Nostr notes by Henning Paul DC4HP

really "brand new" or showroom model?

2026-04-15T05:12:57Z

In reply to nevent1q…gd40
_________________________

really "brand new" or showroom model?

Yes, just make sure to follow the layout guidelines that are ...

2025-11-13T20:11:09Z

In reply to nevent1q…0u8a
_________________________

Yes, just make sure to follow the layout guidelines that are given in the datasheet of the switcher.

3.3V to 1.2V is okay (0.63 watts), going straight from 5 to 1.2 ...

2025-11-13T20:08:22Z

In reply to nevent1q…6f72
_________________________

3.3V to 1.2V is okay (0.63 watts), going straight from 5 to 1.2 would have been problematic.

I have just skimmed the datasheet, but you should use MLCCs as ...

2025-11-13T19:49:31Z

In reply to nevent1q…3r3c
_________________________

I have just skimmed the datasheet, but you should use MLCCs as input and output capacitors (see chapters 8.2.2.2 and 8.2.2.3).

To be honest, the only valid application to use a high-current ...

2025-11-13T19:20:05Z

In reply to nevent1q…3xh6
_________________________

To be honest, the only valid application to use a high-current linear regulator in 2025 is when you need a low-ripple voltage for your analog circuitry. In that case you use a switching regulator to step down to Vout-Vdrop and put the LDO regulator behind that. By doing so you can reduce the voltage ripple of the step-down by the LDO's PSRR.

Disclaimer: I design (and more often review colleagues') analog electronics for a living.

Just don't pull that much current when the voltage delta is ...

2025-11-13T19:16:58Z

In reply to nevent1q…7v2s
_________________________

Just don't pull that much current when the voltage delta is that large...
Low drop regulators need 500mV, maybe 700mV drop to operate. So you add series power resistors to the input that at maximum current Vout+Vdrop is on the input pin. This doesn't improve efficiency, but spreads power dissipation among components.

If you look at the graph, (5-3.3)V*1A=1.7W is still out of spec, ...

2025-11-13T19:04:04Z

In reply to nevent1q…sjhg
_________________________

If you look at the graph, (5-3.3)V*1A=1.7W is still out of spec, and (5-1.2)V*0.3A=1.14W is only ok if the cooling pad is large enough. (Note that the graph assumes 2oz copper.)
Unfortunately for the 1117s the tab isn't GND, so you can't use the GND copper pour as heatsink.

Exactly.

2025-11-13T18:41:25Z

In reply to nevent1q…cfy6
_________________________

Exactly.

5.7 watts on the 3.3V regulator and 4*1.3=5.2 watts on the 5V ...

2025-11-13T18:36:20Z

In reply to nevent1q…uazz
_________________________

5.7 watts on the 3.3V regulator and 4*1.3=5.2 watts on the 5V regulator... Exceeds the maximum allowed power dissipation by far.

Do you plan to use a 9V power supply? Toasty!

2025-11-13T18:26:26Z

In reply to nevent1q…daae
_________________________

Do you plan to use a 9V power supply? Toasty!

Ah, like "Hengst"!

2025-11-04T18:58:24Z

In reply to nevent1q…fmwg
_________________________

Ah, like "Hengst"!

I wonder how Danish "hest" mixes into this.

2025-11-04T18:53:35Z

In reply to nevent1q…xmyf
_________________________

I wonder how Danish "hest" mixes into this.

Phew, if you know how many people fled from the GDR before the ...

2025-06-27T16:59:35Z

In reply to nevent1q…lu9g
_________________________

Phew, if you know how many people fled from the GDR before the wall was built (and even afterwards, many paying with their lives)...

I don't know if this disqualifies her for the job. She never ...

2025-06-27T15:31:26Z

In reply to nevent1q…q4hc
_________________________

I don't know if this disqualifies her for the job. She never met him, not even her father ever met him.
You think maybe there's something evil in her genes?

I just made some measurements with an EPCOS/TDK B62152A1X1 ...

2025-01-02T16:37:13Z

In reply to nevent1q…82s4
_________________________

I just made some measurements with an EPCOS/TDK B62152A1X1 binocular core. It has an A_L of 330nH+-30% (assuming full turns around the center leg). 4 turns measure 5.74µH, so that figures. A single wire through one hole measures 236nH.

1/n

This is obvious when you think harder about it, using that rule ...

2024-12-31T13:19:01Z

In reply to nevent1q…fe0w
_________________________

This is obvious when you think harder about it, using that rule of thumb from the ham radio course, an open loop, i.e., Z->\infty would have been transformed to Z->\infty on the primary side as well, but instead we see the no-load inductance, because the assumption Z << j\omegaL1 no longer holds.
That also implies that a transformer becomes "better" at transforming impedances the higher its no-load inductance is.

This only becomes tractable if we assume an ideal transformer, ...

2024-12-31T13:16:02Z

In reply to nevent1q…5n7q
_________________________

This only becomes tractable if we assume an ideal transformer, i.e., M12^2=L1*L2. Toroids are not so bad in this respect, but still an idealizing assumption.
Deriving this helped me cut the Gordian knot I had in my brain since the ham radio course. A transformer will only transform impedances within certain bounds, i.e., the load impedance has to be significantly smaller than the no-load impedance. Only then, Z is actually multiplied with L1/L2, i.e., n1^2/n2^2.

Now the impedance-transforming thing, the cargo cult of all radio ...

2024-12-31T13:10:37Z

In reply to nevent1q…uqn0
_________________________

Now the impedance-transforming thing, the cargo cult of all radio amateurs.
In general, you see the folloing impedance on side 1 if you terminate side 2 with load Z:
Not so nice, obviously.

The stray inductance is what you have to put in series with an ...

2024-12-31T13:07:54Z

In reply to nevent1q…8dd9
_________________________

The stray inductance is what you have to put in series with an ideal transformer (of reduced inductance) to model a nonideal one (the overall no-load inductance has to be the same).

And because I don't want to promise things I can't keep ...

2024-12-31T13:05:04Z

In reply to nevent1q…pr9e
_________________________

And because I don't want to promise things I can't keep heres how you get these properties: The no-load inductance is trivial: Without load, I2 is 0, therefore the current controlled voltage source on the primary side is 0, too. You only see L1, this is the no-load inductance.
Now the short-circuit case:
Depending on the coupling k, you either fully see L1 (k=0, no coupling), or you see a short circuit (k=1, ideal transformer).

This is intended as a general thought piece, not a reply to your ...

2024-12-31T11:36:45Z

In reply to nevent1q…wqpx
_________________________

This is intended as a general thought piece, not a reply to your post, but I don't know where else to attach it in this thread, so here we go:

This is the general equivalent circuit of a 2-winding transformer. Nonideal, but linear. Everything we assume about them can be derived from it: no-load inductance, short-circuit inductance, stray inductance, impedance transformation properties.

The blue squiggles represent the stray inductance, larger if ...

2024-12-30T22:41:15Z

In reply to nevent1q…p7w2
_________________________

The blue squiggles represent the stray inductance, larger if coupling is worse.
I don't see how the lower left branch should be somehow in series with the output on the right hand side.

I still don't get what you mean by "choke". A series ...

2024-12-30T22:31:37Z

In reply to nevent1q…pp6z
_________________________

I still don't get what you mean by "choke". A series inductor? Between where and where? Or a common mode choke?

A transformer is all about currents. A current in any of the 3 windings will induce a voltage in every of the 3 windings (proportional by M_ij or L_i respectively}. This doesn't do anything per se, as long as this voltage doesn't cause a current (canceling the voltages in the other windings per Lenz's law).

The inductance towards the rail (because the MOSFET is high ...

2024-12-30T21:04:30Z

In reply to nevent1q…clvr
_________________________

The inductance towards the rail (because the MOSFET is high impedance)?

You basically have 3 windings in total, but the coupling between ...

2024-12-30T18:52:10Z

In reply to nevent1q…g5at
_________________________

You basically have 3 windings in total, but the coupling between the tubes is much lower than between each tube and coil, if I'm not mistaken.
Therefore, not much voltage is induced in the branch of the inactive MOSFET and the effect of its body diode can be ignored.

Having that much disposable income for hobby purposes deffo is an ...

2024-12-27T22:17:32Z

In reply to nevent1q…h7xl
_________________________

Having that much disposable income for hobby purposes deffo is an outlier.
I know that others buy yachts and classic cars, but still.

Why the exotic low-noise BC549 and not the general-purpose BC547? ...

2024-12-21T18:25:23Z

In reply to nevent1q…m9k2
_________________________

Why the exotic low-noise BC549 and not the general-purpose BC547? It's a digital circuit after all.

OC outputs are slooooow compared to push-pull outputs in the ...

2024-12-20T14:07:14Z

In reply to nevent1q…fzcu
_________________________

OC outputs are slooooow compared to push-pull outputs in the presence of parasitic capacitances.

I'm currently struggling with dumping an ESMT F50-series QSPI ...

2024-12-18T06:11:06Z

In reply to nevent1q…xmun
_________________________

I'm currently struggling with dumping an ESMT F50-series QSPI NAND flash. flashrom doesn't support it, loading the spinand.ko module on Raspberry Pi detects the chip correctly but fails to access its contents. My suspicion is the SPI controller, so I thought about synthesizing a Cadence QSPI into a Zynq-7000 and accessing it from Petalinux.
I have no idea how much effort it would be to write a SPI-NAND applet for Glasgow.

Exactly.

2024-11-11T16:47:28Z

In reply to nevent1q…2f0g
_________________________

Exactly.

Interesting, I would've put a second piece of stripboard, ...

2024-11-11T16:46:26Z

In reply to nevent1q…ps2z
_________________________

Interesting, I would've put a second piece of stripboard, 90° degree rotated on top of it.

Found it: http://ftec.cfasp.de/bre/spezial-schrauben.htm

2024-10-19T10:49:50Z

In reply to nevent1q…97t8
_________________________

Found it: http://ftec.cfasp.de/bre/spezial-schrauben.htm

Finally someone took the effort to machine them. I first saw them ...

2024-10-19T09:38:05Z

In reply to nevent1q…vrkv
_________________________

Finally someone took the effort to machine them. I first saw them as drawings >20 years ago.

This is a shitpost, right?

2024-10-17T05:51:53Z

In reply to nevent1q…4ff0
_________________________

This is a shitpost, right?

I guess it's just missing the connection between the ...

2024-09-02T16:47:14Z

In reply to nevent1q…n6wp
_________________________

I guess it's just missing the connection between the zener's cathode and the VCC pin.