MC34063 component choice for +475V @ 50mA output, help needed!
Dekatron42
I need help in designing a power supply with the MC34063 which will
deliver +475V @ 50mA for running a bunch of dekatrons.
Calculating the component values is not the problem as there are many
"calculators" for the MC34063 - but knowing what brands and specific
types of components to use is where I need help.
I've been looking at Mike Moorrees' Mk 1.5 and Mk 2 designs with an
external mosfet, as I will need that to handle the power requirements.
I have so far used ordinary linear power supplys for this task but I
would like to shrink the size, and also to have something to do some
more experimenting with.
Aany help is highly appreciated!
/Martin
John Rehwinkel
> deliver +475V @ 50mA for running a bunch of dekatrons.
Wow, that's 24 watts! A pretty beefy supply there. For that kind of voltage and power, I'd
consider a transformer implementation instead of an inductor based boost converter
(with or without a voltage multiplier).
> Calculating the component values is not the problem as there are many
> "calculators" for the MC34063 - but knowing what brands and specific
> types of components to use is where I need help.
>
> I've been looking at Mike Moorrees' Mk 1.5 and Mk 2 designs with an
> external mosfet, as I will need that to handle the power requirements.
The MOSFET and inductor are the critical parts here when scaling up. Your inductor will need to pass all your output power, but with a narrow duty cycle, so you'll want your input voltage to be at the high end (18V or so), and even at DC, that's over an amp. With the narrow duty cycle for the large step-up, you'll want an inductor capable of at least ten amps or so - with a core that won't saturate at this current.
You'll need a pretty low "on" resistance with your MOSFET in order to have decent efficiency at this power level, and you may end up having to parallel a few of them (in which case I'd suggest using a dedicated gate driver chip to turn 'em all on and off solidly).
Your input capacitor will need to withstand a lot of ripple current, so you may end up with a few electrolytics in parallel, along with some higher-frequency types as bypass.
Your wiring will have to be compact and neat, and capable of handling these kinds of currents (if it's on a printed circuit board, use thick traces and heavy copper).
- John
threeneurons
> Hi!
>
> I need help in designing a power supply with the MC34063 which will
> deliver +475V @ 50mA for running a bunch of dekatrons.
>
Wow ! That's over 100 dekatrons ! Whatcha mak'n ?
What Jon says is true. You're gonna need a beefy coil and power FET.
Just some added info. The MC34063 design has a nominal efficiency of
75%; 80% if everything is perfect. 475V x 50mA = 23.8W. At 80% that's
30W in, and 6W of that is waste heat.At 75% that's 32W in; 8W of heat.
That's a lot of waste heat to dissipate in a small area.
The best inefficiencies are achieved with the MAX1771. 85% is
realistic expectation. At 85%, the input power will be 28W, with only
4.2W of waste heat. Here's a link to Nick's MAX1771 page:
http://www.desmith.net/NMdS/Electronics/NixiePSU.html
You can test it with both with, and without, the doubler. Adjust it
for 475V out straight, and test the results, then at 240V out, with a
doubler (to get the 480V), and test it. See which one works better.
Also, lower your voltage to 450V, or even 400V. What kind of dekatrons
are you using ? If you can get away with it, use the 400V. At 50mA,
that's 20W, or 15% less. 23.5W in at 85% efficiency; only 3.5W of
waste heat.
Dekatron42
I'll check it more thoroughly after work!
>
> Wow ! That's over 100 dekatrons ! Whatcha mak'n ?
>
I'm designing a clock! But with far from 100 dekatrons, but enough to
need a hefty PSU.
I might get away with 450V but that is at the lower end of the
specifications for the dekatrons I'm using, and old dekatrons tend to
need a voltage at the specified level or a little above that. If I add
it all up with all cold cathode trigger valves the minimum necessary
current (at its maximum) will be a bit below 50mA but I'd like to have
some headroom. I'm driving a few Nixies too at this voltage which adds
to it.
I am also thinking about using several smaller switching PSUs as there
is really no demand for it being one PSU. It would just be nice to
have it as one PSU, the design is modularized and it could be split
between a few PSUs without too much trouble.
If a single PSU needs a beefier inductor and several power mosfets in
parallel it might actually be easier to make a few PSUs and let them
drive just a few dekatrons each instead of one large PSU powering it
all, it might end up being cheaper too. It will be easier to spread
the heat too from several smaller PSUs too.
I just remembered that Ronald Dekker, www.dos4ever.com, had an
interrsting article about switch mode PSUs and inductors, I'll read
that too.
John Rehwinkel
> specifications for the dekatrons I'm using, and old dekatrons tend to
> need a voltage at the specified level or a little above that. If I add
> it all up with all cold cathode trigger valves the minimum necessary
> current (at its maximum) will be a bit below 50mA but I'd like to have
> some headroom. I'm driving a few Nixies too at this voltage which adds
> to it.
I'd run the nixies and probably the trigger tubes from a lower voltage. If you're using a doubler-style
supply, this is easy - connect the lower voltage stuff at the 250V node, and the dekatrons at the 500V
node. This will be much more efficient, and make using a single (two output) supply more feasible.
- John
Dekatron42
> supply, this is easy - connect the lower voltage stuff at the 250V node, and the dekatrons at the 500V
> node. This will be much more efficient, and make using a single (two output) supply more feasible.
>
I wish I could do that but GSA10Gs and the likes don't allow that if
you are driving Nixies, and the necessary circuitry around them is
also designed for these high voltages. I might figure out a way to do
it but that would be to go outside of the intended use by ETL and I am
really not that eager to test it with these rare dekatrons.
I've been fiddling around with a raised "ground" by supplying it with
a few hundred volts for th etrigger tubes and such, but it does not
change much in the total power needed on the 475V line, and it
complicates the design more than I feel like doing. "Keep it simple"
is what is needed here plus not deviating from the datasheets too
much.
/Martin
threeneurons
> > and the dekatrons at the 500V node.
>
> you are driving Nixies, and the necessary circuitry around them is
Where did you get that many GSA10Gs ?
I'd been collecting dekatrons since 2003, and I don't have a one of
the GSA10 or GCA10. Last haul of odd dekatrons, I got, was a half
dozen GC10/2Ps. That one was lucky on two points. The more important
point is that the supplier turned out to be a real snake
(understatement). On another audio tube forum, an eBay buyer describes
how this same guy swindled him on a $10,000 deal !
I've looked at the Ericsson datasheet on the GCA10 & GSA10. I can tell
you right now, that you can drop that voltage an additional 30V.
Instead of 475V+/-25V, you can use 445+/-20V. Ericsson "Digitrons"
have 200V strike voltages instead of the 170V, of a more common nixie.
Plus they don't show the interstage circuitry. That interstage
circuitry can be powered by 250V or less.
David Forbes
> Hi!
>
> I need help in designing a power supply with the MC34063 which will
> deliver +475V @ 50mA for running a bunch of dekatrons.
>
I built a flyback supply for my oscilloscopes, using a National Simple
Switcher part. You might be able to do the same. The transformer
requires winding, but it's only one long winding.
You'd probably need to use an EE40 or so core size. (I think; it's been
10 years.)
The advantage of the flyback supply is that it is easy to get it to
behave, and it doesn't require a high voltage power MOSFET.
Look at this part: LM2587
--
David Forbes, Tucson AZ
Dekatron42
> Where did you get that many GSA10Gs ?
>
I found most of them on eBay, and bought them as fast as I could! Plus
I got a tip from a person who told me that I could find a few from one
seller, whom I had to coax for more than two years into selling them
to me (first he denied he even had them, but then he mellowed).
I bought some one by one and then two batches of four and eight. I
have a total of twelve GSA10Gs of two different types, eight of them
looks more like the GCA10G in design and four like the more common
"development sample" GSA10G with "wings" between the auxillary anodes,
one GCA10G and one Z572S. Unfortunately one "wing" has come loose on
one of the GSA10Gs, I've been able to shake it so it lies loose inside
tha glass envelope at the top if you keep the valve upside down - so
that one will have to sit upside down to work properly.
I only have nine original B27B sockets for them but a bunch of B26A
sockets so I'll drill a hole and insert one extra pin in the center
for the Anode if I need to, I need these sockets for the GR10J & GR10K
too.
All of them are still intact and working but they work a lot better
near the 475-500V range than below 450V, so I guess that these
engineering samples might be just that - engineering samples not fit
for ordinary sale.
I'm also planning on using six GR10Js for the time display and two
GR10K for the 1Hz oscillator derived from 100Hz (50Hz from the wall
plug) together with two GSA10Gs for that divider (showing the hundreds
of seconds) and then a bunch of Z865Ws for the interstage triggering
and alarm/snooze/time-setting logic plus possibly a few GC10/2Ps for
the snooze function. So in total it will draw a lot of current. The
Z865Ws were especially developed for the GSA/GCA10Gs (if i understand
it correctly) as they trigger at low voltages around 5-8V and not at
25V like the GTE175Ms used for more ordinary dekatrons.
> I'd been collecting dekatrons since 2003, and I don't have a one of
> the GSA10 or GCA10. Last haul of odd dekatrons, I got, was a half
> dozen GC10/2Ps. That one was lucky on two points. The more important
> point is that the supplier turned out to be a real snake
> (understatement). On another audio tube forum, an eBay buyer describes
> how this same guy swindled him on a $10,000 deal !
I've also had my share of swindlers on eBay, but so far the Buyer
protection on PayPal has saved my hide every time.
>
> I've looked at the Ericsson datasheet on the GCA10 & GSA10. I can tell
> you right now, that you can drop that voltage an additional 30V.
> Instead of 475V+/-25V, you can use 445+/-20V. Ericsson "Digitrons"
> have 200V strike voltages instead of the 170V, of a more common nixie.
> Plus they don't show the interstage circuitry. That interstage
> circuitry can be powered by 250V or less.
The interstage circuitry is usually used at lower voltages like you
say, as the trigger tube only need to produce a voltage of
approximately 140v for the GSA/GCA10Gs, but the designs I've looked at
in various old magazines usually use just a single voltage for it all
if possible and then drop it with resistors in the anode circuit of
the trigger tubes. I might have to go this way if it is easier to do
so. Power supplies are not as fun to design as the rest of the logic
parts, but they are necessary, and I had hoped to use just one voltage
if possible keeping the design simpler.
/Martin
MrNixie (UK)
Not popular for supplies in the few-watts range, but if you need this
sort of power maybe a valve amp transformer is going to be your better
bet?
Dekatron42
I regularely use transformers from Tektronix oscilloscopes and valve
amplifiers for my projects, and I am really really careful when
testing stuff.
Valve amp transformers are so large and heavy since they usually
contain one or more heater windings which use up perhaps 50% of the
weight and size. I've actually been thinking about using two small
transformers back to back for supplying the 475V I need for each
module in the design. A Tek transfomer can supply approximately 100mA
at 500V (50VA) but up to 50A for the 6.3V (315VA) windings - which I
don't need in this project as I am going 100% cold cathode tubes
except for the diodes/rectifiers - cold cathode diodes/rectifiers are
not suitable here due to their high working voltage, unless anyone
know of any low voltage types. I've been thinking about using neon
lamps but they too have a too high voltage drop. This way there will
be a lot of transformers and the weight will almost be as much as from
a large transformer, but atleast it will be possible to distribute the
weight better. I know of a guy who can wind transformers as a last
resort so I can partly get rid of the size/weight problem but that
will eat a large hole in my economy.
So far I have only been bitten by a damaged valve tester when it had a
short from a broken relay coil to the case, this relay only shorted
when the relay was actuated which it normally shouldn't be as it works
as an over current protection - but of course I tripped the over
current protection and got bitten a few times. This if you excempt all
shocks I got when I was new to electronics and put my fingers in TV
sets that were turned on, but that was a long time ago and I only have
a few burn scars to show from that period. ;)
I recently bought a pair of rubber gloves specially made for high
voltage use, they can withstand 5KV and comes with a pair of inner
gloves made out of cotton so you shouldn't sweat while wearing them.
I'll use these now as I have three kids to think of nowadays!
Whoops, got a little bit carried away there - but that is the fun with
this hobby!
taylorjpt
This is an easy way to go directly off line so you don't have the
double conversion hit (Assuming you are not running this off
batteries!) of AC to LVDC to HVDC. This applies to any power supply
that puts out the power you want so in this case you would start with
an off the shelf switcher that can handle 24 watts (475V x 50mA =
23.75W). You also need to adjust the spacing to handle the output
voltage, so a dedicated output board is not a bad idea.
jt
Dekatron42
> double conversion hit (Assuming you are not running this off
> batteries!) of AC to LVDC to HVDC. This applies to any power supply
> that puts out the power you want so in this case you would start with
> an off the shelf switcher that can handle 24 watts (475V x 50mA =
> 23.75W). You also need to adjust the spacing to handle the output
> voltage, so a dedicated output board is not a bad idea.
>
Thanks for that tip, that was a nice way to do it!
I immediately sent a question to the guy I know who works with
transformers if he could help me rewinding a transformer, or if he can
make a new transformer from scratch.
I've only seen this done once before and I have tried to dismantle one
of these transformers but I think it was vacuum cured or something so
it was absolutely impossible without breaking it - but I did not apply
any heat to soften the glue/laquer as I thought it wasn't possible to
soften it!
Jon
On Nov 21, 10:52 pm, "MrNixie (UK)" <Laurence.Wilk...@dsl.pipex.com>
wrote:
> Jeepers, Buddy! 475V at 50mA? Did you check your life insurance?
Once you've uprated your life insurance and written your will, you
might want to investigate one of these (or similar):
http://www.ebay.co.uk/itm/170722896560
If I remember my time in the lab right, that's good for up to 500V and
several hundred mA. 400mA probably. Overkill for what you want, but
might enable you to get your project off the ground while you figure
the ins and outs of an optimal power supply.
Usual caveats apply - no connection with the seller, nor
responsibility if you use one of these things to toast yourself,
ignite your workshop etc.
Jon.
Nick
Electrophoresis supplies are great general purpose bench HV PSUs...
Warning: Serious danger of DEATH
Dekatron42
> might want to investigate one of these (or similar):http://www.ebay.co.uk/itm/170722896560
>
> If I remember my time in the lab right, that's good for up to 500V and
> several hundred mA. 400mA probably. Overkill for what you want, but
> might enable you to get your project off the ground while you figure
> the ins and outs of an optimal power supply.
Thanks Jon, but I already have a pair of these:
http://www.vintage-radio.net/forum/showthread.php?t=54336 good for
500V@500mA, -200V@100mA and -200V@1mA which I use.
I asked for help in finding a schematic drawing on the UK forum for
them as they did not work properly. I accidentally found the error
while checking how they were designed. There was a dry joint on one of
the wires to the thyristor controlled rectifiers so the high voltage
regulator were only supplied with a half wave rectified voltage when
the dry joint was open circuit, which happened intermittently. The
same error was found in both of my units so I guess that the same
person had been soldering both of them and made the same mistake. The
dry joint was on the gate connection to one of the thyristors
rectifying one of the sides in the bridge rectifier.
I got a reply from my contact in the transformer business today where
he said that he could help me to either rewind or wind a new
transformer for me. I did not know that he could make switched
transformer too, I have just asked hom about normal E-core transformer
earlier. He also told me to check the "Power Integrations website"
which I will do, I guess it is this one: http://www.powerint.com/ ,
I'll read what they have to say and check if they can design a whole
PSU for a small cost.
Thanks all for all of your input!
I'll report back what I find and also what design I will use.
Nick
Motala to do the Vätternrundan a few years ago. Nice area.
Nick
Nick
design tools at http://www.poweresim.com/
Selecting flyback as the base design, 450V @ 0.05A output, it'll
generate a complete design with all component values & tolerances etc.
Great fun just to play with!
Nick
John Rehwinkel
You're not using the Topswitch part any more? I hadn't realized your transformer was an autotransformer.
- John
David Forbes
I decided to make a design that would run from 12VDC, since it's more
versatile. And it's not an autotransformer, but a regular flyback supply
with primary and secondaries.
This 475V, 50mA flyback seems relatively easy.
Dekatron42
> Motala to do the Vätternrundan a few years ago. Nice area.
>
> Nick
Correct, close to the center of town!
Dekatron42
> design tools athttp://www.poweresim.com/
I tried Poweresim but I soon found out that it did not work well as it
totally missed to assign the proper voltages for the electrolytic
capacitors at the output! It suggested 400V working voltage when it
designed the power supply for 475V so you should really check all of
the components before trying to build something that it has designed.
It was good otherwise and it shows you how a lot of examples.
I've just downloaded the software from "Power Integrations" but I have
not had time to run it yet. I'll report back when I have tested it.
Nick
> > On a more serious note, I think you should look at the wonderful
> > design tools athttp://www.poweresim.com/
>
> I tried Poweresim but I soon found out that it did not work well as it
> totally missed to assign the proper voltages for the electrolytic
> capacitors at the output! It suggested 400V working voltage when it
Its not wrong - its just how you read it - if you click on the
components (C15 & C16), you'll see that it has X2 or X3 at the end of
the description - this means use two or three in series - this is
simply a limitation of the DC ratings of the electrolytic capacitors
that are currently in the component database - the schematic is fine -
you can replace the two series caps with a single higher rated one if
you so wish. You can ask them to add some 500VDC caps if you want!
HTH
Nick
Dekatron42
> components (C15 & C16), you'll see that it has X2 or X3 at the end of
> the description - this means use two or three in series - this is
> simply a limitation of the DC ratings of the electrolytic capacitors
> that are currently in the component database - the schematic is fine -
> you can replace the two series caps with a single higher rated one if
> you so wish. You can ask them to add some 500VDC caps if you want!
Thanks for that clarification, I did not spot that as they were drawn
as a single capacitor in the schematic. They should have added a note
or something as just putting three electrlytics in series without
voltage dividing resistors is not that good.
I'll check it more now that I know of that aspect!
Charles MacDonald
> Once you've uprated your life insurance and written your will, you
> might want to investigate one of these (or similar):
> http://www.ebay.co.uk/itm/170722896560
I see these on e-bay regularly, I bought one of another brand, but am
waiting to have an application where I can test it to see if it has much
ripple/Noise. Powered right up and has a current limiter built in. 400V
on my VOM no problem.
They sell for much cheaper than a regualr 400Volt radio bench supply.
--
Charles MacDonald Stittsville Ontario
cm...@zeusprune.ca Just Beyond the Fringe
http://users.trytel.com/~cmacd/tubes.html
No Microsoft Products were used in sending this e-mail.