 120 hz.? However, the 5 kw PA? did not meet the FCC......'AM noise spec' used for FM broadcast.? The fix was a 2nd LC filter (this time the choke was not resonated).?
(D)? ?PSUD-2? only depicts P-P? ripple at the ripple freq. What's needed is? P-P ripple at all the harmonics of the ripple freq.? IE:? ripple freq is 120 hz.? ?What is the ripple at 240/360/480 / 600 hz? etc.?
(E). Using the? stepped load function of? PSUD-2 works, but doesn't represent what really happens with a dynamic load, like with an audio amplifier or? SSB.? ?I need to be able to differentiate between a static load....and a dynamic load.? ? With a dynamic load, the average? current is low, while the peak current is high.? ?Duty cycle on SSB ( with a lot of speech processing)? is 50% or slightly less on the B+ end of things.? ? The peaks as seen on an RF monitor scope are extremely narrow.? ? Meanwhile, when using the PSUD-2 stepped load option, the best I can do is to toggle between normal idle current..... and the max peak current.? IE: idle current of say 100 ma for XXX msecs, then slam on the full 800 ma load.? ?In reality, on SSB, the duration of the 800 ma peaks is mere msecs.? With enough C in the HV supply, the B+? won't sag as much.?
(F)? on these? HV? B+ supplies,? typ a HV fuse is wired in series with the B+..... followed by a 25-50 ohm 'glitch' resistor.? ?If any mishap occurs, like an anode to grid arc, or B+ to chassis arc, the 50 ohm resistor will limit the fault current to a safe value.? Meanwhile the resulting fault current will open off the hv fuse.....and fast like < 1-2 msecs.? ?IE:? 2.5 kv / 50 ohms? = 50 amps of fault current.? ?50 amps of fault current will open off a 800 ma rated fast HV fuse in 1 msec every time.? The HV fuse and? resistor will? save the HV supply....and also the tube.? ? I install a 2nd HV fuse, this time between one leg of the plate xfmr secondary (either leg, but one leg only) and the input to the rectifier assy.? ?Any B+ to chassis arc, inboard of the 1st HV fuse....and the 2nd fuse? will open off asap.? ? The 800 ma through the 50 ohm resistor will of course cause a 40 Volt,? V drop in the output.? PSUD-2 does not factor in the additional V drop....... but I just subtract it from the final results..... which works, but real difficult to do when a dynamic load is applied.?
I had a few additional ideas from years ago, for enhancements to PSUD, but have lost the paperwork....and can't remember what I had written down.? ??
|
Re: PSUD chart colours (was Adding up all the currents! I'm lost!)
Thanks all for the kind comments, they are very much appreciated :)
?
Regards / 73,
Duncan
MM0KGK
?
|
Re: PSUD chart colours (was Adding up all the currents! I'm lost!)
ditto all below. keep up the good work!
Sent from Samsung Galaxy smartphone.
toggle quoted message
Show quoted text
-------- Original message -------- From: "Ian Eales via groups.io" <ieales@...> Date: 1/5/25 8:40 AM (GMT-08:00) Subject: Re: [duncanampspsud] PSUD chart colours (was Adding up all the currents! I'm lost!)
Me make three!
As a retired developer, I feel your pain at creating something
GREAT!
- Ian
On 1/5/2025 08:32, brian frisbie via
groups.io wrote:
Duncan,
I second Tom’s comment!
Just a fabulous tool with unparalleled customer support!!
Cheers
Brian in Seattle
?
Duncan,
Thanks for all your effort past, present and future.
Projects such as PSUD can take on a life of their own
and there is no way to satisfy everyone's desires
without infinite effort.
Your engaged responses to address the questions that
arise is commendable and, if fact, exceeds the customer
service provided by most commercial companies.
Cheers,
Tom
|
Re: PSUD chart colours (was Adding up all the currents! I'm lost!)
Me make three!
As a retired developer, I feel your pain at creating something
GREAT!
- Ian
On 1/5/2025 08:32, brian frisbie via
groups.io wrote:
toggle quoted message
Show quoted text
Duncan,
I second Tom’s comment!
Just a fabulous tool with unparalleled customer support!!
Cheers
Brian in Seattle
?
Duncan,
Thanks for all your effort past, present and future.
Projects such as PSUD can take on a life of their own
and there is no way to satisfy everyone's desires
without infinite effort.
Your engaged responses to address the questions that
arise is commendable and, if fact, exceeds the customer
service provided by most commercial companies.
Cheers,
Tom
|
Re: PSUD chart colours (was Adding up all the currents! I'm lost!)
Duncan,
I second Tom’s comment!
Just a fabulous tool with unparalleled customer support!!
Cheers
Brian in Seattle
toggle quoted message
Show quoted text
On Jan 5, 2025, at 7:50?AM, TJTomPhillips via groups.io <tjtomphillips@...> wrote:
?
Duncan,
Thanks for all your effort past, present and future.
Projects such as PSUD can take on a life of their own and there is no way to satisfy everyone's desires without infinite effort.
Your engaged responses to address the questions that arise is commendable and, if fact, exceeds the customer service provided by most commercial companies.
Cheers,
Tom
|
Re: PSUD chart colours (was Adding up all the currents! I'm lost!)
Duncan,
Thanks for all your effort past, present and future.
Projects such as PSUD can take on a life of their own and there is no way to satisfy everyone's desires without infinite effort.
Your engaged responses to address the questions that arise is commendable and, if fact, exceeds the customer service provided by most commercial companies.
Cheers,
Tom
|
PSUD chart colours (was Adding up all the currents! I'm lost!)
Thanks for the feedback.
?
As a user of PSUD myself :) I'm painfully aware that the yellow lines aren't helpful, nor is the random cycling. Both of these have been dialled out in the design for PSUD3.
?
While the colours are allocated in pseudo-random order to start with, the colours are darker and they will stick with the trace during the lifecycle of the file. There's a whole bunch of improvements around the charting which are:
?
- Colours are auto-allocated to start with as with PSUD2, but that's where the similarity ends
- Darker colours are used (pls see screen snip below)
- You can pick your own colours for each trace and these will not change from session to session, the choices are saved with the file
- There is an option for different line thicknesses
- There is an option for different line types
?
2/4/5 above will help with printing out on black and white printers - I have a cheap A4 laser printer that has really grown on me as it's convenient and cheap to run.
?
Great news; so where's PSUD3?
?
The problem I have is that the charting component that is used to display the different voltages and currents has some challenges which are holding up PSUD3 right now. The first is that line thicknesses greater than 1 slow the display down considerably, no idea why. And I means slow.
?
The second is that the X axis which shows the time seems to get cluttered up with a bunch of textual rubbish when you zoom right in on a long simulation - this needs to be fixed as it's unreadable.
?
My plan is to write my own chart component that is optimised for PSUD and will give a clean display and good performance. So... this was just to let you know what's happening and that all the feedback is being captured.
?
?
?
Regards,
Duncan
|
Re: Adding up all the currents! I'm lost!
I shouldn't think it too onerous to have a class table where voltages are one style, currents another. Resistor colors sans white could be used: v1 black, v2 red, v3 orange...
- Ian
Jan 4, 2025 15:30:38 Jim VE7RF via groups.io <jim.thom@...>:
toggle quoted message
Show quoted text
On Fri, Jan 3, 2025 at 06:18 PM, <orders2010@...> wrote:
> the other suggestion is on the colors of the graph lines. They change each time you run the simulation!
>It would be nice to make these static with color choices of your own. And I would eliminate YELLOW all together. >Its too hard to see!
?
?
## Yes, get rid of the yellow.? ?I have to keep clicking on one parameter...and it changes through every colour available...till I get the colour I want.? It's no fun when say? Voltage and current are the same colour.?
|
Re: Adding up all the currents! I'm lost!
On Fri, Jan 3, 2025 at 06:18 PM, <orders2010@...> wrote:
> the other suggestion is on the colors of the graph lines. They change each time you run the simulation!
>It would be nice to make these static with color choices of your own. And I would eliminate YELLOW all together. >Its too hard to see!
?
?
## Yes, get rid of the yellow.? ?I have to keep clicking on one parameter...and it changes through every colour available...till I get the colour I want.? It's no fun when say? Voltage and current are the same colour.?
|
Re: Adding up all the currents! I'm lost!
- Ian
On 1/4/2025 07:36, Chris via groups.io
wrote:
toggle quoted message
Show quoted text
Yup, tube
amps really are different from their younger cousins!
EL34world.com
is a great place to learn about this stuff.? Well informed
folks over there.? Also, is reputable.?
chris
thanks for the additional information.
I found the datasheet for the transformer, and collected
the DCR values and needed voltages.
I adjusted the PSUD2 app values as you suggested, and I
removed the zener (compensation) from the formula. Full PT
voltage is 325V.
?
Still does not make much sense to me! The numbers just do
not add up the way I would expect them to. My VC1 should be
a lot higher than 306Vdc.
For grins and giggles I pulled a schematic from a popular
fender 5E3 design site, along with data 9and design tips)
from The Valve Wizard, and input these values into PSUD2.
Even with this new information, the program did not generate
the numbers even close to the actual values that were on
this sample schematic I pulled!
?
Is VT1 RMS voltage of 287.6 the rectified voltage?? that
just seems real low. I think I was measuring way over 350V
on my bench test bed.
?
So at this point, I think I will just leave it as it is,
and finish building the rest of the amp circuit, and then
hook up this supply and carefully bring up the mains and
measure what I get. Adjust as needed to bring it into spec.
I think this Hammond PT, is only around 70mA. I think I
would need closer to 125mA, but you would think you could
control the amount of current being drawn, by lower plate
voltage and biasing the amp cooler! But perhaps I don't
understand it. I'm a solid state engineer.... tubes were
before my time!
What I do not yet understand about tube amps is how the
current changes as you increase signal, and how the OT
impedance affects the current load. So I am concerned that
this little PT will be way under spec, causing it to get
real hot, or even blowing it out. But I will try it anyway,
since it cost me zero $. If needed, I will buy a new
appropriate speced PT.
?
Side note for PSU2 design revision suggestions... It
would be convenient to show the wattage at each of the
nodes. the other suggestion is on the colors of the graph
lines. They change each time you run the simulation! It
would be nice to make these static with color choices of
your own. And I would eliminate YELLOW all together. Its too
hard to see!
?
Thanks.
?
|
Re: Adding up all the currents! I'm lost!
Yup, tube amps really are different from their younger cousins!
EL34world.com is a great place to learn about this stuff.? Well informed folks over there.? Also, is reputable.?
chris
thanks for the additional information.
I found the datasheet for the transformer, and collected the DCR values and needed voltages.
I adjusted the PSUD2 app values as you suggested, and I removed the zener (compensation) from the formula. Full PT voltage is 325V.
?
Still does not make much sense to me! The numbers just do not add up the way I would expect them to. My VC1 should be a lot higher than 306Vdc.
For grins and giggles I pulled a schematic from a popular fender 5E3 design site, along with data 9and design tips) from The Valve Wizard, and input these values into PSUD2. Even with this new information, the program did not generate the numbers even close to the actual values that were on this sample schematic I pulled!
?
Is VT1 RMS voltage of 287.6 the rectified voltage?? that just seems real low. I think I was measuring way over 350V on my bench test bed.
?
So at this point, I think I will just leave it as it is, and finish building the rest of the amp circuit, and then hook up this supply and carefully bring up the mains and measure what I get. Adjust as needed to bring it into spec. I think this Hammond PT, is only around 70mA. I think I would need closer to 125mA, but you would think you could control the amount of current being drawn, by lower plate voltage and biasing the amp cooler! But perhaps I don't understand it. I'm a solid state engineer.... tubes were before my time!
What I do not yet understand about tube amps is how the current changes as you increase signal, and how the OT impedance affects the current load. So I am concerned that this little PT will be way under spec, causing it to get real hot, or even blowing it out. But I will try it anyway, since it cost me zero $. If needed, I will buy a new appropriate speced PT.
?
Side note for PSU2 design revision suggestions... It would be convenient to show the wattage at each of the nodes. the other suggestion is on the colors of the graph lines. They change each time you run the simulation! It would be nice to make these static with color choices of your own. And I would eliminate YELLOW all together. Its too hard to see!
?
Thanks.
?
|
Re: Adding up all the currents! I'm lost!
thanks for the additional information.
I found the datasheet for the transformer, and collected the DCR values and needed voltages.
I adjusted the PSUD2 app values as you suggested, and I removed the zener (compensation) from the formula. Full PT voltage is 325V.
?
Still does not make much sense to me! The numbers just do not add up the way I would expect them to. My VC1 should be a lot higher than 306Vdc.
For grins and giggles I pulled a schematic from a popular fender 5E3 design site, along with data 9and design tips) from The Valve Wizard, and input these values into PSUD2. Even with this new information, the program did not generate the numbers even close to the actual values that were on this sample schematic I pulled!
?
Is VT1 RMS voltage of 287.6 the rectified voltage?? that just seems real low. I think I was measuring way over 350V on my bench test bed.
?
So at this point, I think I will just leave it as it is, and finish building the rest of the amp circuit, and then hook up this supply and carefully bring up the mains and measure what I get. Adjust as needed to bring it into spec. I think this Hammond PT, is only around 70mA. I think I would need closer to 125mA, but you would think you could control the amount of current being drawn, by lower plate voltage and biasing the amp cooler! But perhaps I don't understand it. I'm a solid state engineer.... tubes were before my time!
What I do not yet understand about tube amps is how the current changes as you increase signal, and how the OT impedance affects the current load. So I am concerned that this little PT will be way under spec, causing it to get real hot, or even blowing it out. But I will try it anyway, since it cost me zero $. If needed, I will buy a new appropriate speced PT.
?
Side note for PSU2 design revision suggestions... It would be convenient to show the wattage at each of the nodes. the other suggestion is on the colors of the graph lines. They change each time you run the simulation! It would be nice to make these static with color choices of your own. And I would eliminate YELLOW all together. Its too hard to see!
?
Thanks.
?
|
Re: Adding up all the currents! I'm lost!
As for the DC resistance on the transformer. you are correct, that number is wrong! I thought I had fixed that on PSUD! the data sheet on the PT shows it to be more like 277 ohms. I'll fix that.
?
?
Sounds good, please use the ohms from one leg only, not across the whole HV winding (it's to do with the way PSUD2 works). Also please factor into account the resistance of the primary to give an overall impedance, there's some details in the following text to help with this.
?
If it's possible to measure the resistance of one leg and the primary, and also measure the off-load voltage you can get a good approximation of what the source impedance should be with the Source Impedance Calculator built into PSUD2.
?
Right click on the transformer and select edit:
?
?
This will give you the Edit transformer properties dialog:
?
?
Click the [...] button shown and you will get the Source Impedance Calculator:
?
?
From here, you can enter the mains supply voltage and primary winding resistance, also the secondary voltage and resistance (one leg only). It will then figure out an estimate of the source impedance for you. Please be aware the off load voltage will be higher than the rated secondary voltage of the transformer; that's what it will provide when it's loaded down.
?
Regards,
Duncan
?
?
|
Re: Adding up all the currents! I'm lost!
Excellent suggestions! thanks!
I did not know that about the zeners. I have used them before on other projects, and felt like they worked okay. Other amp builders turned me on to the zener trick.
I looked up the specs on the tubes, and this is where I am coming up with my voltages and currents.
As for the DC resistance on the transformer. you are correct, that number is wrong! I thought I had fixed that on PSUD! the data sheet on the PT shows it to be more like 277 ohms. I'll fix that.
?
I will give this all a try and get back to you with my results!
?
DS
|
Paul Reid
Marc Chavez
Ian Eales