Author Topic: 0-350V 0.2A Bench Power Supply  (Read 8562 times)

0 Members and 5 Guests are viewing this topic.

Offline james_s

  • Super Contributor
  • ***
  • Posts: 2785
  • Country: us
Re: 0-350V 0.2A Bench Power Supply
« Reply #250 on: December 05, 2017, 10:49:32 AM »
If you're worried about the pass device failing, install a meter on the panel connected across the output so you can monitor it. Any HV power supply should be considered live and dangerous regardless what the output is set at. It's like a firearm, always treat any gun as if it is loaded even if you know that it's not. There are other ways for the output to go to maximum uncommanded besides fail shorted of the pass element. A bad pot to set the voltage, a resistor going open, bad connection, someone bumping the knob, foreign object causing a short, this stuff happens.
 

Online Hero999

  • Super Contributor
  • ***
  • Posts: 8989
  • Country: gb
Re: 0-350V 0.2A Bench Power Supply
« Reply #251 on: December 05, 2017, 07:55:35 PM »
Quote
The circuit I posted is no more dangerous, than the one you mentioned. In both cases, the pass device(s) can fail short circuit, causing the output to sit at the full voltage.
http://www.pmillett.com/file_downloads/IP17.pdf
Not at all, it's a problem of probability: the probability of a tube to fail in short circuit is practically nil.  :box:
A tiny ingress of air could cause the valve to arc over internally and the probability of it happening is double because there are two in parallel.

Actually, the first thing I'd do, is use a centre tapped transformer, to give 200V and 400V supplies use another TL431 as a comparator to switch between them, depending on the output voltage  Here's a quick sketch. I haven't done an in depth simulation, hence why I haven't attached the .asc file. It probably needs some modifications to be a practical solution.

Pretty reasonable. :)

Tweaks like threshold current (TL431 is only accurate above 1mA) come down to merely pushing around resistors.

Or use TLV431, since it's cascoded in both instances and the lower voltage limit is not a hindrance.  Saving a couple mA helps a lot at this voltage. :)

Tim
I agree, the TLV431 is much more suitable!

If you're worried about the pass device failing, install a meter on the panel connected across the output so you can monitor it. Any HV power supply should be considered live and dangerous regardless what the output is set at. It's like a firearm, always treat any gun as if it is loaded even if you know that it's not. There are other ways for the output to go to maximum uncommanded besides fail shorted of the pass element. A bad pot to set the voltage, a resistor going open, bad connection, someone bumping the knob, foreign object causing a short, this stuff happens.
Exactly.
 

Offline oldway

  • Super Contributor
  • ***
  • Posts: 1808
Re: 0-350V 0.2A Bench Power Supply
« Reply #252 on: December 05, 2017, 08:17:52 PM »
Quote
A tiny ingress of air could cause the valve to arc over internally and the probability of it happening is double because there are two in parallel.
Another wrong answer....!!!! You said you don't know nothing about vacuum tubes and you proove it... Tubes have a getter to eliminate tiny ingress of air....Breakdown voltage in air is 3KV/mm.....distances between anode and cathode (and their connections) is far more than a mm !

Tubes like P(E)L500, PL504, PL509, PL519 works normally with 6KV peak voltage without any damage....
807 tubes can withstand several KV's.....
EL34 is used in public address audio amplifiers with 700V anode voltage in push pull schematic, that means it withstand an anode / cathode voltage of at least 1400V.....No match at all with semi-conductors !

About safety, primary protection is to prevent dangerous events, not only signalizing it.

Signalizing a danger is only a secundary safety, not the principal one.

Further more, for the same output current and same max output voltage, you must use more Mosfets in parallel than tubes, the risks are also greater for this reason....
« Last Edit: December 05, 2017, 08:36:36 PM by oldway »
 
The following users thanked this post: 001

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 9321
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: 0-350V 0.2A Bench Power Supply
« Reply #253 on: December 05, 2017, 08:36:31 PM »
It used to be that vacuum tubes such as the humble EL509 held one final advantage over semiconductors: performance at high voltages.

Parts like this approach it:
https://media.digikey.com/pdf/Data%20Sheets/IXYS%20PDFs/IXTx02N250(S).pdf
2500V 200mA, 450 ohms.  An EL509 has closer to 50 ohms on-resistance, and around 7kV peak plate voltage handling.

But since then, parts like this have been introduced:
http://ixapps.ixys.com/DataSheet/DS100458B(IXTL2N450).pdf
at a price competitive with a lone EL509 even, let alone including heater and screen supply.

For still-more-specialty applications, there are SiC MOSFETs up to 10kV (though not generally available anywhere, AFAIK), which are significantly better than hard modulator tubes.

Hmm, they're probably not generally available because pulse generators are "munitions".  I suppose because of radar and detonator applications?  Who knows.

Tim
Seven Transistor Labs, LLC
Electronic Design, from Concept to Layout.
Need engineering assistance? Drop me a message!
 

Offline oldway

  • Super Contributor
  • ***
  • Posts: 1808
Re: 0-350V 0.2A Bench Power Supply
« Reply #254 on: December 05, 2017, 08:43:32 PM »
Using a pentode with (relatively !) low voltage G2 is Intrinsically safe because it limits the short circuit current without need of additional components.

If you want to use semiconductors, you must go to the SMPS technology, not the linear technology with pass Mosfet .....
« Last Edit: December 05, 2017, 09:01:58 PM by oldway »
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 2785
  • Country: us
Re: 0-350V 0.2A Bench Power Supply
« Reply #255 on: December 06, 2017, 07:27:32 AM »
The getter will absorb tiny quantities of residual gas but it will quickly be overwhelmed by any sort of leak. I've certainly seen gassy tubes that arced internally, happened to one of the output tubes in my friend's guitar amp several years ago. I've also seen it in a CRT once, a microscopic leak somewhere allowed enough air in that the neck glowed bright purple when it was energized.

I would agree that a tube is less likely to fail shorted than a semiconductor but IMHO this is a moot point.
 

Offline oldway

  • Super Contributor
  • ***
  • Posts: 1808
Re: 0-350V 0.2A Bench Power Supply
« Reply #256 on: December 06, 2017, 07:57:56 AM »
It is extremely rare.....

it's a problem of probability: the probability of a tube to fail in short circuit or arcing is practically nil.

Guitars amplifiers are often abused, transported in trucks without any care, sometimes they even drops them, so it may happen that the glue of the octal socket comes off and there are breaks in the glass at the socket connections which can cause an air inlet.

This can not happen with a bench power supply that is not subject to this kind of mechanical stress.
« Last Edit: December 06, 2017, 08:02:00 AM by oldway »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 9321
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: 0-350V 0.2A Bench Power Supply
« Reply #257 on: December 06, 2017, 10:53:15 AM »
it's a problem of probability: <snip>

Guitars amplifiers are often ... <snip>

"It's not a problem. Except when it is."

Nice conditional probability there.  I give it Bayes out of 10.  :clap:

Tim
Seven Transistor Labs, LLC
Electronic Design, from Concept to Layout.
Need engineering assistance? Drop me a message!
 
The following users thanked this post: TheUnnamedNewbie

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 5739
Re: 0-350V 0.2A Bench Power Supply
« Reply #258 on: December 06, 2017, 01:59:39 PM »
Anyone who thinks the probability of failure of two tubes from the same batch, run in parallel, in adjacent sockets isn't strongly correlated is a fool.

Gassy CRTs are *RARE*.   It requires enough of a leak to let sufficient air in for the tube to support ionised conduction, without there being enough oxygen present for the filament to fail.  In a lifetime in the service trade you might see a few cases, but mostly a leaky tube will be very obvious - flat milky white getter and a blown filament.

If you want to 'gold plate' what was originally intended to be a cheap but usable HV bench supply,  use a dual gang pot and add a crowbar circuit for output overvoltage >10%, and if it uses tubes, loss of negative grid bias rail, or excessive G2 current.

If multiple tubes are used, to prevent overloading, their filaments should be in series so they all cut out if one fails.
 
 

Offline floobydust

  • Frequent Contributor
  • **
  • Posts: 794
  • Country: ca
Re: 0-350V 0.2A Bench Power Supply
« Reply #259 on: December 06, 2017, 04:41:33 PM »
Tube vs Solid-state reliability, each has their own aging and failure mechanisms.
Should we go biplane or helicopter?  ;D

If I needed something built in an hour, this is a junkbox tube design.
If I'm designing this professionally, it's SS with much greater effort, cost and risk.

There is no guidance on how much to derate low-cost switching MOSFET's SOA.
Using a bunch in linear mode, subject to thermal instability due to hot spots, with no DC SOA spec. is like walking into the casino.
Getting around this SS issue is expensive - or use tubes.

We need a "burning HV power supply" contest, with a scope giveaway and DMM's for the runners up.
If community members could place wagers on which design will win, my kind of casino  :popcorn:
 
The following users thanked this post: Ian.M

Offline oldway

  • Super Contributor
  • ***
  • Posts: 1808
Re: 0-350V 0.2A Bench Power Supply
« Reply #260 on: December 06, 2017, 10:37:09 PM »
I found a bunch of Ls50 tubes. Looks good, isn't it?
How to add transistor driver to it?

I heared what some kepco in 70' use some sort of transistor driver and current limiter with two stage pass tubes (one is for preregulator and the second is for stabiliser)
Nothing against the LS50 tube, it is a 40W transmitting pentode, it can be used as pass device without any problem. Heating voltage is 12.6V.
But the socket seems more expensive than the tube itself.

No way to find schematics or service manual of Kepco products  as they are copyrighted.
But you can order by Kepco every old service manual you need for a fee of....$400 !!!!!!
http://www.kepcopower.com/support/svcman.htm

For hybrid solution, see schematic I have posted over pass tube controlled by optpcoupler.
« Last Edit: December 06, 2017, 10:42:21 PM by oldway »
 

Offline oldway

  • Super Contributor
  • ***
  • Posts: 1808
Re: 0-350V 0.2A Bench Power Supply
« Reply #261 on: December 07, 2017, 10:39:15 PM »
Don't you think it would be fair to send a mail to Kepco asking them if they permit you to publish this schematic or not ?  :palm:

Quote
Look easy to build ?
  :-/O You don't even have the voltages of the secundaries of transformers...

It is a 2000V 100mA power supply with floating output.
It is quite different of Heathkit IP17 power supply because it is a much higher voltage and it is a floating output ....IP17 has negative polarity grounded.

The first tubes (you name preregulator) are limiting max Vak voltage (and also max. Pa)  of the regulator tubes.
When Vak of the regulator tubes became higher than a fixed value, preregulator start blocking and limit this voltage.
Control circuit is referenced with + output polarity, for this reason there is no need of high voltage drivers transistors.

As everyone of the polarities may be grounded , the control circuit can be at high voltage compare to ground.

For sure, there are a lot of precautions of safety.....(potentiometers and switches mounted on isolating board, plastic axles, ...)
« Last Edit: December 07, 2017, 10:43:01 PM by oldway »
 

Online Hero999

  • Super Contributor
  • ***
  • Posts: 8989
  • Country: gb
Re: 0-350V 0.2A Bench Power Supply
« Reply #262 on: Yesterday at 10:35:24 AM »
Quote
A tiny ingress of air could cause the valve to arc over internally and the probability of it happening is double because there are two in parallel.
Another wrong answer....!!!! You said you don't know nothing about vacuum tubes and you proove it... Tubes have a getter to eliminate tiny ingress of air....Breakdown voltage in air is 3KV/mm.....distances between anode and cathode (and their connections) is far more than a mm !
I'm no expert on valves but I do know more than nothing. I know about Paschen's law, which means that the breakdown voltage will at first reduce, if any air gets in. The getter is a highly reactive material and will only work up to a certain point, until it's depleted. If there's a leak, it won't go on absorbing air forever. Air will get in, first causing the breakdown voltage to reduce, causing arcing over. It's true that the breakdown voltage will go up again, but it depends on how quick the leak is.

Irrespective of the above arguments: any components which are safety critical need to be appropriately approved and rated. Semiconductors are generally not to be relied upon to protect against electric shock and I doubt thermionic valves are either: do you have any links to safety standards which suggest contrary?
 

Offline oldway

  • Super Contributor
  • ***
  • Posts: 1808
Re: 0-350V 0.2A Bench Power Supply
« Reply #263 on: Yesterday at 04:35:26 PM »
I agree with you that a linear HV power supply with a pass device is not safe and that it is not the right solution.
A modern solution would be a switch mode power supply who is less likely to fail with uncontrolled HV output voltage.

But this is too sphisticated for hobby use.

A lot of HV power supplies have been made with a very simple circuit using a pentode as pass element, and, so far I know, no accident did occured.

Nobody ever said that it was impossible to have an arc in a tube, but that it is an extremely rare event.
For an arc to be possible, there must be a weak air inlet such that the internal pressure remains low.
Generally, oxygen causes the burning of the filament.

As already said, the risk is much less than with a MOSFET.

Those who knew the time of the tubes, like me, know that a gassy tube is extremely rare .... It would be necessary to split the glass or that a solder glass / metal of the connections should fail .... There is no reason for this to happen unless there are exaggerated mechanical stresses on these connections.
It is only in output tubes of guitar amplifiers that I saw this kind of thing happening.
These amplifiers are often abused.
On tube tV set's, I have never seen that happen. (and I repaired hundreds of them in the years 60's and 70's)

 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 9321
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: 0-350V 0.2A Bench Power Supply
« Reply #264 on: Today at 05:32:44 AM »
Those who knew the time of the tubes, like me, know that a gassy tube is extremely rare ....
...
It is only in output tubes of guitar amplifiers that I saw this kind of thing happening.
These amplifiers are often abused.

Which are, gosh, 90% of tube amps today?

Hardly seems "extremely rare"!

Speaking as a professional engineer, I would not hesitate to use a pass regulator design for any application.  If high reliability were required, I would use redundant design methods.

I would use a switching design where efficiency is more important, that's all.

If high safety were required, that would be solved through a combination of protective circuits (e.g. crowbar) and mechanical interlocks.  You're a fool to think there's anything remotely safe, in practice or in legal terms, about a >70V power supply.  Such systems also go well beyond the scope of this thread.

But, as professional advice goes, it's only worth what you paid for it, right?...

Tim
Seven Transistor Labs, LLC
Electronic Design, from Concept to Layout.
Need engineering assistance? Drop me a message!
 

Offline floobydust

  • Frequent Contributor
  • **
  • Posts: 794
  • Country: ca
Re: 0-350V 0.2A Bench Power Supply
« Reply #265 on: Today at 08:03:50 AM »
Air in a tube gives lower and lower transconductance, then the filament fails. Never seen arcing from it. Air ingress only due to physical abuse or many decades of age.

One of six pass-MOSFETS shorted in that B&K 9185. Did nobody notice a commercial product's catastrophic failure?
Something got a full-blast 650VDC surprise instead of whatever it was set to  :o should I call my lawyer?

SS vs tube for safety, this seems like just a pissing match.
 
The following users thanked this post: oldway

Offline oldway

  • Super Contributor
  • ***
  • Posts: 1808
Re: 0-350V 0.2A Bench Power Supply
« Reply #266 on: Today at 10:55:21 AM »
Air in a tube gives lower and lower transconductance, then the filament fails. Never seen arcing from it. Air ingress only due to physical abuse or many decades of age.

One of six pass-MOSFETS shorted in that B&K 9185. Did nobody notice a commercial product's catastrophic failure?
Something got a full-blast 650VDC surprise instead of whatever it was set to  :o should I call my lawyer?

SS vs tube for safety, this seems like just a pissing match.
How to transfer years of experience and knowledge about tubes to beginners in the field, who, at most, have made a single project with tubes in their life?

That's the problem with this frorum, there are lots of people writing about things they do not know.

We have confidence in the tubes thanks to our experience of many years in projects, repairs and tests of tubes gears.
Only those who are over 65/70 years old have such an experience ....
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 2785
  • Country: us
Re: 0-350V 0.2A Bench Power Supply
« Reply #267 on: Today at 11:14:54 AM »
Publish the info? Answer questions to those who ask? We all accumulate information throughout our lives and it's depressing to think how much knowledge is lost as older segments of the population die off. At the same time it's not realistic to expect all of it to remain useful or of value to younger folks. I expect that by the time I'm 60-70 very few people will care about my knowledge and experience repairing CRT displays, setting up the convergence and purity on a color tube, knowing how the scan circuits video amplifiers work, but times change and like other vacuum tubes CRTs will have largely disappeared from daily life. Again like other vacuum tubes I'm sure there will remain loyal pockets of people who like them for niche applications but otherwise society will have moved on and found new ways to do things regardless of what merits some of the old tech may still have. That's just life, things change, but people will manage.

I still don't understand why this debate is even still going. When dealing with a 350V power supply design any number of things can go wrong besides catastrophic failure of the pass element, regardless of what topology is used it should incorporate features to prevent a failure of any sort from causing injury. Even with every possible failsafe one should never put themselves in a situation where they will be harmed if the output unexpectedly goes to full voltage. When you work on the wiring in your house you wouldn't fret about the possibility of a light switch failing and causing a shock, you'd (hopefully) shut off and tag the circuit breaker in the panel and still avoid grabbing onto exposed conductors.
 

Offline Alex Eisenhut

  • Super Contributor
  • ***
  • Posts: 1651
  • Country: ca
  • If you can buy it for 4$ on eBay, why design it?
Re: 0-350V 0.2A Bench Power Supply
« Reply #268 on: Today at 12:14:46 PM »


How to transfer years of experience and knowledge about tubes to beginners in the field, who, at most, have made a single project with tubes in their life?

Here's a good start:
http://www.john-a-harper.com/tubes201/
« Last Edit: Today at 12:25:04 PM by Alex Eisenhut »
 

Offline floobydust

  • Frequent Contributor
  • **
  • Posts: 794
  • Country: ca
Re: 0-350V 0.2A Bench Power Supply
« Reply #269 on: Today at 12:39:00 PM »
You would put the pro's and con's of the part/architecture, weighing things like cost, size, reliability, complexity etc. into  a 'decision matrix' to choose.
The forum (format) can't accomplish this- the requirements are poorly defined and the endless debate continues...

We're up against the stigma of an old technology where people don't know why it's "old"- it is to be scorned. Although it's solid and proven to work.
We're up against the marketing hype of new technology where people don't know the datasheets are full of deceptive bullshit. It's guess work as far as reliability. SPICE SOA models are unfortunately not there yet.


 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf