Author Topic: 0-350V 0.2A Bench Power Supply  (Read 53262 times)

0 Members and 1 Guest are viewing this topic.

Offline 001Topic starter

  • Super Contributor
  • ***
  • Posts: 1170
  • Country: aq
Re: 0-350V 0.2A Bench Power Supply
« Reply #125 on: November 24, 2017, 03:00:30 pm »
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12807
Re: 0-350V 0.2A Bench Power Supply
« Reply #126 on: November 24, 2017, 04:19:15 pm »
Beware of the maximum heater-cathode breakdown voltage.   In the test jig, you may want to put a beefy TVS diode between the cathode and ground to clamp it below 250V to protect your bench power supply, which should *NOT* be ground referenced when used for the filament.  Also put a 1A fast fuse rated for a minimum of 500V DC in series with the the output of the improvised variable HT supply I described above.  Ordinary glass fuses wont reliably interrupt 500V DC.
 

Offline chris_leyson

  • Super Contributor
  • ***
  • Posts: 1541
  • Country: wales
Re: 0-350V 0.2A Bench Power Supply
« Reply #127 on: November 24, 2017, 06:43:00 pm »
I've been looking at the T-reg regulator and there is an update on Jan Didden's web site https://linearaudio.nl/sites/linearaudio.net/files/T-reg%203%20article.pdf. I like the floating regulator design, the error amplifier is simple and using a separate transformer for the regulator supply and tube heater avoids heater cathode breakdon problems. The 1mA current source design for setting the output voltage looks a bit elaborate and I'm sure there are better ways of doing this, also you need a -50V supply for the tube version. Tried out a spice model and added a simple current limit, seems to be OK but will need to do more work on it. Current source for the voltage setting resistor was quickly put together and I need to open up the control loop and check stability. Load resistor R5 swept from 1 ohm to 25k ohm and output voltage is plotted against cathode current. EL34/6CA7 spice model from Duncan Amps. Gave up on the foldback current limit. Foldback limiting is OK for a fixed supply but a bit tricky on variable supplies. Also, I wouldn't use some of the transistors shown in the spice model, they just happened to be convenient.
« Last Edit: November 24, 2017, 07:24:00 pm by chris_leyson »
 
The following users thanked this post: 001

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: 0-350V 0.2A Bench Power Supply
« Reply #128 on: November 24, 2017, 06:48:06 pm »
Ugh...
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline chris_leyson

  • Super Contributor
  • ***
  • Posts: 1541
  • Country: wales
Re: 0-350V 0.2A Bench Power Supply
« Reply #129 on: November 24, 2017, 06:49:50 pm »
Quote
Ugh...
Indeed, jist doodling why not
 

Offline 001Topic starter

  • Super Contributor
  • ***
  • Posts: 1170
  • Country: aq
Re: 0-350V 0.2A Bench Power Supply
« Reply #130 on: November 24, 2017, 06:51:03 pm »


Thanx a lot!
Is it ok to imrove it with TL431 instead diodes?
 

Offline chris_leyson

  • Super Contributor
  • ***
  • Posts: 1541
  • Country: wales
Re: 0-350V 0.2A Bench Power Supply
« Reply #131 on: November 24, 2017, 07:06:24 pm »
Quote
Is it ok to imrove it with TL431 instead diodes?
Thanks 001, that's a really neat solution for a current souce I never thought of that. To be honest I just threw something together very quickly for the current source as I was more concerned about SOA,  power dissipation and stability and the current source I could leave until later. It's a neat design, four transistors a TL431 and a tube or tubes for the pass element. Of course mosfets could be used instead and that would save the -50V supply, but hey, if anyone needs a bench supply for tube work then you need a negative bias supply anyway.
 

Offline chris_leyson

  • Super Contributor
  • ***
  • Posts: 1541
  • Country: wales
Re: 0-350V 0.2A Bench Power Supply
« Reply #132 on: November 24, 2017, 08:59:27 pm »
Quote
Is it ok to imrove it with TL431 instead diodes?
Sorry, wasn't thinking. No, you would have to use LM385 or LM4041 for a transistor current source, LM385 and LM4041 feedback is referenced to cathode. TL431 will only do current sink as feedback is referenced to anode. Linear Technology LM385 is only two pin device so no good.
« Last Edit: November 24, 2017, 09:13:31 pm by chris_leyson »
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19345
  • Country: gb
  • 0999
Re: 0-350V 0.2A Bench Power Supply
« Reply #133 on: November 24, 2017, 10:22:14 pm »
Why not? What's wrong with tying the grids together?

...

It works perfectly with both grids connected together. I don't know if this works with all tetrodes or if it's specific to the simulation model. I suppose the data sheet is the best place to start.

Needing to bias the grid negative to cutoff is expected behaviour. Connecting the grids together is an unusual thing to do (as in, almost never done) and I suspect that the simulation model is simply not representative in that configuration. I have no idea what the valve would do in the real world. What range of grid-cathode voltage does your simulation regulate with?

Usual valve operation is that it is "on" with the grid at 0 volts relative to the cathode, and you progressively reduce the current as you bring the grid more negative with respect to the cathode. This behaviour  is like an N-channel JFET or depletion-mode MOESFET. If you make the grid positive with respect to the cathode then you can on some types turn it slightly harder on but too positive and it will start to collect electrons, and current will flow into the grid. This is generally bad because the grid is not able to dissipate much power. Grid 1 current / power limits tend to be very low (but I can't actually find it in the datasheet). Sorry if some of this is obvious but you said you don't have much experience with valves. Hopefully it's helpful to someone anyway.

PL509 datasheet is here: http://www.r-type.org/pdfs/pl509.pdf

Page 7 shows current as a function of grid 1 voltage with the valve "triode-strapped", i.e. grid 2 connected to the anode and a fixed anode voltage of 160 V. Note that the range of Vg1 is 0 to -50 V.

Pages 8 and 9 show pentode operation with grid 2 held at a fixed +160 V, 175 V and 190 V relative to the cathode but the anode voltage varied with constant Vg1. The Vg1 range is 0 to -40 V.

edit: Ian.M has posted whilst I was typing. He makes a good point about the alignment of G1 and G2, but we're in agreement that the simulation model is probably junk for this configuration.
Thanks. I can see why my current limiting circuit is fatally flawed. I've managed to bodge it so it kind of works but it's still poor. The BJT limits the current first, then the short circuit current is determined by the characteristics of the valve. I imagine the valve's cut-off is extremely variable, like a JFET, so it won't work.

Last bad nubie valve circuit I promise.  ;)

EDIT:
Bad circuit attached, see the link below for the corrected one:
https://www.eevblog.com/forum/projects/looking-for-yours-opinions!-hv-stabilized-power-supply/msg1359532/#msg1359532
« Last Edit: November 26, 2017, 11:24:12 pm by Hero999 »
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12807
Re: 0-350V 0.2A Bench Power Supply
« Reply #134 on: November 25, 2017, 12:51:25 am »
If you are going to simulate this stuff please use a current source as the load (see component 'load') so you can use a simple .dc sweep rather than stepping a parametrised resistive load in a .op sim.  It will reduce the simulation run time by at least an order of magnitude.

You'll need an anti-parallel ideal diode across the current source load so it can only sink current and cant drive the output negative.   Use:
Code: [Select]
.model Dideal D(Ron=10n Roff=.1G Vfwd=0) ;Ideal diodeN.B. LTspice slows down drastically if the diode Ron and Roff are 12 or more decades apart.

Also, please label important intermediate nodes (e.g. round the valve and LM317) as its a PITA to be dealing with node numbers.

Now that's fixed, plot the dissipation (alt click) of Q1, and its collector current.  You've actually 'invented' a crappy shunt regulator.   Good luck finding a PNP BJT that can dissipate over 21W and withstand over 350V.  Most will be *FAR* outside their DC SOA.
« Last Edit: November 25, 2017, 01:17:46 am by Ian.M »
 
The following users thanked this post: 001

Offline oldway

  • Super Contributor
  • ***
  • !
  • Posts: 2172
Re: 0-350V 0.2A Bench Power Supply
« Reply #135 on: November 25, 2017, 11:14:13 am »
I wrote that I would not participate in this topic but you make me pity .... you do not know anything about electronic tubes and you try to make an absurd project to do the same thing as a simple power supply Heathkit IP17 ....  :palm:
Start by understanding how it works and what happens when you short-circuit the output of this HeathkitIP17 power supply ?

Is it protected against short circuits? If yes, how?.... :box:
It has not even a fuse in the HV rail.... :-DD

To help you answer to my question...  :-/O
What is the purpose of the characteristic curves of the tubes given by the manufacturers if you do not even consult them?  |O

And if you consult them, do not you notice an essential difference between a pentode and a triode?  :scared:

So, why do you want to use a pentode as a triode? To explode your power supply in case of short circuit?

A pentode limits the current independently of the anode voltage .... it is necessary to be stupid to not use this property.
At least it help to limit the instantaneous current of short circuit. This is intrinsically safe... :-+

The instantaneous power dissipated in the anode will be very much above the maximum power, but that does not matter, the anode has sufficient thermal inertia to accept this overload for a short time, a few seconds.

This is a sufficient time to blow a fuse, or disconnect the power supply, or have another electronic or electro-mechanical protection act.

In my project, I feed G2 of EL34 / 6CA7with 140V...I have only the curves for G2 = 250V....
« Last Edit: November 25, 2017, 11:44:43 am by oldway »
 
The following users thanked this post: 001

Offline oldway

  • Super Contributor
  • ***
  • !
  • Posts: 2172
Re: 0-350V 0.2A Bench Power Supply
« Reply #136 on: November 25, 2017, 11:41:44 am »
I wrote:
Quote
The idea of an hybrid power supply with solid state control and vacuum tube as a series regulator is excellent.
That's the way to go.
And what is the best way to combine high voltage power stage and semi conductor control ?
HV OPTOCOUPLER  :-+

The principle is very simple: To provide a bias of more or less -100V to G1 of EL34 to ensure full blocking of the tube and to reduce this bias by the control signal transmitted by an HV optocopler.
G1 will thus vary between -100V and 0V.

If you use several tubes in =, you must provide a cathode resistor to equalize currents between tubes.

NB: G3 internaly connected to cathode.

NB: schematic only for understanding of the principles of the project....

NB2: as I wrote:
Quote
If we have a winding 230V, or 325V rectified, we can do 2 ranges: one from 0 to 320V and one from 280 to 450V.
The anodic dissipation at 200mA will then be: range 1, 60W max, range 2, 40W max, which is quite reasonable and even allows to envisage a current greater than 200 mA.
The 2 ranges are selected by a switch. The other contacts of the switch are used in the control circuit.

NB3: heater circuit is not designed on the schematic. Heater voltage used is 12V with two heater filaments in serie. That's the reason why I choosed 4 EL34 and not 3.
« Last Edit: November 25, 2017, 12:42:21 pm by oldway »
 

Offline chris_leyson

  • Super Contributor
  • ***
  • Posts: 1541
  • Country: wales
Re: 0-350V 0.2A Bench Power Supply
« Reply #137 on: November 25, 2017, 12:45:06 pm »
Quote
If you are going to simulate this stuff please use a current source as the load
Quote
It will reduce the simulation run time by at least an order of magnitude

Hi Iam.M are you refering to my spice circuit or hero999 or both maybe ?
I tried sweeping the load resistor R5 just to see how well it worked and was surprised. The simulation time in this case is probably only 100 or 200ms at most, it's done as soon as I click the mouse button. It takes orders of magnitude longer to change the plot axis from time to current ;) The problem with swept current souces or sinks is that results are sometimes unpredictable if the source or sink current exceeds the limit current even with an anti-parallel diode fitted. BTW thanks for the model suggestion for the ideal anit-parallel diode.
One thing that I did notice was that if the grid pull down resistor was greater than 250k then the regulator is unstable at or near short circuit conditions so that's why I used 100k, also probably has a lot to do with the location of the compensation cap and to some extent the spice model for the EL34/6CA7. Instability increases simulation time by orders of magnitude  :)

Thanks oldway, you just reminded me I need to look at the screen grid bias, I wired the EL34 as a triode because it was quick and sorting out proper bias and looking at screen grid power dissipation is a job for later. To be honest my spice circuit is intended as a quick look and see what happens.

Thanks 001 for the TL431 suggestion for the current source. Used an LM385 and it works well, model is available from LTSpice forum.

In all honesty I need to start over and somehow open up the control loop, check phase and gain margins for stability, check closed loop gain and hum rejection and so on. Not to worried about DC accuracy and the DC drift should be good. The hum and noise performance of the original design is really very good for such a simple circuit. BTW my circuit as posted is not intended as a working design so don't build it ! Just noticed I left R6 and R7 in the circuit, I don't need those as they are for the Mosfet version, I should take them out  :palm:
« Last Edit: November 25, 2017, 12:55:24 pm by chris_leyson »
 

Offline oldway

  • Super Contributor
  • ***
  • !
  • Posts: 2172
Re: 0-350V 0.2A Bench Power Supply
« Reply #138 on: November 25, 2017, 01:21:49 pm »
Indeed, why make it simple and safe, if you can do complicated and dangerous .... :-DD

Our generation has been several times on the moon with very limited technological means ..... we have learned to do a lot with little ... We were ingenious and inventive engineers. It seems that this quality has been lost.  :palm:

EDIT: By using my power circuit, adjustable voltage control and current limiting circuits become extremely simple.
Just add a shunt in the negative rail to measure the current.
Control can be done by operational circuits as in any linear power supply.
There is no longer any semiconductor subjected to high voltages.
A simple + / - 12V auxiliary power supply is enough to power the control circuit.
A TL431 can be used as a reference if necessary.
« Last Edit: November 25, 2017, 01:35:14 pm by oldway »
 
The following users thanked this post: 001

Offline 001Topic starter

  • Super Contributor
  • ***
  • Posts: 1170
  • Country: aq
Re: 0-350V 0.2A Bench Power Supply
« Reply #139 on: November 25, 2017, 01:35:07 pm »
Yes

It is wery strange to see tubes that used this way  :-//
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12807
Re: 0-350V 0.2A Bench Power Supply
« Reply #140 on: November 25, 2017, 01:38:00 pm »
There's more on the Heathkit IP17 here: http://www.sgitheach.org.uk/ip17.html, including the schematic and manual.   Heathkit got a lot out of only six tubes - two paralleled 6L6GC for the pass element, a 6BH6 to drive them and for feedback regulation, two 150V stabiliser tubes and a double diode to rectify the supply to the stabilised -300V rail.

However the complexity was the custom transformers.  They would be uneconomic to reproduce for a one-off build unless you are equipped to wind your own and are experienced at doing so.  The filament supply one is pretty simple, it only needs five 6.3V secondaries (actually three + a CT 12.6V winding), and two of those are for the user's load.  It could easily be replaced with two off-the-shelf transformers.   Its the main HT transformer that's the P.I.T.A. Its got three secondaries: 175V for the  6L6GC G2 (screen grid) bias supply, that's referenced to the cathodes, 210V for the raw unreg 600V HT rail, via a voltage doubling rectifier (*STEAL* that idea!), and 600V CT for the raw -390V that feeds the regulated -300V bias rail.   

Worst case, a 6L6GC needs less than 20mA of G2 current, so in a modern design it would be acceptable to use an isolated flyback converter to derive a 50mA G2 bias supply, possibly with a floating IC regulator to further stabilise it, and the same could be done for the -300V bias supply.  That's probably going to be significantly cheaper than the custom transformer.  A 220V:220V isolating transformer would probably be close enough for the main HT supply, though it would be preferable to use one with both 220V and 240V primary taps so it could be tapped down a bit.

@Chris Leyson,
No, I was referring to Hero999's sim.  However, there is an alternative if you want to sweep a resistor using the .dc command - use a load resistor but edit its value to R=I(Vctrl) to make a behavioral resistor, then create a voltage source Vctrl to control it.  Set the series (internal) resistance to -1 ohms  :wtf: (Yes the NEGATIVE resistance is intentional to get positive current through the source for positive source voltage), and of course, ground one side of it. 

Another approach is not to short the voltage source or set its Rser, but instead label its output node, and change the behavioural resistor expression to refer to that node voltage.  However, you then have two names to keep consistent.

One can then sweep Vctrl in a .dc command.
.dc dec Vctrl 10k 10 covers the same range as Hero999's original .stepped sim.

There's a pitfall for the unwary:  You'd think you could simply sweep a shorted current source, but LTspice deletes shorted current sources before generating the netlist so the .dc command will always bitch that it cant find your control source!  |O
One fix is don't short the source, instead use a 1 ohm  resistor, but that's more complexity
« Last Edit: November 25, 2017, 01:42:12 pm by Ian.M »
 

Offline oldway

  • Super Contributor
  • ***
  • !
  • Posts: 2172
Re: 0-350V 0.2A Bench Power Supply
« Reply #141 on: November 25, 2017, 01:39:17 pm »
Yes

It is wery strange to see tubes that used this way  :-//
What do you mean with this ?
In the years 60, there was no optocoupler and the solution I am proposing was not possible.
It was possible to use direct coupling because 500 or 600V was no problem at all for the driving vacuum tube.
 

Offline 001Topic starter

  • Super Contributor
  • ***
  • Posts: 1170
  • Country: aq
Re: 0-350V 0.2A Bench Power Supply
« Reply #142 on: November 25, 2017, 01:41:17 pm »
-

It is not about Your circuit but about strange simulation results above  :palm:
 

Offline oldway

  • Super Contributor
  • ***
  • !
  • Posts: 2172
Re: 0-350V 0.2A Bench Power Supply
« Reply #143 on: November 25, 2017, 01:45:37 pm »
There's more on the Heathkit IP17 here: http://www.sgitheach.org.uk/ip17.html, including the schematic and manual.   Heathkit got a lot out of only six tubes - two paralleled 6L6GC for the pass element, a 6BH6 to drive them and for feedback regulation, two 150V stabiliser tubes and a double diode to rectify the supply to the stabilised -300V rail.
...

I already have  posted this link....
https://www.eevblog.com/forum/projects/looking-for-yours-opinions!-hv-stabilized-power-supply/msg1355691/#msg1355691

Of course, in my schematic, there is no need of negative rail.
Nul current in the optocoupler = fully blocked pass tubes with -100V Grid 1 negative bias.
« Last Edit: November 25, 2017, 01:52:25 pm by oldway »
 

Offline 001Topic starter

  • Super Contributor
  • ***
  • Posts: 1170
  • Country: aq
Re: 0-350V 0.2A Bench Power Supply
« Reply #144 on: November 25, 2017, 01:59:22 pm »
Classic horowitz & hill ideas
 

Online floobydust

  • Super Contributor
  • ***
  • Posts: 6924
  • Country: ca
Re: 0-350V 0.2A Bench Power Supply
« Reply #145 on: November 25, 2017, 04:57:10 pm »
Oldway, thanks for bring grid-bias (supply) back ;) That circuit looks good, add a bit of current-limiting to protect it. OP's 470uF filter cap stores 37.6J which may hurt things, that's a bit big. IP-17 is 70uF/2 lol.

I wasn't sure why the Heathkit IP-17 has a separate dedicated rail for the screens. It might be to keep 6L6 gain up at low drops? I guess we haven't looked at dropout voltage.
I recall the 6AS7/6080 was specifically designed for power supplies, but it worked with low plate voltages ~250V.

For real-life load testing HV power supplies, I use surplus 813 tetrodes with a cooling fan. I can't remember if secondary emission was a problem, thought the dynode region existed on the RCA ones I have- it might have been RF as I didn't put care into it and it's a bit tense on the workbench when the plates are glowing.

IXYS linear MOSFET IXTB30N100L 1,000V 30A looks tough but over $60 and Ciss=13,700pF (hard to shut off quickly) and TO-264 with no mounting hole
 

Offline oldway

  • Super Contributor
  • ***
  • !
  • Posts: 2172
Re: 0-350V 0.2A Bench Power Supply
« Reply #146 on: November 25, 2017, 05:31:24 pm »
Quote
Oldway, thanks for bring grid-bias (supply) back ;) That circuit looks good, add a bit of current-limiting to protect it.
It is protected because there are two current limits.

One instantaneous not depending of any other component than the caracteristics of the vacuum tube, wich is limited to about 100 mA/tube, another one who is slower and using the current information of a shunt resistor in the negative rail.

Here are the caracteristics of EL34 with 250V and 350V G2 voltage.....I am feeding G2 with only 140V...

With 250V G2 voltage and 0V G1 voltage (in my circuit, G1 can't go positive), Ia is limited at about 300mA with 500V anode voltage.
With 350V G2 voltage and 0V G1 voltage, Ia is limited at about 550 mA

With G2 at 140V, I am expecting a current limitation at about 100 mA/tube.

Of course, with Ua = 500V  and Ia = 0.1A, we have an anode dissipation of 50W, far too much for a 20W power tube.

But a second current limitation from the semi-conductor control unit using a shunt resistor in the negative rail will reduce further this current to the set value of 200 mA output current or 50mA/tube.
« Last Edit: November 25, 2017, 05:33:45 pm by oldway »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: 0-350V 0.2A Bench Power Supply
« Reply #147 on: November 25, 2017, 05:56:44 pm »
I wasn't sure why the Heathkit IP-17 has a separate dedicated rail for the screens. It might be to keep 6L6 gain up at low drops? I guess we haven't looked at dropout voltage.

Beware that the screen goes toaster mode if you let it reach dropout.

Sort of like the old bipolar LDOs that increased their quiescent current a hundredfold in dropout.

Quote
IXYS linear MOSFET IXTB30N100L 1,000V 30A looks tough but over $60 and Ciss=13,700pF (hard to shut off quickly) and TO-264 with no mounting hole

Well, obviously it's not a good choice then, is it?  If that's your only comparison, tubes may well be cheaper. :P

Best cost per watt tends to be either the very old designs -- IRFxxx and the like -- or new SuperJunction FETs (Fairchild QMOS, ST MDmesh, Infineon (and licensors) CoolMos, etc.) with modest ratings.  350V * 0.2A = 70W which is okay to burn in a single TO-247 or larger (around $5 and up), or two TO-220 in parallel (maybe $2 each).

The kind without mounting hole are better in all respects.  They are easier to mount, as you don't need to worry about the screw tilting the device, concentrating mounting pressure away from the die.  Instead, you use a readily available spring clip.  (Well, you can -- and should -- use a spring clip with conventional screw mount devices, too, of course!)

This isn't a hard project, and just adding the filament transformer alone already blows out any cost savings you might imagine, using toob.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Online floobydust

  • Super Contributor
  • ***
  • Posts: 6924
  • Country: ca
Re: 0-350V 0.2A Bench Power Supply
« Reply #148 on: November 26, 2017, 02:22:32 am »
Quote
Oldway, thanks for bring grid-bias (supply) back ;) That circuit looks good, add a bit of current-limiting to protect it.
It is protected because there are two current limits.

One instantaneous not depending of any other component than the caracteristics of the vacuum tube, wich is limited to about 100 mA/tube, another one who is slower and using the current information of a shunt resistor in the negative rail.

Here are the caracteristics of EL34 with 250V and 350V G2 voltage.....I am feeding G2 with only 140V...

With 250V G2 voltage and 0V G1 voltage (in my circuit, G1 can't go positive), Ia is limited at about 300mA with 500V anode voltage.
With 350V G2 voltage and 0V G1 voltage, Ia is limited at about 550 mA

With G2 at 140V, I am expecting a current limitation at about 100 mA/tube.

Of course, with Ua = 500V  and Ia = 0.1A, we have an anode dissipation of 50W, far too much for a 20W power tube.

But a second current limitation from the semi-conductor control unit using a shunt resistor in the negative rail will reduce further this current to the set value of 200 mA output current or 50mA/tube.

OK now I see how you are doing it. I saw the cathode resistors and pondered adding a current-sense transistor to help out.
 

Online floobydust

  • Super Contributor
  • ***
  • Posts: 6924
  • Country: ca
Re: 0-350V 0.2A Bench Power Supply
« Reply #149 on: November 26, 2017, 02:27:13 am »
I wasn't sure why the Heathkit IP-17 has a separate dedicated rail for the screens. It might be to keep 6L6 gain up at low drops? I guess we haven't looked at dropout voltage.

Beware that the screen goes toaster mode if you let it reach dropout.

Sort of like the old bipolar LDOs that increased their quiescent current a hundredfold in dropout.

Quote
IXYS linear MOSFET IXTB30N100L 1,000V 30A looks tough but over $60 and Ciss=13,700pF (hard to shut off quickly) and TO-264 with no mounting hole

Well, obviously it's not a good choice then, is it?  If that's your only comparison, tubes may well be cheaper. :P

Best cost per watt tends to be either the very old designs -- IRFxxx and the like -- or new SuperJunction FETs (Fairchild QMOS, ST MDmesh, Infineon (and licensors) CoolMos, etc.) with modest ratings.  350V * 0.2A = 70W which is okay to burn in a single TO-247 or larger (around $5 and up), or two TO-220 in parallel (maybe $2 each).

The kind without mounting hole are better in all respects.  They are easier to mount, as you don't need to worry about the screw tilting the device, concentrating mounting pressure away from the die.  Instead, you use a readily available spring clip.  (Well, you can -- and should -- use a spring clip with conventional screw mount devices, too, of course!)

This isn't a hard project, and just adding the filament transformer alone already blows out any cost savings you might imagine, using toob.

Tim


I think it's difficult for a thread to converge on a solution that OP and others are happy with.

The design requirements - budget- low cost, low component count, current-limiting, noise/precision, robustness, building one, building 1,000 etc. seems to throw the design from MOSFET, tube, switching PSU, or a hybrid of these. I don't know what "homebrewed polymer films" need for a OP's PSU.

The tube solution is simple, proven, robust but needs an additional filament+grid-bias power transformer (windings), inefficient, older parts and has the stigma of being old technology.

The MOSFET solution is fragile, complicated SOA protection and "trial by fire". Schrapnel from a TO-247 says "no, not quite right" and hopefully the load survives as well. Replace MOSFETs and try again.

The SMPS solution is a lot of engineering. The custom HV magnetics are a bear to make, unless you have cores, bobbins, magnet wire, insulating tape etc. and a lot of patience.


I mention the Ixys linear power MOSFET to show that one (moderately exotic) pass transistor could work (with fast enough SOA protection). There are 500V 15A parts around $10 each.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf