On 06/06/2021 5:02 am, jlarkin@highlandsniptechnology.com wrote:
> On Sat, 05 Jun 2021 18:40:03 -0400, legg <legg@nospam.magma.ca> wrote:
>
>> On Sat, 05 Jun 2021 14:45:30 -0700, jlarkin@highlandsniptechnology.com
>> wrote:
>>
>>> On Sat, 5 Jun 2021 13:38:55 -0700 (PDT), Fred Bloggs
>>> <bloggs.fredbloggs.fred@gmail.com> wrote:
>>>
>>>> On Saturday, June 5, 2021 at 4:23:26 PM UTC-4, jla...@highlandsniptechnology.com wrote:
>>>>> One of our customers returned a bunch of fried VME waveform generators
>>>>>
>>>>> http://www.highlandtechnology.com/DSS/V375DS.shtml
>>>>>
>>>>> They were installed in a VME crate that has a kinda clumsy push-push
>>>>> power switch. If you fumble with it just right, turn if off for about
>>>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>>>>
>>>> How the hell does that happen???
>>>
>>> Looks like someone bought some switching supplies and hacked into them
>>> to add a low current pushbutton power switch on the front panel.
>>> Badly.
>>>
>>> If I fix them, maybe for free, I'll make some friends.
>>
>>
>> Hack into the low power push-botton switch.
>>
>> There's a bunch of POR detectors with delays in the second range.
>>
>> PT7M6315 2000ms
>> MCP130T 700ms
>> STM6524 4000-7500ms
>> SR2LA 4000-7500ms
>>
>> Fighting the overshoot with an active clamp shouldn't be as
>> power-intensive as you're thinking. a simple to247 doing its
>> zener clamping to 4VGS + a couple of diode drops should do it.
>>
>> Do both. Belt and braces.
>>
>> RL
>
> Pulling the crates and reworking them would be a big deal.
>
> I'm laying out a board now, with ten beefy DPAK fets with a bit of
> resistance in the drains to share the dissipation.
>
> Since nobody here wants to discuss the design of a gate driver, I did
> it myself.
>
>
>

Must be countless ways to do that. Here is my first stab at it:

<https://www.dropbox.com/s/qw4hq8xgdeevnlh/VME_Shunt.jpg?dl=0>

piglet

On Sat, 5 Jun 2021 22:54:00 -0700 (PDT), Ed Lee
<edward.ming.lee@gmail.com> wrote:

>On Saturday, June 5, 2021 at 10:45:26 PM UTC-7, RosemontCrest wrote:
>> On 6/5/21 10:20 PM, Ed Lee wrote:
>> > On Saturday, June 5, 2021 at 2:56:37 PM UTC-7, jla...@highlandsniptechnology.com wrote:
>> >> On Sat, 5 Jun 2021 14:09:29 -0700 (PDT), Fred Bloggs
>> >> <bloggs.fred...@gmail.com> wrote:
>> >>
>> >>> On Saturday, June 5, 2021 at 4:23:26 PM UTC-4, jla...@highlandsniptechnology.com wrote:
>> >>>> One of our customers returned a bunch of fried VME waveform generators
>> >>>>
>> >>>> http://www.highlandtechnology.com/DSS/V375DS.shtml
>> >>>>
>> >>>> They were installed in a VME crate that has a kinda clumsy push-push
>> >>>> power switch. If you fumble with it just right, turn if off for about
>> >>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>> >>>>
>> >>>> Our modules don't like that. Some of the older FPGAs are powered
>> >>>> directly from +5.
>> >>>>
>> >>>> So I've volunteered to make a clamp module. I'm thinking about three
>> >>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
>> >>>
>> >>> Don't do that. Put a delay-on-break relay in line with the mains. It won't let power be re-applied for the timeout etc...
>> >>>
>> >>>
>> >> It would be a major project to remove all the crates from a bunch of
>> >> sites and add all that stuff. A plugin clamp would be an easy field
>> >> addition.
>> >>
>> >> Looks like 10 or so surface-mount dpak fets would be nice. We have
>> >> FDD86367 in stock. Each will conduct about 100 amps with 5.5v on the
>> >> gate.
>> >
>> > I can't find any info on this FET.
>> https://www.onsemi.com/pdf/datasheet/fdd86367-d.pdf
>
>OK, that's fine. But it says 2v to 4v gate voltage, not 5.5v

That's the gate threshold voltage, what it takes to get 250 uA drain
current.

I want hundreds of amps.

--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.

On Sat, 5 Jun 2021 19:47:16 -0700 (PDT), Flyguy
<soar2morrow@yahoo.com> wrote:

>On Saturday, June 5, 2021 at 7:27:04 PM UTC-7, Cydrome Leader wrote:
>> Don Y <blocked...@foo.invalid> wrote:
>> > On 6/5/2021 2:21 PM, Tom Del Rosso wrote:
>> >> Don Y wrote:
>> >>> On 6/5/2021 1:23 PM, jla...@highlandsniptechnology.com wrote:
>> >>>> One of our customers returned a bunch of fried VME waveform
>> >>>> generators http://www.highlandtechnology.com/DSS/V375DS.shtml
>> >>>>
>> >>>> They were installed in a VME crate that has a kinda clumsy push-push
>> >>>> power switch. If you fumble with it just right, turn if off for about
>> >>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>> >>>>
>> >>>> Our modules don't like that. Some of the older FPGAs are powered
>> >>>> directly from +5.
>> >>>>
>> >>>> So I've volunteered to make a clamp module. I'm thinking about three
>> >>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
>> >>>>
>> >>>> IXFH400N075 maybe.
>> >>>>
>> >>>> If I make a VME module and use the P1 and P2 connectors, I'll have
>> >>>> six pins for the +5.
>> >>>>
>> >>>> Anybody want to have a whack at the gate driver circuit?
>> >>>
>> >>> Seems like it would be easier to just bolt on a "switch
>> >>> smartenerupper" to ensure power never gets cycled in ways that upset
>> >>> the power supplies.
>> >>
>> >> Yeah, I'm an amateur hobbyist know-nothing, but I would think an OVP
>> >> should be inside the supply, not plugged into the bus. If you need four
>> >> 1000 amp MOSFET's then won't there be some drop along the backplane
>> >> which would cause another board to get 7 or 8 volts? And what about
>> >> would it do to the connector? Oxidation? Melting?
>> >
>> > Yeah, seems like a bigger effort with more unknowns than
>> > simply addressing the (apparent) "cause" of the problem.
>> >
>> > Like beefing up the bumpers on your car because your
>> > brakes aren't good...
>> yeah, this entire project sounds stupid. It sounds like the repair fees
>> for the blown cards is pretty low though, so everyone should be happy.
>Here are a bunch of SCR crowbar circuits:
>https://www.google.com/search?rlz=1C1RXQR_enUS934US934&sxsrf=ALeKk01ShGR6oemCtdKzR0_qkEPgkjA8bg:1622947548395&source=univ&tbm=isch&q=scr+crowbar+voltage+clamp+module+5v&sa=X&ved=2ahUKEwjGr_Hg_oHxAhWGbs0KHcRCAacQjJkEegQIFBAB&biw=1280&bih=625

Most of those circuits will work once, namely destroy the SCR.

--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.

On 6/6/21 6:06 pm, jlarkin@highlandsniptechnology.com wrote:
> On Sat, 5 Jun 2021 22:54:00 -0700 (PDT), Ed Lee
> <edward.ming.lee@gmail.com> wrote:
>
>> On Saturday, June 5, 2021 at 10:45:26 PM UTC-7, RosemontCrest wrote:
>>> On 6/5/21 10:20 PM, Ed Lee wrote:
>>>> On Saturday, June 5, 2021 at 2:56:37 PM UTC-7, jla...@highlandsniptechnology.com wrote:
>>>>> On Sat, 5 Jun 2021 14:09:29 -0700 (PDT), Fred Bloggs
>>>>> <bloggs.fred...@gmail.com> wrote:
>>>>>
>>>>>> On Saturday, June 5, 2021 at 4:23:26 PM UTC-4, jla...@highlandsniptechnology.com wrote:
>>>>>>> One of our customers returned a bunch of fried VME waveform generators
>>>>>>>
>>>>>>> http://www.highlandtechnology.com/DSS/V375DS.shtml
>>>>>>>
>>>>>>> They were installed in a VME crate that has a kinda clumsy push-push
>>>>>>> power switch. If you fumble with it just right, turn if off for about
>>>>>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>>>>>>>
>>>>>>> Our modules don't like that. Some of the older FPGAs are powered
>>>>>>> directly from +5.
>>>>>>>
>>>>>>> So I've volunteered to make a clamp module. I'm thinking about three
>>>>>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
>>>>>>
>>>>>> Don't do that. Put a delay-on-break relay in line with the mains. It won't let power be re-applied for the timeout etc...
>>>>>>
>>>>>>
>>>>> It would be a major project to remove all the crates from a bunch of
>>>>> sites and add all that stuff. A plugin clamp would be an easy field
>>>>> addition.
>>>>>
>>>>> Looks like 10 or so surface-mount dpak fets would be nice. We have
>>>>> FDD86367 in stock. Each will conduct about 100 amps with 5.5v on the
>>>>> gate.
>>>>
>>>> I can't find any info on this FET.
>>> https://www.onsemi.com/pdf/datasheet/fdd86367-d.pdf
>>
>> OK, that's fine. But it says 2v to 4v gate voltage, not 5.5v
>
> That's the gate threshold voltage, what it takes to get 250 uA drain
> current.
>
> I want hundreds of amps.

Why don't you just turn off your device instead of short circuiting the
supply?

CH

On 06/06/2021 9:06 am, jlarkin@highlandsniptechnology.com wrote:
> On Sat, 5 Jun 2021 22:54:00 -0700 (PDT), Ed Lee
> <edward.ming.lee@gmail.com> wrote:
>>
>> OK, that's fine. But it says 2v to 4v gate voltage, not 5.5v
>
> That's the gate threshold voltage, what it takes to get 250 uA drain
> current.
>
> I want hundreds of amps.
>
>

Yes, and by 6-7V when the payload silicon is starting to hurt the FET
will be conducting very well. I think the scheme will work fine.

piglet

On 06/06/2021 9:26 am, Clifford Heath wrote:
> On 6/6/21 6:06 pm, jlarkin@highlandsniptechnology.com wrote:
>> On Sat, 5 Jun 2021 22:54:00 -0700 (PDT), Ed Lee
>> <edward.ming.lee@gmail.com> wrote:
>>
>>> On Saturday, June 5, 2021 at 10:45:26 PM UTC-7, RosemontCrest wrote:
>>>> On 6/5/21 10:20 PM, Ed Lee wrote:
>>>>> On Saturday, June 5, 2021 at 2:56:37 PM UTC-7,
>>>>> jla...@highlandsniptechnology.com wrote:
>>>>>> On Sat, 5 Jun 2021 14:09:29 -0700 (PDT), Fred Bloggs
>>>>>> <bloggs.fred...@gmail.com> wrote:
>>>>>>
>>>>>>> On Saturday, June 5, 2021 at 4:23:26 PM UTC-4,
>>>>>>> jla...@highlandsniptechnology.com wrote:
>>>>>>>> One of our customers returned a bunch of fried VME waveform
>>>>>>>> generators
>>>>>>>>
>>>>>>>> http://www.highlandtechnology.com/DSS/V375DS.shtml
>>>>>>>>
>>>>>>>> They were installed in a VME crate that has a kinda clumsy
>>>>>>>> push-push
>>>>>>>> power switch. If you fumble with it just right, turn if off for
>>>>>>>> about
>>>>>>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>>>>>>>>
>>>>>>>> Our modules don't like that. Some of the older FPGAs are powered
>>>>>>>> directly from +5.
>>>>>>>>
>>>>>>>> So I've volunteered to make a clamp module. I'm thinking about
>>>>>>>> three
>>>>>>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
>>>>>>>
>>>>>>> Don't do that. Put a delay-on-break relay in line with the mains.
>>>>>>> It won't let power be re-applied for the timeout etc...
>>>>>>>
>>>>>>>
>>>>>> It would be a major project to remove all the crates from a bunch of
>>>>>> sites and add all that stuff. A plugin clamp would be an easy field
>>>>>> addition.
>>>>>>
>>>>>> Looks like 10 or so surface-mount dpak fets would be nice. We have
>>>>>> FDD86367 in stock. Each will conduct about 100 amps with 5.5v on the
>>>>>> gate.
>>>>>
>>>>> I can't find any info on this FET.
>>>> https://www.onsemi.com/pdf/datasheet/fdd86367-d.pdf
>>>
>>> OK, that's fine.  But it says 2v to 4v gate voltage, not 5.5v
>>
>> That's the gate threshold voltage, what it takes to get 250 uA drain
>> current.
>>
>> I want hundreds of amps.
>
> Why don't you just turn off your device instead of short circuiting the
> supply?
>
> CH

I think JL wants a simple pragmatic retrofit option to an existing
product. A beefy shunt clamp could be simple enough for a user to fit
themselves.

piglet

søndag den 6. juni 2021 kl. 10.06.39 UTC+2 skrev jla...@highlandsniptechnology.com:
> On Sat, 5 Jun 2021 22:54:00 -0700 (PDT), Ed Lee
> <edward....@gmail.com> wrote:
>
> >On Saturday, June 5, 2021 at 10:45:26 PM UTC-7, RosemontCrest wrote:
> >> On 6/5/21 10:20 PM, Ed Lee wrote:
> >> > On Saturday, June 5, 2021 at 2:56:37 PM UTC-7, jla...@highlandsniptechnology.com wrote:
> >> >> On Sat, 5 Jun 2021 14:09:29 -0700 (PDT), Fred Bloggs
> >> >> <bloggs.fred...@gmail.com> wrote:
> >> >>
> >> >>> On Saturday, June 5, 2021 at 4:23:26 PM UTC-4, jla...@highlandsniptechnology.com wrote:
> >> >>>> One of our customers returned a bunch of fried VME waveform generators
> >> >>>>
> >> >>>> http://www.highlandtechnology.com/DSS/V375DS.shtml
> >> >>>>
> >> >>>> They were installed in a VME crate that has a kinda clumsy push-push
> >> >>>> power switch. If you fumble with it just right, turn if off for about
> >> >>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
> >> >>>>
> >> >>>> Our modules don't like that. Some of the older FPGAs are powered
> >> >>>> directly from +5.
> >> >>>>
> >> >>>> So I've volunteered to make a clamp module. I'm thinking about three
> >> >>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
> >> >>>
> >> >>> Don't do that. Put a delay-on-break relay in line with the mains. It won't let power be re-applied for the timeout etc...
> >> >>>
> >> >>>
> >> >> It would be a major project to remove all the crates from a bunch of
> >> >> sites and add all that stuff. A plugin clamp would be an easy field
> >> >> addition.
> >> >>
> >> >> Looks like 10 or so surface-mount dpak fets would be nice. We have
> >> >> FDD86367 in stock. Each will conduct about 100 amps with 5.5v on the
> >> >> gate.
> >> >
> >> > I can't find any info on this FET.
> >> https://www.onsemi.com/pdf/datasheet/fdd86367-d.pdf
> >
> >OK, that's fine. But it says 2v to 4v gate voltage, not 5.5v
> That's the gate threshold voltage, what it takes to get 250 uA drain
> current.
>
> I want hundreds of amps.

how big is the supply?

and if you just short out the supply rather than just a linear shunt, isn't it likely to do weird things?

On 06/06/2021 06:23, jlarkin@highlandsniptechnology.com wrote:
> One of our customers returned a bunch of fried VME waveform generators
>
> http://www.highlandtechnology.com/DSS/V375DS.shtml
>
> They were installed in a VME crate that has a kinda clumsy push-push
> power switch. If you fumble with it just right, turn if off for about
> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>
> Our modules don't like that. Some of the older FPGAs are powered
> directly from +5.
>
> So I've volunteered to make a clamp module. I'm thinking about three
> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
>
> IXFH400N075 maybe.
>
> If I make a VME module and use the P1 and P2 connectors, I'll have six
> pins for the +5.
>
> Anybody want to have a whack at the gate driver circuit?

The gate driver I did here regulates to a constant voltage, but is
probably needlessly complicated for your clamping needs:

https://bitbucket.org/chrisgj198/projects/raw/master/eload/kicad/eload1a.pdf

On Sunday, June 6, 2021 at 1:06:39 AM UTC-7, jla...@highlandsniptechnology.com wrote:
> On Sat, 5 Jun 2021 22:54:00 -0700 (PDT), Ed Lee
> <edward....@gmail.com> wrote:
>
> >On Saturday, June 5, 2021 at 10:45:26 PM UTC-7, RosemontCrest wrote:
> >> On 6/5/21 10:20 PM, Ed Lee wrote:
> >> > On Saturday, June 5, 2021 at 2:56:37 PM UTC-7, jla...@highlandsniptechnology.com wrote:
> >> >> On Sat, 5 Jun 2021 14:09:29 -0700 (PDT), Fred Bloggs
> >> >> <bloggs.fred...@gmail.com> wrote:
> >> >>
> >> >>> On Saturday, June 5, 2021 at 4:23:26 PM UTC-4, jla...@highlandsniptechnology.com wrote:
> >> >>>> One of our customers returned a bunch of fried VME waveform generators
> >> >>>>
> >> >>>> http://www.highlandtechnology.com/DSS/V375DS.shtml
> >> >>>>
> >> >>>> They were installed in a VME crate that has a kinda clumsy push-push
> >> >>>> power switch. If you fumble with it just right, turn if off for about
> >> >>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
> >> >>>>
> >> >>>> Our modules don't like that. Some of the older FPGAs are powered
> >> >>>> directly from +5.
> >> >>>>
> >> >>>> So I've volunteered to make a clamp module. I'm thinking about three
> >> >>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
> >> >>>
> >> >>> Don't do that. Put a delay-on-break relay in line with the mains. It won't let power be re-applied for the timeout etc...
> >> >>>
> >> >>>
> >> >> It would be a major project to remove all the crates from a bunch of
> >> >> sites and add all that stuff. A plugin clamp would be an easy field
> >> >> addition.
> >> >>
> >> >> Looks like 10 or so surface-mount dpak fets would be nice. We have
> >> >> FDD86367 in stock. Each will conduct about 100 amps with 5.5v on the
> >> >> gate.
> >> >
> >> > I can't find any info on this FET.
> >> https://www.onsemi.com/pdf/datasheet/fdd86367-d.pdf
> >
> >OK, that's fine. But it says 2v to 4v gate voltage, not 5.5v
> That's the gate threshold voltage, what it takes to get 250 uA drain
> current.

These have much lower gate drive voltage for rated current, perhaps higher for surge, if you want to stay with 3v to 4v drive.

Diode DMN3020UTS TSSOP-8 14A
Diode DMN2015UFDF DFN2020-6 14A
Vishay SiA448DJ SC-70-8L 12A
AlpOmg AO6404 TSOP6 8.6A
Rohm RQ1C075UN TSMT8 7.5A
ST STT5N2VH5 SOT23 5A

> I want hundreds of amps.

For how long?

On Sunday, June 6, 2021 at 1:11:00 AM UTC-7, jla...@highlandsniptechnology.com wrote:
> On Sat, 5 Jun 2021 19:47:16 -0700 (PDT), Flyguy
> <soar2...@yahoo.com> wrote:
> >On Saturday, June 5, 2021 at 7:27:04 PM UTC-7, Cydrome Leader wrote:
> >> Don Y <blocked...@foo.invalid> wrote:
> >> > On 6/5/2021 2:21 PM, Tom Del Rosso wrote:
> >> >> Don Y wrote:
> >> >>> On 6/5/2021 1:23 PM, jla...@highlandsniptechnology.com wrote:
> >> >>>> One of our customers returned a bunch of fried VME waveform
> >> >>>> generators http://www.highlandtechnology.com/DSS/V375DS.shtml
> >> >>>>
> >> >>>> They were installed in a VME crate that has a kinda clumsy push-push
> >> >>>> power switch. If you fumble with it just right, turn if off for about
> >> >>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
> >> >>>>
> >> >>>> Our modules don't like that. Some of the older FPGAs are powered
> >> >>>> directly from +5.
> >> >>>>
> >> >>>> So I've volunteered to make a clamp module. I'm thinking about three
> >> >>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
> >> >>>>
> >> >>>> IXFH400N075 maybe.
> >> >>>>
> >> >>>> If I make a VME module and use the P1 and P2 connectors, I'll have
> >> >>>> six pins for the +5.
> >> >>>>
> >> >>>> Anybody want to have a whack at the gate driver circuit?
> >> >>>
> >> >>> Seems like it would be easier to just bolt on a "switch
> >> >>> smartenerupper" to ensure power never gets cycled in ways that upset
> >> >>> the power supplies.
> >> >>
> >> >> Yeah, I'm an amateur hobbyist know-nothing, but I would think an OVP
> >> >> should be inside the supply, not plugged into the bus. If you need four
> >> >> 1000 amp MOSFET's then won't there be some drop along the backplane
> >> >> which would cause another board to get 7 or 8 volts? And what about
> >> >> would it do to the connector? Oxidation? Melting?
> >> >
> >> > Yeah, seems like a bigger effort with more unknowns than
> >> > simply addressing the (apparent) "cause" of the problem.
> >> >
> >> > Like beefing up the bumpers on your car because your
> >> > brakes aren't good...
> >> yeah, this entire project sounds stupid. It sounds like the repair fees
> >> for the blown cards is pretty low though, so everyone should be happy.
> >Here are a bunch of SCR crowbar circuits:
> >https://www.google.com/search?rlz=1C1RXQR_enUS934US934&sxsrf=ALeKk01ShGR6oemCtdKzR0_qkEPgkjA8bg:1622947548395&source=univ&tbm=isch&q=scr+crowbar+voltage+clamp+module+5v&sa=X&ved=2ahUKEwjGr_Hg_oHxAhWGbs0KHcRCAacQjJkEegQIFBAB&biw=1280&bih=625
> Most of those circuits will work once, namely destroy the SCR.

You need a fuse to save the SCR as well. Sustained hundreds of amps could very well destroy the PCB traces also.

On Sun, 6 Jun 2021 18:26:29 +1000, Clifford Heath <no.spam@please.net>
wrote:

>On 6/6/21 6:06 pm, jlarkin@highlandsniptechnology.com wrote:
>> On Sat, 5 Jun 2021 22:54:00 -0700 (PDT), Ed Lee
>> <edward.ming.lee@gmail.com> wrote:
>>
>>> On Saturday, June 5, 2021 at 10:45:26 PM UTC-7, RosemontCrest wrote:
>>>> On 6/5/21 10:20 PM, Ed Lee wrote:
>>>>> On Saturday, June 5, 2021 at 2:56:37 PM UTC-7, jla...@highlandsniptechnology.com wrote:
>>>>>> On Sat, 5 Jun 2021 14:09:29 -0700 (PDT), Fred Bloggs
>>>>>> <bloggs.fred...@gmail.com> wrote:
>>>>>>
>>>>>>> On Saturday, June 5, 2021 at 4:23:26 PM UTC-4, jla...@highlandsniptechnology.com wrote:
>>>>>>>> One of our customers returned a bunch of fried VME waveform generators
>>>>>>>>
>>>>>>>> http://www.highlandtechnology.com/DSS/V375DS.shtml
>>>>>>>>
>>>>>>>> They were installed in a VME crate that has a kinda clumsy push-push
>>>>>>>> power switch. If you fumble with it just right, turn if off for about
>>>>>>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>>>>>>>>
>>>>>>>> Our modules don't like that. Some of the older FPGAs are powered
>>>>>>>> directly from +5.
>>>>>>>>
>>>>>>>> So I've volunteered to make a clamp module. I'm thinking about three
>>>>>>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
>>>>>>>
>>>>>>> Don't do that. Put a delay-on-break relay in line with the mains. It won't let power be re-applied for the timeout etc...
>>>>>>>
>>>>>>>
>>>>>> It would be a major project to remove all the crates from a bunch of
>>>>>> sites and add all that stuff. A plugin clamp would be an easy field
>>>>>> addition.
>>>>>>
>>>>>> Looks like 10 or so surface-mount dpak fets would be nice. We have
>>>>>> FDD86367 in stock. Each will conduct about 100 amps with 5.5v on the
>>>>>> gate.
>>>>>
>>>>> I can't find any info on this FET.
>>>> https://www.onsemi.com/pdf/datasheet/fdd86367-d.pdf
>>>
>>> OK, that's fine. But it says 2v to 4v gate voltage, not 5.5v
>>
>> That's the gate threshold voltage, what it takes to get 250 uA drain
>> current.
>>
>> I want hundreds of amps.
>
>Why don't you just turn off your device instead of short circuiting the
>supply?
>
>CH

Good idea. The customer could just cut the AC line cord.

--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.

On Sun, 6 Jun 2021 03:50:12 -0700 (PDT), Lasse Langwadt Christensen
<langwadt@fonz.dk> wrote:

>søndag den 6. juni 2021 kl. 10.06.39 UTC+2 skrev jla...@highlandsniptechnology.com:
>> On Sat, 5 Jun 2021 22:54:00 -0700 (PDT), Ed Lee
>> <edward....@gmail.com> wrote:
>>
>> >On Saturday, June 5, 2021 at 10:45:26 PM UTC-7, RosemontCrest wrote:
>> >> On 6/5/21 10:20 PM, Ed Lee wrote:
>> >> > On Saturday, June 5, 2021 at 2:56:37 PM UTC-7, jla...@highlandsniptechnology.com wrote:
>> >> >> On Sat, 5 Jun 2021 14:09:29 -0700 (PDT), Fred Bloggs
>> >> >> <bloggs.fred...@gmail.com> wrote:
>> >> >>
>> >> >>> On Saturday, June 5, 2021 at 4:23:26 PM UTC-4, jla...@highlandsniptechnology.com wrote:
>> >> >>>> One of our customers returned a bunch of fried VME waveform generators
>> >> >>>>
>> >> >>>> http://www.highlandtechnology.com/DSS/V375DS.shtml
>> >> >>>>
>> >> >>>> They were installed in a VME crate that has a kinda clumsy push-push
>> >> >>>> power switch. If you fumble with it just right, turn if off for about
>> >> >>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>> >> >>>>
>> >> >>>> Our modules don't like that. Some of the older FPGAs are powered
>> >> >>>> directly from +5.
>> >> >>>>
>> >> >>>> So I've volunteered to make a clamp module. I'm thinking about three
>> >> >>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
>> >> >>>
>> >> >>> Don't do that. Put a delay-on-break relay in line with the mains. It won't let power be re-applied for the timeout etc...
>> >> >>>
>> >> >>>
>> >> >> It would be a major project to remove all the crates from a bunch of
>> >> >> sites and add all that stuff. A plugin clamp would be an easy field
>> >> >> addition.
>> >> >>
>> >> >> Looks like 10 or so surface-mount dpak fets would be nice. We have
>> >> >> FDD86367 in stock. Each will conduct about 100 amps with 5.5v on the
>> >> >> gate.
>> >> >
>> >> > I can't find any info on this FET.
>> >> https://www.onsemi.com/pdf/datasheet/fdd86367-d.pdf
>> >
>> >OK, that's fine. But it says 2v to 4v gate voltage, not 5.5v
>> That's the gate threshold voltage, what it takes to get 250 uA drain
>> current.
>>
>> I want hundreds of amps.
>
>how big is the supply?

I'm not sure what its peak current capability might be. Hundreds of
amps maybe for some milliseconds.
>
>and if you just short out the supply rather than just a linear shunt, isn't it likely to do weird things?

I'm proposing to clamp the +5 rail at about 5.5, not crowbar it.

--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.

On Sun, 6 Jun 2021 06:24:47 -0700 (PDT), Ed Lee
<edward.ming.lee@gmail.com> wrote:

>On Sunday, June 6, 2021 at 1:11:00 AM UTC-7, jla...@highlandsniptechnology.com wrote:
>> On Sat, 5 Jun 2021 19:47:16 -0700 (PDT), Flyguy
>> <soar2...@yahoo.com> wrote:
>> >On Saturday, June 5, 2021 at 7:27:04 PM UTC-7, Cydrome Leader wrote:
>> >> Don Y <blocked...@foo.invalid> wrote:
>> >> > On 6/5/2021 2:21 PM, Tom Del Rosso wrote:
>> >> >> Don Y wrote:
>> >> >>> On 6/5/2021 1:23 PM, jla...@highlandsniptechnology.com wrote:
>> >> >>>> One of our customers returned a bunch of fried VME waveform
>> >> >>>> generators http://www.highlandtechnology.com/DSS/V375DS.shtml
>> >> >>>>
>> >> >>>> They were installed in a VME crate that has a kinda clumsy push-push
>> >> >>>> power switch. If you fumble with it just right, turn if off for about
>> >> >>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>> >> >>>>
>> >> >>>> Our modules don't like that. Some of the older FPGAs are powered
>> >> >>>> directly from +5.
>> >> >>>>
>> >> >>>> So I've volunteered to make a clamp module. I'm thinking about three
>> >> >>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
>> >> >>>>
>> >> >>>> IXFH400N075 maybe.
>> >> >>>>
>> >> >>>> If I make a VME module and use the P1 and P2 connectors, I'll have
>> >> >>>> six pins for the +5.
>> >> >>>>
>> >> >>>> Anybody want to have a whack at the gate driver circuit?
>> >> >>>
>> >> >>> Seems like it would be easier to just bolt on a "switch
>> >> >>> smartenerupper" to ensure power never gets cycled in ways that upset
>> >> >>> the power supplies.
>> >> >>
>> >> >> Yeah, I'm an amateur hobbyist know-nothing, but I would think an OVP
>> >> >> should be inside the supply, not plugged into the bus. If you need four
>> >> >> 1000 amp MOSFET's then won't there be some drop along the backplane
>> >> >> which would cause another board to get 7 or 8 volts? And what about
>> >> >> would it do to the connector? Oxidation? Melting?
>> >> >
>> >> > Yeah, seems like a bigger effort with more unknowns than
>> >> > simply addressing the (apparent) "cause" of the problem.
>> >> >
>> >> > Like beefing up the bumpers on your car because your
>> >> > brakes aren't good...
>> >> yeah, this entire project sounds stupid. It sounds like the repair fees
>> >> for the blown cards is pretty low though, so everyone should be happy.
>> >Here are a bunch of SCR crowbar circuits:
>> >https://www.google.com/search?rlz=1C1RXQR_enUS934US934&sxsrf=ALeKk01ShGR6oemCtdKzR0_qkEPgkjA8bg:1622947548395&source=univ&tbm=isch&q=scr+crowbar+voltage+clamp+module+5v&sa=X&ved=2ahUKEwjGr_Hg_oHxAhWGbs0KHcRCAacQjJkEegQIFBAB&biw=1280&bih=625
>> Most of those circuits will work once, namely destroy the SCR.
>
>You need a fuse to save the SCR as well. Sustained hundreds of amps could very well destroy the PCB traces also.

If it blows a fuse, it works just once.

And I don't want to remove the VME crates and disassemble the power
supplies and add a huge fuse somewhere. May as well redesign the
supply to not spike when the switch is teased.

The hazard seem to happen with an MTBF per crate measured in decades,
just often enough to blow up our modules in one of many crates every
few years.

--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.

On Saturday, June 5, 2021 at 5:45:39 PM UTC-4, jla...@highlandsniptechnology.com wrote:
> On Sat, 5 Jun 2021 13:38:55 -0700 (PDT), Fred Bloggs
> <bloggs.fred...@gmail.com> wrote:
>
> >On Saturday, June 5, 2021 at 4:23:26 PM UTC-4, jla...@highlandsniptechnology.com wrote:
> >> One of our customers returned a bunch of fried VME waveform generators
> >>
> >> http://www.highlandtechnology.com/DSS/V375DS.shtml
> >>
> >> They were installed in a VME crate that has a kinda clumsy push-push
> >> power switch. If you fumble with it just right, turn if off for about
> >> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
> >
> >How the hell does that happen???
> Looks like someone bought some switching supplies and hacked into them
> to add a low current pushbutton power switch on the front panel.
> Badly.

Sounds like they hacked into the ENABLE of a buck controller, letting something float, so that during start-up it's letting the unregulated DC get thru the LC onto the output before it starts regulating.

>
> If I fix them, maybe for free, I'll make some friends.
> --
>
> John Larkin Highland Technology, Inc
>
> The best designs are necessarily accidental.

On Sunday, June 6, 2021 at 7:03:46 AM UTC-7, jla...@highlandsniptechnology.com wrote:
> On Sun, 6 Jun 2021 06:24:47 -0700 (PDT), Ed Lee
> <edward....@gmail.com> wrote:
>
> >On Sunday, June 6, 2021 at 1:11:00 AM UTC-7, jla...@highlandsniptechnology.com wrote:
> >> On Sat, 5 Jun 2021 19:47:16 -0700 (PDT), Flyguy
> >> <soar2...@yahoo.com> wrote:
> >> >On Saturday, June 5, 2021 at 7:27:04 PM UTC-7, Cydrome Leader wrote:
> >> >> Don Y <blocked...@foo.invalid> wrote:
> >> >> > On 6/5/2021 2:21 PM, Tom Del Rosso wrote:
> >> >> >> Don Y wrote:
> >> >> >>> On 6/5/2021 1:23 PM, jla...@highlandsniptechnology.com wrote:
> >> >> >>>> One of our customers returned a bunch of fried VME waveform
> >> >> >>>> generators http://www.highlandtechnology.com/DSS/V375DS.shtml
> >> >> >>>>
> >> >> >>>> They were installed in a VME crate that has a kinda clumsy push-push
> >> >> >>>> power switch. If you fumble with it just right, turn if off for about
> >> >> >>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
> >> >> >>>>
> >> >> >>>> Our modules don't like that. Some of the older FPGAs are powered
> >> >> >>>> directly from +5.
> >> >> >>>>
> >> >> >>>> So I've volunteered to make a clamp module. I'm thinking about three
> >> >> >>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
> >> >> >>>>
> >> >> >>>> IXFH400N075 maybe.
> >> >> >>>>
> >> >> >>>> If I make a VME module and use the P1 and P2 connectors, I'll have
> >> >> >>>> six pins for the +5.
> >> >> >>>>
> >> >> >>>> Anybody want to have a whack at the gate driver circuit?
> >> >> >>>
> >> >> >>> Seems like it would be easier to just bolt on a "switch
> >> >> >>> smartenerupper" to ensure power never gets cycled in ways that upset
> >> >> >>> the power supplies.
> >> >> >>
> >> >> >> Yeah, I'm an amateur hobbyist know-nothing, but I would think an OVP
> >> >> >> should be inside the supply, not plugged into the bus. If you need four
> >> >> >> 1000 amp MOSFET's then won't there be some drop along the backplane
> >> >> >> which would cause another board to get 7 or 8 volts? And what about
> >> >> >> would it do to the connector? Oxidation? Melting?
> >> >> >
> >> >> > Yeah, seems like a bigger effort with more unknowns than
> >> >> > simply addressing the (apparent) "cause" of the problem.
> >> >> >
> >> >> > Like beefing up the bumpers on your car because your
> >> >> > brakes aren't good...
> >> >> yeah, this entire project sounds stupid. It sounds like the repair fees
> >> >> for the blown cards is pretty low though, so everyone should be happy.
> >> >Here are a bunch of SCR crowbar circuits:
> >> >https://www.google.com/search?rlz=1C1RXQR_enUS934US934&sxsrf=ALeKk01ShGR6oemCtdKzR0_qkEPgkjA8bg:1622947548395&source=univ&tbm=isch&q=scr+crowbar+voltage+clamp+module+5v&sa=X&ved=2ahUKEwjGr_Hg_oHxAhWGbs0KHcRCAacQjJkEegQIFBAB&biw=1280&bih=625
> >> Most of those circuits will work once, namely destroy the SCR.
> >
> >You need a fuse to save the SCR as well. Sustained hundreds of amps could very well destroy the PCB traces also.
> If it blows a fuse, it works just once.

Same if it blows the PCB traces. How big are you planning on your traces to handle hundreds of amps? I think at least 50mils for sustained 5A.

> And I don't want to remove the VME crates and disassemble the power
> supplies and add a huge fuse somewhere. May as well redesign the
> supply to not spike when the switch is teased.

You have to tap it at the supply lines anyway, if you don't want to blow up your VME bus lines.

> The hazard seem to happen with an MTBF per crate measured in decades,
> just often enough to blow up our modules in one of many crates every
> few years.

Could be power surge on the AC line.

On Sun, 6 Jun 2021 08:36:09 +0100, piglet <erichpwagner@hotmail.com>
wrote:

>On 06/06/2021 5:02 am, jlarkin@highlandsniptechnology.com wrote:
>> On Sat, 05 Jun 2021 18:40:03 -0400, legg <legg@nospam.magma.ca> wrote:
>>
>>> On Sat, 05 Jun 2021 14:45:30 -0700, jlarkin@highlandsniptechnology.com
>>> wrote:
>>>
>>>> On Sat, 5 Jun 2021 13:38:55 -0700 (PDT), Fred Bloggs
>>>> <bloggs.fredbloggs.fred@gmail.com> wrote:
>>>>
>>>>> On Saturday, June 5, 2021 at 4:23:26 PM UTC-4, jla...@highlandsniptechnology.com wrote:
>>>>>> One of our customers returned a bunch of fried VME waveform generators
>>>>>>
>>>>>> http://www.highlandtechnology.com/DSS/V375DS.shtml
>>>>>>
>>>>>> They were installed in a VME crate that has a kinda clumsy push-push
>>>>>> power switch. If you fumble with it just right, turn if off for about
>>>>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>>>>>
>>>>> How the hell does that happen???
>>>>
>>>> Looks like someone bought some switching supplies and hacked into them
>>>> to add a low current pushbutton power switch on the front panel.
>>>> Badly.
>>>>
>>>> If I fix them, maybe for free, I'll make some friends.
>>>
>>>
>>> Hack into the low power push-botton switch.
>>>
>>> There's a bunch of POR detectors with delays in the second range.
>>>
>>> PT7M6315 2000ms
>>> MCP130T 700ms
>>> STM6524 4000-7500ms
>>> SR2LA 4000-7500ms
>>>
>>> Fighting the overshoot with an active clamp shouldn't be as
>>> power-intensive as you're thinking. a simple to247 doing its
>>> zener clamping to 4VGS + a couple of diode drops should do it.
>>>
>>> Do both. Belt and braces.
>>>
>>> RL
>>
>> Pulling the crates and reworking them would be a big deal.
>>
>> I'm laying out a board now, with ten beefy DPAK fets with a bit of
>> resistance in the drains to share the dissipation.
>>
>> Since nobody here wants to discuss the design of a gate driver, I did
>> it myself.
>>
>>
>>
>
>Must be countless ways to do that. Here is my first stab at it:
>
><https://www.dropbox.com/s/qw4hq8xgdeevnlh/VME_Shunt.jpg?dl=0>
>
>piglet

That's nice. I used a fixed bandgap, an LM4040-5 1% part.

https://www.dropbox.com/sh/oeqfilk8gdbr19n/AACLy_K-C5nPTHWjmtXwzwd_a?dl=0

--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.

On Saturday, June 5, 2021 at 5:30:50 PM UTC-4, Don Y wrote:
> On 6/5/2021 2:09 PM, Fred Bloggs wrote:
> > On Saturday, June 5, 2021 at 4:23:26 PM UTC-4,
> > jla...@highlandsniptechnology.com wrote:
> >> One of our customers returned a bunch of fried VME waveform generators
> >>
> >> http://www.highlandtechnology.com/DSS/V375DS.shtml
> >>
> >> They were installed in a VME crate that has a kinda clumsy push-push power
> >> switch. If you fumble with it just right, turn if off for about 3/4 of a
> >> second and turn it back on, the +5 bus spikes to about +9.
> >>
> >> Our modules don't like that. Some of the older FPGAs are powered directly
> >> from +5.
> >>
> >> So I've volunteered to make a clamp module. I'm thinking about three or
> >> four 1000 watt, 1000 amp mosfets and some sort of gate driver.
> >
> > Don't do that. Put a delay-on-break relay in line with the mains. It won't
> > let power be re-applied for the timeout etc...
> "Time" is likely a second-order manifestation. You likely want to watch
> some (or all) of the supplies and ensure they have discharged to some
> "appropriate" level before allowing power to be reapplied.
>
> Note that how quickly the supplies discharge will likely be reflective
> of the actual load they are PRESENTLY seeing. So, the "time" could
> change if the loading on the backplane changed (e.g., if one or more
> modules were removed).
>
> Using the current state of the supplies is more consistent with
> <whatever> is causing them to misbehave in this short-cycled
> application.

"Normally" supplies don't do stuff like this under any circumstances. Some hacker got into them and moved some of the control around without thinking it through completely.

On Sat, 05 Jun 2021 21:02:49 -0700, jlarkin@highlandsniptechnology.com
wrote:

>On Sat, 05 Jun 2021 18:40:03 -0400, legg <legg@nospam.magma.ca> wrote:
>
>>On Sat, 05 Jun 2021 14:45:30 -0700, jlarkin@highlandsniptechnology.com
>>wrote:
>>
>>>On Sat, 5 Jun 2021 13:38:55 -0700 (PDT), Fred Bloggs
>>><bloggs.fredbloggs.fred@gmail.com> wrote:
>>>
>>>>On Saturday, June 5, 2021 at 4:23:26 PM UTC-4, jla...@highlandsniptechnology.com wrote:
>>>>> One of our customers returned a bunch of fried VME waveform generators
>>>>>
>>>>> http://www.highlandtechnology.com/DSS/V375DS.shtml
>>>>>
>>>>> They were installed in a VME crate that has a kinda clumsy push-push
>>>>> power switch. If you fumble with it just right, turn if off for about
>>>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>>>>
>>>>How the hell does that happen???
>>>
>>>Looks like someone bought some switching supplies and hacked into them
>>>to add a low current pushbutton power switch on the front panel.
>>>Badly.
>>>
>>>If I fix them, maybe for free, I'll make some friends.
>>
>>
>>Hack into the low power push-botton switch.
>>
>>There's a bunch of POR detectors with delays in the second range.
>>
>>PT7M6315 2000ms
>>MCP130T 700ms
>>STM6524 4000-7500ms
>>SR2LA 4000-7500ms
>>
>>Fighting the overshoot with an active clamp shouldn't be as
>>power-intensive as you're thinking. a simple to247 doing its
>>zener clamping to 4VGS + a couple of diode drops should do it.
>>
>>Do both. Belt and braces.
>>
>>RL
>
>Pulling the crates and reworking them would be a big deal.
>
>I'm laying out a board now, with ten beefy DPAK fets with a bit of
>resistance in the drains to share the dissipation.
>
>Since nobody here wants to discuss the design of a gate driver, I did
>it myself.

Well, you're going to have to pull something, Shirley.

Something like an HUF75652 has a gate-source voltage that's
pretty steady around 4V from 5 to 50A. A zener from the
load-connected drain to the gate and a 3K9 gate shunt
resistor should clamp a start-up transient.

How was the low power on/off hardware configured? Hacking
that 'package' makes more sense, if all it needs is a delay.
Just stick the POR circuit in-line.

RL

jlarkin@highlandsniptechnology.com wrote:
>
> I'm proposing to clamp the +5 rail at about 5.5, not crowbar it.

There's a difference?

Can they access the PS connector to the backplane? Put a device between
there. The module connectors are going to weld or something if this is
triggered a number of times.

And what if they plug it into the backplane at the end opposite the PS?
What will the other modules see?

--
Defund the Thought Police

On Sun, 6 Jun 2021 07:12:26 -0700 (PDT), Fred Bloggs
<bloggs.fredbloggs.fred@gmail.com> wrote:

>On Saturday, June 5, 2021 at 5:45:39 PM UTC-4, jla...@highlandsniptechnology.com wrote:
>> On Sat, 5 Jun 2021 13:38:55 -0700 (PDT), Fred Bloggs
>> <bloggs.fred...@gmail.com> wrote:
>>
>> >On Saturday, June 5, 2021 at 4:23:26 PM UTC-4, jla...@highlandsniptechnology.com wrote:
>> >> One of our customers returned a bunch of fried VME waveform generators
>> >>
>> >> http://www.highlandtechnology.com/DSS/V375DS.shtml
>> >>
>> >> They were installed in a VME crate that has a kinda clumsy push-push
>> >> power switch. If you fumble with it just right, turn if off for about
>> >> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>> >
>> >How the hell does that happen???
>> Looks like someone bought some switching supplies and hacked into them
>> to add a low current pushbutton power switch on the front panel.
>> Badly.
>
>Sounds like they hacked into the ENABLE of a buck controller, letting something float, so that during start-up it's letting the unregulated DC get thru the LC onto the output before it starts regulating.
>
>

Something in the loop compensation probably doesn't have time to
discharge properly. I doubt that the schematics of the power supply
are available, and the crates are bolted into test racks across three
countries. A no-assembly-required plug-in fix is preferred, if it
doesn't don't melt the connector pins.

It's good politics to make the effort. Interesting too.

--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.

jlarkin@highlandsniptechnology.com wrote:
>
> Looks like someone bought some switching supplies and hacked into them
> to add a low current pushbutton power switch on the front panel.
> Badly.
>
> If I fix them, maybe for free, I'll make some friends.

Take one of the supplies out and find another way to hack into it.

Maybe fix the switch.

Maybe find a shutdown or feedback input.

--
Defund the Thought Police

On Sun, 6 Jun 2021 07:15:12 -0700 (PDT), Fred Bloggs
<bloggs.fredbloggs.fred@gmail.com> wrote:

>On Saturday, June 5, 2021 at 5:30:50 PM UTC-4, Don Y wrote:
>> On 6/5/2021 2:09 PM, Fred Bloggs wrote:
>> > On Saturday, June 5, 2021 at 4:23:26 PM UTC-4,
>> > jla...@highlandsniptechnology.com wrote:
>> >> One of our customers returned a bunch of fried VME waveform generators
>> >>
>> >> http://www.highlandtechnology.com/DSS/V375DS.shtml
>> >>
>> >> They were installed in a VME crate that has a kinda clumsy push-push power
>> >> switch. If you fumble with it just right, turn if off for about 3/4 of a
>> >> second and turn it back on, the +5 bus spikes to about +9.
>> >>
>> >> Our modules don't like that. Some of the older FPGAs are powered directly
>> >> from +5.
>> >>
>> >> So I've volunteered to make a clamp module. I'm thinking about three or
>> >> four 1000 watt, 1000 amp mosfets and some sort of gate driver.
>> >
>> > Don't do that. Put a delay-on-break relay in line with the mains. It won't
>> > let power be re-applied for the timeout etc...
>> "Time" is likely a second-order manifestation. You likely want to watch
>> some (or all) of the supplies and ensure they have discharged to some
>> "appropriate" level before allowing power to be reapplied.
>>
>> Note that how quickly the supplies discharge will likely be reflective
>> of the actual load they are PRESENTLY seeing. So, the "time" could
>> change if the loading on the backplane changed (e.g., if one or more
>> modules were removed).
>>
>> Using the current state of the supplies is more consistent with
>> <whatever> is causing them to misbehave in this short-cycled
>> application.
>
>"Normally" supplies don't do stuff like this under any circumstances. Some hacker got into them and moved some of the control around without thinking it through completely.

Very likely. Expensive blunder. It could have been a power supply
design error too.

--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.

On Sunday, June 6, 2021 at 7:41:35 AM UTC-7, jla...@highlandsniptechnology.com wrote:
> On Sun, 6 Jun 2021 07:15:12 -0700 (PDT), Fred Bloggs
> <bloggs.fred...@gmail.com> wrote:
>
> >On Saturday, June 5, 2021 at 5:30:50 PM UTC-4, Don Y wrote:
> >> On 6/5/2021 2:09 PM, Fred Bloggs wrote:
> >> > On Saturday, June 5, 2021 at 4:23:26 PM UTC-4,
> >> > jla...@highlandsniptechnology.com wrote:
> >> >> One of our customers returned a bunch of fried VME waveform generators
> >> >>
> >> >> http://www.highlandtechnology.com/DSS/V375DS.shtml
> >> >>
> >> >> They were installed in a VME crate that has a kinda clumsy push-push power
> >> >> switch. If you fumble with it just right, turn if off for about 3/4 of a
> >> >> second and turn it back on, the +5 bus spikes to about +9.
> >> >>
> >> >> Our modules don't like that. Some of the older FPGAs are powered directly
> >> >> from +5.
> >> >>
> >> >> So I've volunteered to make a clamp module. I'm thinking about three or
> >> >> four 1000 watt, 1000 amp mosfets and some sort of gate driver.
> >> >
> >> > Don't do that. Put a delay-on-break relay in line with the mains. It won't
> >> > let power be re-applied for the timeout etc...
> >> "Time" is likely a second-order manifestation. You likely want to watch
> >> some (or all) of the supplies and ensure they have discharged to some
> >> "appropriate" level before allowing power to be reapplied.
> >>
> >> Note that how quickly the supplies discharge will likely be reflective
> >> of the actual load they are PRESENTLY seeing. So, the "time" could
> >> change if the loading on the backplane changed (e.g., if one or more
> >> modules were removed).
> >>
> >> Using the current state of the supplies is more consistent with
> >> <whatever> is causing them to misbehave in this short-cycled
> >> application.
> >
> >"Normally" supplies don't do stuff like this under any circumstances. Some hacker got into them and moved some of the control around without thinking it through completely.
> Very likely. Expensive blunder. It could have been a power supply
> design error too.

So, you are hacking their hack. Shunting hundreds of amps might save your module, but moving this expensive fuse to the VME bus power rails as fuse. Can the bus rails handle hundreds of amps?

On Sun, 6 Jun 2021 07:12:59 -0700 (PDT), Ed Lee
<edward.ming.lee@gmail.com> wrote:

>On Sunday, June 6, 2021 at 7:03:46 AM UTC-7, jla...@highlandsniptechnology.com wrote:
>> On Sun, 6 Jun 2021 06:24:47 -0700 (PDT), Ed Lee
>> <edward....@gmail.com> wrote:
>>
>> >On Sunday, June 6, 2021 at 1:11:00 AM UTC-7, jla...@highlandsniptechnology.com wrote:
>> >> On Sat, 5 Jun 2021 19:47:16 -0700 (PDT), Flyguy
>> >> <soar2...@yahoo.com> wrote:
>> >> >On Saturday, June 5, 2021 at 7:27:04 PM UTC-7, Cydrome Leader wrote:
>> >> >> Don Y <blocked...@foo.invalid> wrote:
>> >> >> > On 6/5/2021 2:21 PM, Tom Del Rosso wrote:
>> >> >> >> Don Y wrote:
>> >> >> >>> On 6/5/2021 1:23 PM, jla...@highlandsniptechnology.com wrote:
>> >> >> >>>> One of our customers returned a bunch of fried VME waveform
>> >> >> >>>> generators http://www.highlandtechnology.com/DSS/V375DS.shtml
>> >> >> >>>>
>> >> >> >>>> They were installed in a VME crate that has a kinda clumsy push-push
>> >> >> >>>> power switch. If you fumble with it just right, turn if off for about
>> >> >> >>>> 3/4 of a second and turn it back on, the +5 bus spikes to about +9.
>> >> >> >>>>
>> >> >> >>>> Our modules don't like that. Some of the older FPGAs are powered
>> >> >> >>>> directly from +5.
>> >> >> >>>>
>> >> >> >>>> So I've volunteered to make a clamp module. I'm thinking about three
>> >> >> >>>> or four 1000 watt, 1000 amp mosfets and some sort of gate driver.
>> >> >> >>>>
>> >> >> >>>> IXFH400N075 maybe.
>> >> >> >>>>
>> >> >> >>>> If I make a VME module and use the P1 and P2 connectors, I'll have
>> >> >> >>>> six pins for the +5.
>> >> >> >>>>
>> >> >> >>>> Anybody want to have a whack at the gate driver circuit?
>> >> >> >>>
>> >> >> >>> Seems like it would be easier to just bolt on a "switch
>> >> >> >>> smartenerupper" to ensure power never gets cycled in ways that upset
>> >> >> >>> the power supplies.
>> >> >> >>
>> >> >> >> Yeah, I'm an amateur hobbyist know-nothing, but I would think an OVP
>> >> >> >> should be inside the supply, not plugged into the bus. If you need four
>> >> >> >> 1000 amp MOSFET's then won't there be some drop along the backplane
>> >> >> >> which would cause another board to get 7 or 8 volts? And what about
>> >> >> >> would it do to the connector? Oxidation? Melting?
>> >> >> >
>> >> >> > Yeah, seems like a bigger effort with more unknowns than
>> >> >> > simply addressing the (apparent) "cause" of the problem.
>> >> >> >
>> >> >> > Like beefing up the bumpers on your car because your
>> >> >> > brakes aren't good...
>> >> >> yeah, this entire project sounds stupid. It sounds like the repair fees
>> >> >> for the blown cards is pretty low though, so everyone should be happy.
>> >> >Here are a bunch of SCR crowbar circuits:
>> >> >https://www.google.com/search?rlz=1C1RXQR_enUS934US934&sxsrf=ALeKk01ShGR6oemCtdKzR0_qkEPgkjA8bg:1622947548395&source=univ&tbm=isch&q=scr+crowbar+voltage+clamp+module+5v&sa=X&ved=2ahUKEwjGr_Hg_oHxAhWGbs0KHcRCAacQjJkEegQIFBAB&biw=1280&bih=625
>> >> Most of those circuits will work once, namely destroy the SCR.
>> >
>> >You need a fuse to save the SCR as well. Sustained hundreds of amps could very well destroy the PCB traces also.
>> If it blows a fuse, it works just once.
>
>Same if it blows the PCB traces. How big are you planning on your traces to handle hundreds of amps? I think at least 50mils for sustained 5A.

My 4-layer clamp board will have giant power and ground pours. The
limit will be welding the six +5V pins on the P1/P2 backplane DIN
connectors. I'll add maybe 10 milliohms of trace resistance per pin,
before I hit my pours, to help equalize the currents per pin.

>
>> And I don't want to remove the VME crates and disassemble the power
>> supplies and add a huge fuse somewhere. May as well redesign the
>> supply to not spike when the switch is teased.
>
>You have to tap it at the supply lines anyway, if you don't want to blow up your VME bus lines.
>
>> The hazard seem to happen with an MTBF per crate measured in decades,
>> just often enough to blow up our modules in one of many crates every
>> few years.
>
>Could be power surge on the AC line.

The spike is invoked by precise fiddling with the clumsy push-push
power switch. Turning the supply off for just about 1 second makes the
+5 spike to +9. The techs manage to do this every few years.

--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.

On Sun, 6 Jun 2021 10:38:35 -0400, "Tom Del Rosso"
<fizzbintuesday@that-google-mail-domain.com> wrote:

>jlarkin@highlandsniptechnology.com wrote:
>>
>> I'm proposing to clamp the +5 rail at about 5.5, not crowbar it.
>
>There's a difference?

"Crowbar" usually means to fire an SCR (or some mechanical device) to
short a bus to ground. It doesn't recover until the power source is
cycled off and on again.

"Clamp" means to limit the voltage on the bus, to about 5.5 in my
case.

>
>Can they access the PS connector to the backplane? Put a device between
>there. The module connectors are going to weld or something if this is
>triggered a number of times.

Adding the shunt to the backplane is one option, but plugging into a
VME card slot is the preferred field upgrade. Something in series with
the power supply would work but would be harder to install in the
field.

Maybe some magical power strip could work too. Gotta think about that.

>
>And what if they plug it into the backplane at the end opposite the PS?
>What will the other modules see?

Backplanes ususlly use big PCB power pours or actual copper busbars
for power dustribution... pretty low overall resistance.

--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.