MIC5270 negative LDO misbehaving...

[Tim Williams]

\"Phil Hobbs\" <pcdhSpamMeSenseless@electrooptical.net> wrote in message
news:193da1c7-de94-c7c3-43db-d8680951a4fd@electrooptical.net...

Loop instability usually leads to oscillation at some frequency near the
unity gain cross, with a fairly well-defined period and a distorted
sinusoidal waveform.

This one is a sawtooth whose frequency depends very strongly on the load,
and whose waveform is very far from periodic. That\'s why I was looking
for some interaction with the (much faster) switching junk on the LDO\'s
input.

Often when this happens, it\'s because the crossover is much higher (low
MHz?), and the poles are so far right that it overshoots deep into cutoff.
So you end up with a relaxation oscillator. You have to load it hard enough
(whether by brute-force load current to get it back to threshold sooner, or
with a big enough cap (give or take ESR?) to absorb the huge output current
peak) that that nonlinearity goes away and the actual loop response becomes
apparent.

That\'s the usual thing with LDOs (sometimes it\'s even in the datasheet, you
often see a sawtooth bounce on the light-load transient plot), and a lot of
other regulators (I\'ve seen it plenty of times with TL431 regulated
flybacks). But if you can\'t load it (or unload it, for that matter) enough
to get into the near-linear range, that\'s a bit of a problem...

Also, it seems to be bipolar, so RF rectification isn\'t a bad guess.

On an only somewhat related note, I\'ve see integrated switching regs that
went cuckoo under induced noise. Specifically, a few kV of 1.5/50us surge,
in a wiring harness an inch above the board. Those were some oddball
regulator from Intel (via acquisition from a brand I don\'t remember), high
frequency, integrated inductor, tiny. Best I can tell, the ADJ pin (which
was hardly exposed on the board, maybe 0.01pF to the offending wire?) was
sensitive to that tiniest bit of overvoltage and shut itself down. Symptom
was, CPU resetting erratically. Great, huh?...

Tim

--
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Design
Website: https://www.seventransistorlabs.com/

 

[John Doe]

Piotr Wyderski <peter.pan@neverland.mil> wrote:

John Doe wrote:

That means the maximum current required at the input is 2700 nano
amps? Just curious.

It means that this \"infinite impedance non-inverting opamp input\" is
injecting 2.7uA into the circuit connected to it.

No results found for \"infinite impedance non-inverting opamp input\".

How can an input inject current into a circuit?

 

[Martin Brown]

On 08/10/2020 00:27, Phil Hobbs wrote:

On 2020-10-07 16:16, Fred Bloggs wrote:
On Wednesday, October 7, 2020 at 3:42:12 PM UTC-4, Phil Hobbs wrote:
So I have this little power supply board, which is sort of a collection
of our previous solutions to weird power supply needs, mainly needing
very quiet supplies.

It uses an LMR23630 sync buck at 2.1 MHz to make +12 from +24, which is
good for a couple of amps. Its output feeds a 450-kHz AOZ1282CI async
buck making -16.

There are linears and cap multipliers to make +15, -12, quiet +-15,
quiet +-5, and one or two others.

It all works very nicely except that the Micrel MIC5270YM5 LDO making
-12V produces an irregular sawtooth about a volt p-p which nothing seems
to fix.

Hanging a 300-ohm load on it makes the sawtooth speed up. That\'s
expected, because once the pass transistor kicks it out of regulation,
the capacitor has to bleed down until it happens again, and that happens
much faster with a load.

1 nF and 10 nF caps from output to FB don\'t help.

I thought it might be the switching spikes on the input confusing it. A
330-ohm high current bead (BLMAG331SN1) and a 1-uF bypass on the input
knocked the spikes down pretty well but didn\'t help the oscillation.

The part is specified for ceramic output caps of at least 1 uF, and the
datasheet claims that the value can be increased without limit. I tried
4.7 uF and 1 uF, both with good C(V) curves so that the actual
capacitance was reasonably well known. Didn\'t help.

I tried changing chips, but that made no difference.

Micrel brags about how stable and quiet the chip is, so presumably it
works in some regimes--in fact we use it in another gizmo to make -4
from -5 to -6, and it works fine. Apparently this lulled me into a
false sense of security.

Must be too much ESR:
\"High-ESR capacitors may cause instability. Capacitors
with an ESR of 3Ω or greater at 100kHz may cause a high
frequency oscillation.\"
Low ESR out to 100KHz means a high frequency low loss type.

Thanks.

Loop instability usually leads to oscillation at some frequency near the
unity gain cross, with a fairly well-defined period and a distorted
sinusoidal waveform.

This one is a sawtooth whose frequency depends very strongly on the
load, and whose waveform is very far from periodic.  That\'s why I was
looking for some interaction with the (much faster) switching junk on
the LDO\'s input.

  If the sawtooth is stable it\'s unlikely to be a coupling issue.

The maximum frequency sawtooth I\'ve seen is around 20 +- 20 kHz.  (It
really is very irregular.)

Tomorrow I\'ll disconnect it from the switchers and run it off a lab supply.

Is there anything special about the load that the device is facing -
extremely fast switching times for instance (even at low current).

The other thought is if it is load current related does the behaviour
become any more regular as the DC load current is steadily increased.

--
Regards,
Martin Brown

 

[Piotr Wyderski]

John Doe wrote:

> How can an input inject current into a circuit?

Asking this question implies that you have selected the red pill, Neo.

All opamps do that, but in this case the current is exceptionally high
(by the standards I am used to). In the case of, say, LT6242 it is 1pA,
while the LT8262 has 2700000pA -- value high enough to visibly distort
the circuit driving it.

Best regards, Piotr

 

[John Doe]

Piotr Wyderski <peter.pan@neverland.mil> wrote:

John Doe wrote:

How can an input inject current into a circuit?

Asking this question implies that you have selected the red pill,
Neo.

You need a better translator. I have no idea what \"red pill\" and
\"Neo\" refer to.

All opamps do that, but in this case the current is exceptionally
high (by the standards I am used to). In the case of, say, LT6242
it is 1pA, while the LT8262 has 2700000pA -- value high enough to
visibly distort the circuit driving it.

Looks like another translation problem. How does an input inject
current?

 

[Phil Hobbs]

On 2020-10-07 16:40, John Larkin wrote:

On Wed, 7 Oct 2020 15:42:01 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

So I have this little power supply board, which is sort of a collection
of our previous solutions to weird power supply needs, mainly needing
very quiet supplies.

It uses an LMR23630 sync buck at 2.1 MHz to make +12 from +24, which is
good for a couple of amps. Its output feeds a 450-kHz AOZ1282CI async
buck making -16.

There are linears and cap multipliers to make +15, -12, quiet +-15,
quiet +-5, and one or two others.

It all works very nicely except that the Micrel MIC5270YM5 LDO making
-12V produces an irregular sawtooth about a volt p-p which nothing seems
to fix.

We\'ve seen that, a linear reg making sawtooths. I think we used some
other Micrel reg that was bad news.

Maybe the current limiter circuit is less well-designed than the main
loop?


Hanging a 300-ohm load on it makes the sawtooth speed up. That\'s
expected, because once the pass transistor kicks it out of regulation,
the capacitor has to bleed down until it happens again, and that happens
much faster with a load.

1 nF and 10 nF caps from output to FB don\'t help.

I thought it might be the switching spikes on the input confusing it. A
330-ohm high current bead (BLMAG331SN1) and a 1-uF bypass on the input
knocked the spikes down pretty well but didn\'t help the oscillation.

The part is specified for ceramic output caps of at least 1 uF, and the
datasheet claims that the value can be increased without limit. I tried
4.7 uF and 1 uF, both with good C(V) curves so that the actual
capacitance was reasonably well known. Didn\'t help.

I tried changing chips, but that made no difference.

Micrel brags about how stable and quiet the chip is, so presumably it
works in some regimes--in fact we use it in another gizmo to make -4
from -5 to -6, and it works fine. Apparently this lulled me into a
false sense of security.

Razza frazza $#@*!!!


Did you try the Williams stabilization technique, namely using a huge
aluminum lytic?

The board is only 35 mm square and needs to be heatsinked via a gap pad,
so there\'s not a lot of space for BFCs, more\'s the pity.

Here\'s my latest idea about supplies for my DDG rev C.

https://www.dropbox.com/s/gu9fsljyju558w4/P505C_Regs_1.jpg?raw=1

Nice.

There are a few half-erased areas, which I also used to suffer from. I
swear by my Staedtler plug-in electric eraser--$15 on eBay some years
back, and an excellent investment. If you haven\'t used one, it\'ll
change your life. The little battery-powered ones are useless by
comparison.

Designing this is like solving a 7-sided Rubics cube in the dark. I\'ll
have 17 regulator chips, four of them duals, and a couple of
references.

The big Zynq is rude and crude and draws nasty current bursts as code
executes. The +1 volt switcher was oscillating at 80 KHz and made the
Arms intermittent. The trigger FPGA is super sensitive to supply
noise. Both have sequencing rules. There are thermal issues. The
pick-and-place machine has run out of feeder slots, so I can\'t just
add a lot of parts.

A couple of places, we are using a 56uF polymer cap in series with a 1
ohm resistor, as a damper. That\'s crazy but works.

I\'ve done that sort of thing too. The problem usually is finding a
resistor which has an adequate pulse rating, in case somebody shorts the
output or connects the box to a car battery.

Switchers driving switchers have fun negative-resistance issues.

My LTM8078 positive to dual negative switcher idea has occasional
hangup issues too. Not in Spice of course. Hence the 79M05 and the
question mark.

Not complaining, of course.

Naturally not. Building gizmos is such fun that it\'s fairly easy to
take difficulties in stride.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Phil Hobbs]

On 2020-10-08 05:59, Tim Williams wrote:

\"Phil Hobbs\" <pcdhSpamMeSenseless@electrooptical.net> wrote in message
news:193da1c7-de94-c7c3-43db-d8680951a4fd@electrooptical.net...
Loop instability usually leads to oscillation at some frequency near the
unity gain cross, with a fairly well-defined period and a distorted
sinusoidal waveform.

This one is a sawtooth whose frequency depends very strongly on the
load, and whose waveform is very far from periodic.  That\'s why I was
looking for some interaction with the (much faster) switching junk on
the LDO\'s input.

Often when this happens, it\'s because the crossover is much higher (low
MHz?), and the poles are so far right that it overshoots deep into
cutoff. So you end up with a relaxation oscillator.  You have to load it
hard enough (whether by brute-force load current to get it back to
threshold sooner, or with a big enough cap (give or take ESR?) to absorb
the huge output current peak) that that nonlinearity goes away and the
actual loop response becomes apparent.

The sharp edge of the sawtooth actually happens in two goes, about 2 us
apart and of about equal size. So whatever is going wrong seems to be
happening at very roughly 500 kHz.

It does it over the full range of loads, ~300 uA to 100 mA where the
current limit kicks in.

I also tried reducing the input voltage and feedback ratio to match the
one that works on the other board--no change to the instability. This
morning I\'ll try running it off a bench supply and see if that fixes it.

That\'s the usual thing with LDOs (sometimes it\'s even in the datasheet,
you often see a sawtooth bounce on the light-load transient plot), and a
lot of other regulators (I\'ve seen it plenty of times with TL431
regulated flybacks).  But if you can\'t load it (or unload it, for that
matter) enough to get into the near-linear range, that\'s a bit of a
problem...

Also, it seems to be bipolar, so RF rectification isn\'t a bad guess.

On an only somewhat related note, I\'ve see integrated switching regs
that went cuckoo under induced noise.  Specifically, a few kV of
1.5/50us surge, in a wiring harness an inch above the board.  Those were
some oddball regulator from Intel (via acquisition from a brand I don\'t
remember), high frequency, integrated inductor, tiny.  Best I can tell,
the ADJ pin (which was hardly exposed on the board, maybe 0.01pF to the
offending wire?) was sensitive to that tiniest bit of overvoltage and
shut itself down.  Symptom was, CPU resetting erratically.  Great, huh?...

I really miss dumb chips sometimes.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Phil Hobbs]

On 2020-10-08 06:51, Martin Brown wrote:

On 08/10/2020 00:27, Phil Hobbs wrote:
On 2020-10-07 16:16, Fred Bloggs wrote:
On Wednesday, October 7, 2020 at 3:42:12 PM UTC-4, Phil Hobbs wrote:
So I have this little power supply board, which is sort of a collection
of our previous solutions to weird power supply needs, mainly needing
very quiet supplies.

It uses an LMR23630 sync buck at 2.1 MHz to make +12 from +24, which is
good for a couple of amps. Its output feeds a 450-kHz AOZ1282CI async
buck making -16.

There are linears and cap multipliers to make +15, -12, quiet +-15,
quiet +-5, and one or two others.

It all works very nicely except that the Micrel MIC5270YM5 LDO making
-12V produces an irregular sawtooth about a volt p-p which nothing
seems
to fix.

Hanging a 300-ohm load on it makes the sawtooth speed up. That\'s
expected, because once the pass transistor kicks it out of regulation,
the capacitor has to bleed down until it happens again, and that
happens
much faster with a load.

1 nF and 10 nF caps from output to FB don\'t help.

I thought it might be the switching spikes on the input confusing it. A
330-ohm high current bead (BLMAG331SN1) and a 1-uF bypass on the input
knocked the spikes down pretty well but didn\'t help the oscillation.

The part is specified for ceramic output caps of at least 1 uF, and the
datasheet claims that the value can be increased without limit. I tried
4.7 uF and 1 uF, both with good C(V) curves so that the actual
capacitance was reasonably well known. Didn\'t help.

I tried changing chips, but that made no difference.

Micrel brags about how stable and quiet the chip is, so presumably it
works in some regimes--in fact we use it in another gizmo to make -4
from -5 to -6, and it works fine. Apparently this lulled me into a
false sense of security.

Must be too much ESR:
\"High-ESR capacitors may cause instability. Capacitors
with an ESR of 3Ω or greater at 100kHz may cause a high
frequency oscillation.\"
Low ESR out to 100KHz means a high frequency low loss type.

Thanks.

Loop instability usually leads to oscillation at some frequency near
the unity gain cross, with a fairly well-defined period and a
distorted sinusoidal waveform.

This one is a sawtooth whose frequency depends very strongly on the
load, and whose waveform is very far from periodic.  That\'s why I was
looking for some interaction with the (much faster) switching junk on
the LDO\'s input.

  If the sawtooth is stable it\'s unlikely to be a coupling issue.

The maximum frequency sawtooth I\'ve seen is around 20 +- 20 kHz.  (It
really is very irregular.)

Tomorrow I\'ll disconnect it from the switchers and run it off a lab
supply.

Is there anything special about the load that the device is facing -
extremely fast switching times for instance (even at low current).

Nah, just resistors at the moment.

The other thought is if it is load current related does the behaviour
become any more regular as the DC load current is steadily increased.

The oscillation frequency goes up because the cap bleeds down faster
with a heavier load, but no other noticeable change.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Phil Hobbs]

On 2020-10-08 08:15, Piotr Wyderski wrote:

John Doe wrote:

How can an input inject current into a circuit?

Asking this question implies that you have selected the red pill, Neo.

All opamps do that, but in this case the current is exceptionally high
(by the standards I am used to). In the case of, say, LT6242 it is 1pA,
while the LT8262 has 2700000pA -- value high enough to visibly distort
the circuit driving it.

ADC inputs are also outputs--they kick out enough charge to seriously
discombobulate some op amps. That RC on the input is not an optional extra.

I have a fairly swoopy Krohn-Hite tunable filter box with plugins that
I\'d use a lot more if its kickout weren\'t so hideous. Over the years
I\'ve learned to connect the inputs of any new instrument to a scope to
spot problems like that.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[John Doe]

Phil Hobbs wrote:

John Larkin wrote:

Here\'s my latest idea about supplies for my DDG rev C.

https://www.dropbox.com/s/gu9fsljyju558w4/P505C_Regs_1.jpg?raw=1

Nice.

There are a few half-erased areas, which I also used to suffer
from. I swear by my Staedtler plug-in electric eraser--$15 on
eBay some years back, and an excellent investment. If you haven\'t
used one, it\'ll change your life. The little battery-powered ones
are useless by comparison.

https://www.amazon.com/s?k=Staedtler+plug-in+electric+eraser&ref=nb_sb_noss

 

[John Doe]

Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

Piotr Wyderski wrote:
John Doe wrote:

How can an input inject current into a circuit?

Asking this question implies that you have selected the red pill,
Neo.

As if political zealotry injecting itself into an argument is cute.

> ADC inputs are also outputs

What does that look like?

 

[Phil Hobbs]

On 2020-10-08 09:12, John Doe wrote:

Phil Hobbs wrote:

John Larkin wrote:

Here\'s my latest idea about supplies for my DDG rev C.

https://www.dropbox.com/s/gu9fsljyju558w4/P505C_Regs_1.jpg?raw=1

Nice.

There are a few half-erased areas, which I also used to suffer
from. I swear by my Staedtler plug-in electric eraser--$15 on
eBay some years back, and an excellent investment. If you haven\'t
used one, it\'ll change your life. The little battery-powered ones
are useless by comparison.

https://www.amazon.com/s?k=Staedtler+plug-in+electric+eraser&ref=nb_sb_noss

Mine\'s a 527-00, which is discontinued but still available. There are
lots of older ones on the used market for cheap, but AFAICT they mostly
take the smaller-diameter eraser refills, which are unobtainium. The
527-00 takes the normal white Staedtler eraser refills.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Phil Hobbs]

On 2020-10-08 09:23, John Doe wrote:

Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

Piotr Wyderski wrote:
John Doe wrote:

How can an input inject current into a circuit?

Asking this question implies that you have selected the red pill,
Neo.



As if political zealotry injecting itself into an argument is cute.

That isn\'t politics, silly--Piotr\'s just teasing you. (Very gently, I
might add.)

You\'ve been labouring under the delusion that just because something is
called an \'input\', all it does is sit there. The reality is often far
different, hence the Matrix reference.

ADC inputs are also outputs

What does that look like?

Nasty nanosecond spikes of up to tens of picocoulombs. The internal
sampling cap of the ADC gets switched onto the input terminal, so it
charges up in a hurry. The capacitance is often 10-40 pF, so this is
not an inconsiderable amount of current.

With those RRIO amps JL was talking about, the big current pulse can
cause them to lose control of their outputs and dump a bunch of charge
into their internal compensation caps, so that they\'re confused for
microseconds.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Phil Hobbs]

On 2020-10-07 23:30, Steve Wilson wrote:

jlarkin@highlandsniptechnology.com wrote:

On Wed, 7 Oct 2020 20:04:47 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

Thanks. The sawtooth happens even at very light loads.

Cheers

Phil Hobbs

I wonder if one edge is an internal RF oscillation, and the other edge
is recovery.

Squegging?

Nah, doesn\'t have a grid leak.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[John Doe]

Phil Hobbs wrote:

John Doe wrote:
Phil Hobbs wrote:
Piotr Wyderski wrote:
John Doe wrote:

How can an input inject current into a circuit?

Asking this question implies that you have selected the red
pill, Neo.



As if political zealotry injecting itself into an argument is
cute.

That isn\'t politics, silly--Piotr\'s just teasing you. (Very
gently, I might add.)

You\'ve been labouring under the delusion that just because
something is called an \'input\', all it does is sit there. The
reality is often far different, hence the Matrix reference.

I didn\'t know the \"red pill\" thing came from The Matrix (1999).
Wasn\'t much into it. No time to derive clear meaning from everything
people say on the Internet. Now I know.






--

ADC inputs are also outputs

What does that look like?

Nasty nanosecond spikes of up to tens of picocoulombs. The
internal sampling cap of the ADC gets switched onto the input
terminal, so it charges up in a hurry. The capacitance is often
10-40 pF, so this is not an inconsiderable amount of current.

With those RRIO amps JL was talking about, the big current pulse
can cause them to lose control of their outputs and dump a bunch
of charge into their internal compensation caps, so that they\'re
confused for microseconds.

Cheers

Phil Hobbs

 

[server]

On Thu, 8 Oct 2020 09:42:27 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-10-07 23:30, Steve Wilson wrote:
jlarkin@highlandsniptechnology.com wrote:

On Wed, 7 Oct 2020 20:04:47 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

Thanks. The sawtooth happens even at very light loads.

Cheers

Phil Hobbs

I wonder if one edge is an internal RF oscillation, and the other edge
is recovery.

Squegging?

Nah, doesn\'t have a grid leak.

Cheers

Phil Hobbs

The input bypassing would affect that sort of oscillation.



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard

 

[server]

On Thu, 8 Oct 2020 08:43:27 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-10-07 16:40, John Larkin wrote:
On Wed, 7 Oct 2020 15:42:01 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

So I have this little power supply board, which is sort of a collection
of our previous solutions to weird power supply needs, mainly needing
very quiet supplies.

It uses an LMR23630 sync buck at 2.1 MHz to make +12 from +24, which is
good for a couple of amps. Its output feeds a 450-kHz AOZ1282CI async
buck making -16.

There are linears and cap multipliers to make +15, -12, quiet +-15,
quiet +-5, and one or two others.

It all works very nicely except that the Micrel MIC5270YM5 LDO making
-12V produces an irregular sawtooth about a volt p-p which nothing seems
to fix.

We\'ve seen that, a linear reg making sawtooths. I think we used some
other Micrel reg that was bad news.

Maybe the current limiter circuit is less well-designed than the main
loop?


Hanging a 300-ohm load on it makes the sawtooth speed up. That\'s
expected, because once the pass transistor kicks it out of regulation,
the capacitor has to bleed down until it happens again, and that happens
much faster with a load.

1 nF and 10 nF caps from output to FB don\'t help.

I thought it might be the switching spikes on the input confusing it. A
330-ohm high current bead (BLMAG331SN1) and a 1-uF bypass on the input
knocked the spikes down pretty well but didn\'t help the oscillation.

The part is specified for ceramic output caps of at least 1 uF, and the
datasheet claims that the value can be increased without limit. I tried
4.7 uF and 1 uF, both with good C(V) curves so that the actual
capacitance was reasonably well known. Didn\'t help.

I tried changing chips, but that made no difference.

Micrel brags about how stable and quiet the chip is, so presumably it
works in some regimes--in fact we use it in another gizmo to make -4
from -5 to -6, and it works fine. Apparently this lulled me into a
false sense of security.

Razza frazza $#@*!!!


Did you try the Williams stabilization technique, namely using a huge
aluminum lytic?

The board is only 35 mm square and needs to be heatsinked via a gap pad,
so there\'s not a lot of space for BFCs, more\'s the pity.


Here\'s my latest idea about supplies for my DDG rev C.

https://www.dropbox.com/s/gu9fsljyju558w4/P505C_Regs_1.jpg?raw=1

Nice.

There are a few half-erased areas, which I also used to suffer from. I
swear by my Staedtler plug-in electric eraser--$15 on eBay some years
back, and an excellent investment. If you haven\'t used one, it\'ll
change your life. The little battery-powered ones are useless by
comparison.

Oh, I have an AC-powered electric eraser and an electric pencil
sharpener. That sketch would make a superb blue-line copy, but it\'s a
lot easier to photograph with a cell phone camera, drawing lying on
the floor. We\'re still working mostly by email.

I have a long USB cable on my phone that lets me shoot drawings and my
whiteboard without disconnecting from my PC. It won\'t quite reach my
scopes.

I like to draw on vellum, at my drafting table in the big window. It\'s
functional enough.

I\'m going to re-do a lot of that later this morning. Too many
possibilities. Someone pointed out the the optional OCXO can pull a
full amp during warmup. Fixing that of course cascades in all
directions. We have a lot of switched +5 available, so I guess I\'ll do
a home-made LDO to make 3.3Q, the quiet 3.3 supply.

Designing this is like solving a 7-sided Rubics cube in the dark. I\'ll
have 17 regulator chips, four of them duals, and a couple of
references.

The big Zynq is rude and crude and draws nasty current bursts as code
executes. The +1 volt switcher was oscillating at 80 KHz and made the
Arms intermittent. The trigger FPGA is super sensitive to supply
noise. Both have sequencing rules. There are thermal issues. The
pick-and-place machine has run out of feeder slots, so I can\'t just
add a lot of parts.

A couple of places, we are using a 56uF polymer cap in series with a 1
ohm resistor, as a damper. That\'s crazy but works.

I\'ve done that sort of thing too. The problem usually is finding a
resistor which has an adequate pulse rating, in case somebody shorts the
output or connects the box to a car battery.


Switchers driving switchers have fun negative-resistance issues.

My LTM8078 positive to dual negative switcher idea has occasional
hangup issues too. Not in Spice of course. Hence the 79M05 and the
question mark.

Not complaining, of course.

Naturally not. Building gizmos is such fun that it\'s fairly easy to
take difficulties in stride.

Power supplies are pretty close to grunt work, with hazards. There\'s
no eureka-mode upside to getting them right.



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard

 

[server]

On Thu, 8 Oct 2020 07:52:06 +0200, Piotr Wyderski
<peter.pan@neverland.mil> wrote:

John Larkin wrote:

LM8261 is specifically c-loadable.

And hence I am puzzled with what I see. I admit I didn\'t have time to
debug this hopefully minor issue, but this little subcircuit should have
worked correctly from the very beginning. The part is genuine, and the
meter readings exceed the datasheet worst cases many times. Weird.

I will follow Phil\'s piece of advice.

Best regards, Piotr

If an opamp doesn\'t like a capacitive load, there is the option to add
a small series resistor to the c-load and split the feedback into AC
and DC paths. I use that with TCA0372 a lot, with a 1 ohm series
resistor.



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard

 

[server]

On Thu, 8 Oct 2020 09:08:41 +0200, Piotr Wyderski
<peter.pan@neverland.mil> wrote:

piglet wrote:

I am about to use that part so thanks for the alert. Have you eliminated
input bias current, datasheet worst case is 2700nA?

No, my dear piglet, not at all. Moreover, I have just identified this as
the root cause. Holy cow, didn\'t know they are still making parts like this.

It\'s not current or thermal limited either.



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard

 

[Martin Brown]

On 08/10/2020 13:51, Phil Hobbs wrote:

On 2020-10-08 06:51, Martin Brown wrote:

Is there anything special about the load that the device is facing -
extremely fast switching times for instance (even at low current).

Nah, just resistors at the moment.

Oh well. If it can\'t drive that then there really is something wrong.

The other thought is if it is load current related does the behaviour
become any more regular as the DC load current is steadily increased.

The oscillation frequency goes up because the cap bleeds down faster
with a heavier load, but no other noticeable change.

Sounds like the control loop has a pole in the wrong place for whatever
reason. I notice the weasel words \"designed for ceramic capacitors\" and
\"typically only requires 1uF\" on the output for stability. I\'d be
tempted to put a small fast decoupling capacitor in close either side.

--
Regards,
Martin Brown