New SE With Triode Mode KT120/150

[Cressy Snr]

I doubt that all of those caps are 2 ohms ESR .. in fact I'm not sure any would be quite that high.

That’s the default and I couldn’t be arsed to change them.

That's entirely up to you Steve, but the more accurate the information you give the chart the better it is for the end result.

It’s only a guide. This is after all, valves we’re talking about here. It’s not a fancy phono stage supply.

[pre65]

I looked at the data sheet for your capacitors, and they state low ESR, but don't give a figure.

I tend to use 0.5 ohm if I can't find an accurate figure.

[Cressy Snr]

I looked at the data sheet for your capacitors, and they state low ESR, but don't give a figure.

I tend to use 0.5 ohm if I can't find an accurate figure.

Well yes, but it’s still guesswork, so we are still not ‘accurate’ are we? The graph using 0.5 Ohm ESR, is no more representative of real life than one with 2 Ohms, given that we don’t know the ESR of the JJ caps in the first place?

And what about the Sprague Atoms’ in the pre/driver supply. Nothing for them either. So what are we supposed to do? Use the default value and don’t worry about it I say.

[Paul Barker]

Could get an esr meter For 15 quid odd. Or not bother with sims like me. Just shoot from the hip. If you must sim everything, at least put the right figures in.

[Paul Barker]

Though I have no right to say such like, since the Matrix wouldn’t have birthed if it weren’t for Stephies use of sims. Except that, no point me guilding the Lilly I couldn’t add anything to the Matrix signal circuit which was conceived by turning a failed forgotten technique only corrected by Sim use.

[Paul Barker]

Reminds me of the heresay about Henry Ford. He had various buttons on his desk one for each expert. He didn’t have to Sim anything, he pressed the appropriate button, which freed him up to do his role in the formula that brought success. You can’t do everything yourself anything like as well and you’d lose the overall plot lost in the detail.

Sim City on Intact Audio is the last public place Stephie communicates. If you’re into Sims don’t neglect that place.

I have no doubt clever people here sim their creations well because they sound well and solidly reliable. But a loaded gun in the hands of a child with guessed data.

[pre65]

I have no doubt clever people here sim their creations well because they sound well and solidly reliable. But a loaded gun in the hands of a child with guessed data.

I'd like to be able to use Spice, but my understanding at the time I last tried was insufficient. PSUD is (in my opinion) easier to use, and even when guesstimating certain parameters (like capacitor ESR) is reasonably accurate.

In this instance my use of the word "guesstimating" is indicative of something more accurate than a pure guess.

[Thermionic Idler]

I have no doubt clever people here sim their creations well because they sound well and solidly reliable. But a loaded gun in the hands of a child with guessed data.

I'd like to be able to use Spice, but my understanding at the time I last tried was insufficient. PSUD is (in my opinion) easier to use, and even when guesstimating certain parameters (like capacitor ESR) is reasonably accurate.

In this instance my use of the word "guesstimating" is indicative of something more accurate than a pure guess.

For my 300B project, I'm using a combination of LTSpice for the signal and PSUD for the power supply - basically because PSUD has proper built in transformer and rectifier models, whereas in Spice you pretty much have to build the transformer models from scratch, and the figures I got were so wildly out of whack with the expected ones that it was obvious I'd got something wrong.

[izzy wizzy]

For my 300B project, I'm using a combination of LTSpice for the signal and PSUD for the power supply - basically because PSUD has proper built in transformer and rectifier models, whereas in Spice you pretty much have to build the transformer models from scratch, and the figures I got were so wildly out of whack with the expected ones that it was obvious I'd got something wrong.

I do same but take the psu components a bit further back to the output of the rectifier or first choke of choke input esp in single ended circuits as there's more signal in the power supply than differential/PP.

[Cressy Snr]

Well, despite all this talk about capacitor ESR. I will still be using the default values. I see no reason to complicate matters.

ESR seems to be a minefield. The few examples I’ve seen specify ESR at 100KHz. An average 100uF electrolytic cap’s impedance at 100Hz seems to be around 16 Ohms. As ESR is an AC value then why can’t it be the same as the capacitor impedance. I suppose the “effective” bit of the ESR term is the fly in the ointment and that’s where I glaze over I’m afraid.

Resistance in AC terms IS impedance as far as I have been led to believe. But the usual “except on Tuesdays“ qualification probably comes in; specifically designed to to trap the people who think that maths is maths, trust the ‘experts’ learn how to work something out, feel quite pleased about it and then some wag tells them it either doesn’t apply, it’s not as simple as that or ‘you’ve missed a bit.’

That’s the frustration for someone like me, always was, and that’s why it took the head of maths at my old school, 3 months of private tuition and cramming every night for a full spring term, to get me a to a grade 6 Maths O Level. Basically by the time I reached the fifth form, (Y11 in new money) the head of maths, Mr Kitching discovered from his initial baselines, that I was barely above Year 6 level maths.

To get from there to O Level standard in 3 months was frankly an incredible achievement both on my part for sticking with it at 16 years old, and especially on Mr Kitching’s part. I owe that guy a lot for what he did for me, and my dad for forking out the money. OTOH it says a lot about the crap quality of some of the maths teaching and teachers, and the non-existent formative assessment, that was going on in the 60s and 70s.

Why my lack of progress wasn’t picked up earlier is frankly criminal IMO, but a certain standard of ‘intelligence’ was expected if you went to Grammar school and if you didn’t measure up, they didn’t want to know, and you were put in the thick stream, for maths and made to feel like a complete no-hoper. In fact it was the new, remarkably enthusiastic ‘thick stream’ maths teacher Mrs Griffin, who sounded the alert, as she had the nous to look beyond her own subject and couldn’t understand how a kid could be good at physics, but crap at maths.

Thinking back, the fact that I was able, under the radar, to get to the fifth form having made no maths progress, was definitely the reason why I became a top grade Primary Assessment Coordinator (outstanding from OFSTED) and went to other schools, training staff in assessment best practice.

Anyway, back on topic: if Duncan Munro chose 2 Ohms, it had to be for a reason and I’m not about to argue with him. I’m not qualified to dispute his values. Run as is, the software has got me a superb pair of power supplies, that don’t inject buzz, don’t hum and the sound is the best I’ve had to date. That’s enough for a complete thicko in the mathematical modelling field.

That I’ve got as far as I have since I started this hobby from scratch in 2005 is, in itself, remarkable.

[IslandPink]

In a film capacitor you can think of the ESR as being caused by the very thin metallised film on the plastic layers . You lose some power getting the charge in and out of those thin layers, just like going down thin wires to a speaker. There's something similar in an electrolytic but maybe to some extent it also includes the gunk between the layers.
I just use whatever I can find on a similar capacitor on a datasheet, I don't measure the things. This ESR is not the same as the AC impedance, it's more of a physical resistance due to the construction.

The only one on which it matters a bit is the first cap - the ESR there affects things like the peak rate of change of current and voltage going into the choke.

[Cressy Snr]

Thanks Mark. An explanation I can actually understand!
I was beginning to despair.

[pre65]

Thanks Mark. An explanation I can actually understand!

Yes, thanks Mark.

Motor run capacitors, which I have often used as a first capacitor (before choke) would seem to have an ESR of 0.5 ohm (or less). I hope, Steve, you have not taken any of this ESR discussion as being directed at you personally, more to help anyone who uses, or hopes to use PSUD in the future.

[Nick]

As Mark says, all passive components can be regarded as a network or L,R, and C

So you can consider a cap as a cap as a perfect cap in series with a perfect resistor and a perfect inductor (there is also a perfect resistor in parallel across the perfect cap, but for most caps this can be considered as infinite. The in series resistor is created by the lead out wired and the resistance of the plates and as its a perfect resistance its got no frequency dependent terms. The perfect inductor is again former d by the lead out wires and the shape of the plates. Its a perfect inductor so it is frequency dependent.

The ESR is just what it says, the effective series resistance and is not frequency dependent and is the value of the imaginary perfect resistor.

The Effective Series Inductance will create a frequency dependent impedance value, that while measured in ohms is not the same as resistance as it includes a phase angle.

And of course there will be all sorts of other parasitic components that can be drawn into that circuit as the cap is not floating in deep space so there are caps between the leads and the outside works making it more than a two terminal device. And the leads can be considered as transmission lines above a certain frequency, so bunched LCR networks, but that way lies madness.

With respect Steve, you seem to be tilting at windmills, reality is what it is, and its just lucky that we can model it with maths at all else we would still be in the dark ages. Its not as it is there to frustrate you, but you can chose to be happy with what works or worry over what you probably don't need to know. Most of us just try to find the level of interpretation that works for us and does what we need.

[Nick]

There's something similar in an electrolytic but maybe to some extent it also includes the gunk between the layers.

The liquid (electrolyte) between the plates are in theory just a conductive solution. The actual cap is formed by the very thin insulating layer created between one of the plates and the liquid by the polarising voltage.As you say, it will have some resistance.

And of course that effective resistance will start to dissipate power when AC current passes through the cap.

IMHO, for most valve power supply uses, 2R is fine. But when you are using it to model lower voltage higher current supplies, its better to get the correct values from the datasheet as the surrounding impedances will be much lower so the 2R value will start to matter more than if you are looking at 100ma and 500v