New power supply design

[taylorjpt]

Just finished my 50W HVPS design, should be able to push it to 75W at 36V in.
150-200V out.
5-36VDC in.
200V/0.250A at 24V in.
Same form factor as 1364.

This was designed to power my nixie Sudoku project: 91 direct drive IN12Bs (2.5mA/tube) with an IR grid input system.

jt

[John Rehwinkel]

> Just finished my 50W HVPS design, should be able to push it to 75W at 36V in.

Nifty!

> 150-200V out.
> 5-36VDC in.
> 200V/0.250A at 24V in.

Okay, all sounds good.

> Same form factor as 1364.

Okay, now I'm impressed. I suspect some deep magic in your magnetics.

Planning on selling these beasties? They'd be great for small vacuum tube projects.

- John

[Michel van der Meij]

> Same form factor as 1364.

 

That is a jaw dropping claim. It is like 6 times more power than the 1364! I'd like to see a picture of that one.

The 1363/64 are nice units but at a half load they do tend to make quite some audible noise, which is a bit annoying. I swapped mine with a simple NE555 HV supply and no noise at all.

 

How is that for your 50W unit?

 

Michel

 

 

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send an email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/1AC571FE-1DE4-4C24-BEB7-1ABADA3C7877%40mac.com?hl=en-GB.
For more options, visit https://groups.google.com/groups/opt_out.

 

[blkadder]

I'd be interested in purchasing one of these as well.  A picture would be awesome.

Ron

[taylorjpt]

Proto/Test Bench:  http://ppl.ug/SJYXP4kw6u4/
50W test data:       http://ppl.ug/dwkhS2L32kI/

85% efficiency from 20mA to 260mA at 204V out and 24VDC input.  It has the same adjustment scheme as the 1364 as well as enable logic.  Very quiet and very low EMI since the transformer is a closed torroid.  A difference from the 1364 is there is no true shut-down, i.e. the output will track the input when the power supply is disabled.

The footprint is the same as the 1364.  The Z-axis is a little bit thicker and will be potted in a thermally conductive epoxy with a threaded insert in the center for mounting a heatsink. (At 85% efficiency and 50W out, there is 9W of heat to pull out).

jt

[taylorjpt]

77W test data:  http://ppl.ug/A2EYKy2BwmY/

383mA/201Vout at 36V input (48V abs max);  Had to pre-load my load bank to get over the 250mA limit.

jt
 

[David Forbes]

John,

Congratulations on getting so much power out of a tiny gizmo.

It's nice to see what someone who does this for a living can achieve.


--
David Forbes, Tucson AZ

[Nick]

On Monday, 12 August 2013 05:45:17 UTC+1, taylorjpt wrote:

77W test data:  http://ppl.ug/A2EYKy2BwmY/

383mA/201Vout at 36V input (48V abs max);  Had to pre-load my load bank to get over the 250mA limit.

Vout = 201.596350 (9 digits) - do you *really* have a 30-bit A/D on that :)

Seriously, very nice results. As you point out, with compact units, power dissipation  can be an issue...

[taylorjpt]

I use a mux in front of an HP34401 multimeter that auto ranges and so can read from 0.1uV to 1kV. It was just easier to display 6 decimals on the quickie UI I wrote tonight.

[Dekatron42]

Nice!

 

I'll definately buy some if you will sell it in the future (near future I hope)!

 

Will it be possible to get 300V from this power supply without a voltage doubler?

 

From what you write, and looking at the picture, this looks like a standard switcher which doesn't use a transformer - will you make a transformer based power supply with isolated output too? I mainly try to use isolated power supplies in my circuits since I won't have to worry about a lot of stuff then.

 

/Martin

[taylorjpt]

Its only 20 bits of accuracy, the extra bits are a result of the slope conversion that the meter uses.

[Jan Rychter]

Those are impressive numbers.

As I'm designing my own supply (getting about 90% efficiency for a boost from 12V to 180V at up to 80mA, will share the design once done), I'd be interested in knowing more about the design? Are you willing to share some information?

The biggest design problem I encountered was that the Nixie load can be highly variable (think dimming, multiplexing, etc), so with a booster design you can't just operate in Continuous Conduction Mode (CCM) all the time, which is what most high-power designs do. What's worse, the transition between CCM and DCM (Discontinuous Conduction Mode) is poorly defined. This is why I decided to finally use a smaller inductance and stay in DCM all the time, even with full load.

Actually, is your design a boost at all? Or do you use a flyback transformer?

When you measure output power, do you just stick a meter and measure average current/voltage, or do you do something fancy to take ripple into account?

--J.

> --
> You received this message because you are subscribed to the Google Groups "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
> To post to this group, send an email to neoni...@googlegroups.com.

> To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/9944e216-7d04-4c1e-8a5f-8e37f5c4f687%40googlegroups.com?hl=en-GB.

[taylorjpt]

>Actually, is your design a boost at all? Or do you use a flyback transformer?

I suppose you mean single winding inductor boost which is a particular variation of a flyback converter, i.e. any converter where energy is stored in the inductance out of phase with the energy released to the load.  The 1364 used a coupled inductor, aka transformer, whith no direct path from the input to output so that the output goes to zero when not switching.  To pick up a few percentage points of efficiency on this design I went with a tapped inductor.  The main reason for either of these two variations vs the single inductor has to do with the voltage stress on the main switching transistor:  A single inductor flyback exposes the switching transistor to the output voltage which is traded for by an increase resistance.  By using the transformer or tapped inductor, the switching transistor drain only sees a voltage during the off time of Vout x (Nprimary/Nsecondary) or Vout x (N1/N2) respectively where N1 is the turns from Vin to Vsw and N2 is the turns from Vsw to Vout.  Since the output voltage is stacked on top of the input and the input voltage portion is delivered at near 100% efficiency, switchers with high input voltage benefit more from a tapped inductor than those expected to operate at low input voltages.

 

 >When you measure output power, do you just stick a meter and measure average current/voltage, or do you do something fancy to take ripple into account?

The meter I use (HP34401A) has a slope converter which intrinsically averages the readings when in DC mode, however there is also an RMS AC function as well, it just wasn't part of my UI as I used a scope to watch that real time for monitoring stability.  The p-p ripple target for this design at 75W load is < 1V at 200V out (<0.5%).

jt

[taylorjpt]

>Will it be possible to get 300V from this power supply without a voltage doubler?

I am designing this one with >400V out without an external circuit.

>...will you make a transformer based power supply with isolated output too?

I don't understand the question about isolation. By using a DC input, mains isolation comes at the expense of lower system efficiency by the decision was made to let someone else deal with UL and focus on the high voltage side.

jt

[Dekatron42]

Really nice with >400V output voltage!

 

Regarding isolation, some circuits are easier to build if the different high voltages needed are isolated from the low voltages (digital and/or analogue) and in the case with Russian Dekatrons like A-201 (Polyatron) it is a lot easier to design a circuit with isolated power supplies than those that use a common ground.

 

Isolation can of course be made with optocouplers for the other signals but in some circumstances the original design used transformers as isolation between the different voltage levels to make sure that it was easy to drive the Dekatrons/Polyatrons. Right now I am using small Traco switchers to isolate the high voltages from the low voltages and also optocouplers in some cases in combination with this.

 

There might be other ways of doing this but I am trying to follow the original designs to get things to work before I start to experiment with other solutions, and in some caes I want to stick to the original designs and not use modern components except for the power supplies which I don't regard as part of the original circuit (I also want to minimize power usage and weight).

 

/Martin

[taylorjpt]

>Regarding isolation, some circuits are easier to build if the different high voltages needed are isolated from the low voltages (digital and/or analogue) and in the case with Russian Dekatrons like A-201 (Polyatron) it is a lot easier to design a circuit with isolated power supplies than those that use a common ground.

Got it... It should be possible to get any isolated voltage or set of voltages by using a secondary that is ratio metrically related to a sensed output winding.  For a variable input voltage a grounded secondary (vs a tapped winding) is needed or else the output will vary as the input changes unless you subtract the input voltage from the feedback equation.  For instance, to get 600V on an isolated winding you can develop 200V across a 60 turn secondary that is fed back to the controller and then use a floating 180 turn secondary to generate the 600V output.  As long as there is minimal loading on both windings, line and load regulation is actually pretty good.  This is used in a lot of newer AC adapters to get rid of the optocoupler by simply measuring the voltage on the primary during flyback to get a sense of the secondary voltage.   I'll give that a try when my boards come in next week.

jt

[Dekatron42]

Fantastic!

 

Having the possibility of getting a set of voltages would be very nice, that way all voltages needed for driving a Dekatron/Polyatron like guide bias voltage, guide stepping voltage, reset/set pulse voltage could be had from one small transformer (and possibly the voltage for driving a Nixie too - Anode and bias voltages). For ordinary Dekatrons all of these voltages have the same ground reference and can usally be solved by simple resistive voltage dividers, but with some types, like the Polyatrons, the circuit is simplified by having separate grounds.

 

I think the market for such a power supply is not that big but it might mean that more people use Dekatrons to build clocks (which of course will drive the price up on the usable Dekatrons like the Russian A-101 to 103 and the European/american GS10C/6476 and similar selector types).

 

Maybe off-topic, but I am thinking of what I would like to build - a nice small bench PSU. For anyone who wants to build one or more small bench power supplies for experiments with Dekatrons with good regulation it is always welcome to have separate voltages available on the "primary" side of the pass element so you don't need to cool off too much heat in the pass element. Having the possibility to change range for the input voltage to the pass element reduces the size of the heatsink you need. Making a 0-600V powersupply and using only 5V output voltage means that you drop 595V across the pass element at whatever current you draw and that amounts to a lot of heat, but with a switchable range of "primary" voltages you will have to cool off less heat.

 

/Martin

[taylorjpt]

16V/50W http://ppl.ug/g3NZo2iSQrs/
24V/77W/1uF http://ppl.ug/05s0AvqBL1U/
24V/77W/22uF http://ppl.ug/ddanLbw4wRc/

The transformer design is finished. I am now getting a solid 50W out with a 16V input.

The output ripple is a function of the output capacitor, with the 1uF on my test board I'm getting 5Vp-p at 77W out. The final value will be 4.7uF to get near my target of 0.5% ripple at overload but external C can be added as well as shown in the example with an external 22uF electrolytic.

[Dekatron42]

The links doesn't work for me, so I took the liberty to edit them so they work for me:

 

16V/50W                http://ppl.ug/g3NZo2iSQrs/ 

24V/77W/1uF         http://ppl.ug/05s0AvqBL1U/

24V/77W/22uF       http://ppl.ug/ddanLbw4wRc/

 

/Martin

[taylorjpt]

24Vin, 80W out at 308V: http://ppl.ug/iOtLwdq3aFM/

This is scaling really nicely. The board design is setup to do up to 450V out on board or drive a second output winding for higher ratio metrically related outputs. Planning to use this driver for an 8kV project as well.

jt

[John Rehwinkel]

> This is scaling really nicely. The board design is setup to do up to 450V out on board or drive a second output winding for higher ratio metrically related outputs. Planning to use this driver for an 8kV project as well.

Sir, you had my curiosity, now you have my interest! I'm currently building a monoscope project* that will need several supplies of a few kilovolts. I'm also working on a field emission electron microscope that will want 9kV or so. Commercial regulated high voltage supplies are too expensive for my tastes. Rolling my own is going more slowly than I like (that CRT power supply board is my design, but it has ... issues). So I'm quite interested in your engineering here!

* http://www.flickr.com/photos/23304890@N00/7268621172/in/set-72157629911800685

- John

[Dalibor Farný]

I will be also interested in buying some!

Dalibor Farný
http://dalibor.farny.cz

sent from Samsung Galaxy Pad

Dne 13.8.2013 9:33 "taylorjpt" <j...@tayloredge.com> napsal(a):

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send an email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/c0d19a0b-51b0-482a-af4b-46ad7e662e34%40googlegroups.com?hl=en-GB.

[Nick]

John - I've got several CK1414s variants myself - glorious beasts - what mask set do yours have?

Not had much time to do anything with mine yet - they are all alphanumeric with a few special characters - probably used for airport information displays etc.

Nick

[John Rehwinkel]

> John - I've got several CK1414s variants myself - glorious beasts - what mask set do yours have?
>
> Not had much time to do anything with mine yet - they are all alphanumeric with a few special characters - probably used for airport information displays etc.

Yeah, mine also has alphanumeric with some symbols (there are some pics of the target in the same flickr set). I'm thinking of making a CRT-based FLWK out of it.

- John

[Marcin Adamski]

That would be awesome. Please add me to the list of buyers.
Marcin

[Tidak Ada]

I am higly interested in some of these supplies for some dekatron projects,
but I am not yet aware what voltage I need.
May be 600V for an EZ10-B and 450-500 V for the following stages So is it
possible to make a tap on the secondary of the transformer or soemthing like
that to get the two voltages. Or is it better to use the 1663/1664 with the
suggested voltage multiplier?
Further your new development would be interesting for supplying a Philbrick
K2xx Operational amplifier (600V needed).
So I like to me informed about the developments and possibilities.

eric

--
You received this message because you are subscribed to the Google Groups
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to neonixie-l+...@googlegroups.com.
To post to this group, send an email to neoni...@googlegroups.com.
To view this discussion on the web, visit

https://groups.google.com/d/msgid/neonixie-l/520A6877.6000905%40gmail.com?hl
=en-GB.

[taylorjpt]

If you have 600V out of a 160T winding, to get 450V you would simply put a tap at 120T. Completely isolated outputs work the same but instead of a tap, you just have a separate winding for each output. This also points the way to get a low impedance feedback network (Less noise pickup) with a predictable frequency response without dumping too much power in the divider: You can set the output voltage for the previous example with a tap at 6T for a 30V feedback so for a 30K = 1mA feedback divider only 30mW needs to be wasted where feedback at the full 600V would consume 600mW for the same 1mA or 12W for the same 30K. This becomes an important consideration as the input parasitic capacitance of the regulator feedback is a constant, so increasing the feedback impedance to match the output voltage means the network low pass corner frequency moves lower as the output voltage increases.

Using a toroid core has a big advantage for a multi winding configuration with regard to the voltage ratio relationship: Unlike the open magnetic path of a drum core which can have fractional turns, toroids can only have an integer number of turns because only wires that pass through the hole at the center count as that section of the wire is the only one surrounded by a circulating magnetic field. Try it on a 60Hz transformer by unwinding a turn and observing there is little voltage change (relative to pulling the wire out of the hole) as the wire is pulled away from the outside of the core. In the case of an E-I core, you can have 1/2 turn increments as there are two holes. If you want a 1/4 turn from an E-I or a half turn from a toroid, you can drill a hole at the core center and pull the wire through it.

As long as there is some load on each of the taps, the output voltages track fairly well: No load on un-sensed windings turns the rectifier/capacitor on that winding into a transient peak detector.

jt

[Dekatron42]

All this seems easy in writing, but for people like me who are not used to making switching power supplies it would be very kind of you if you could help me (us) with a complete solution when it comes to which toroid core to use and how to wind it.

 

I don't know what your ideas are on how to produce or sell this switcher, but if you could either sell a kit with the switcher as such and a toroid which could be used with instructions on how to wind it to get the voltages each of us are interrested in and/or sell one or more prewound toroid(s) with the most common voltages for Dekatrons (within reasonable limits) it would be a great help.

 

Making a multi-tap toroid with some voltages that are usually used for Dekatrons/Nixies with some extra windings to add or subtract voltages to be able to make adjustments up or down from the standard taps would be a good idea I think.

 

For those who want to use really high voltages past 1KV there will probably have to be special solutions to ensure that the isolation requirements are met properly.

 

I hope that you will find a good solution for this that will make it easy to use this switcher and that it will sell well!

 

/Martin

[taylorjpt]

For custom voltage supplies that you can do as roll-your-own projects, I have selected a ferrite UU core set that is good to 50W when gapped properly and allows for easy winding off the core (It was doing >92% eff at 30W due to the lower core losses of the ferrite). I am working on a multi-section bobbin for this core set like I used on a 30kV neon supply I did in the 80's that can keep the turn-turn voltage stress down to a manageable 30V. This core set is significantly larger than the toroid I'm using in the self contained unit (And significantly less exotic) and is intended as an external transformer attached to a thermally potted driver only section: Unlike the 1364, the 1474 brings out the switching nodes to allow for an external transformer. A half size version can be made by inverting one of the U cores and cutting off the legs (Still looking for an off the shelf "I").

Using the UU cores, VERY high isolation factors can be achieved without exotic creapage mechanicals.

jt

[Oscilloclock]

I've used off-the-shelf and readily-available TDK cores in the past. I recently wrote a 4-part series on how to design, choose a core, and wind your own flyback HV trafo. By no means a professional treatise, and certainly the trafo won't have jt's very high efficiency, but it should be enough to experiment with!

http://oscilloclock.com/archives/329

- Aaron

[taylorjpt]

I just sent out for PWBs for a new version of the 1364 that uses a controller that operates much more quietly; will be getting those beginning of next week. This controller worked really well on the 75W converter and operated very quietly from 0 to 383mA @ 200V as well as 0-250mA @ 300V.

If they work as expected, I'll be phasing out the 1364X and 1363A for the 1364A and 1363C. The form factor is exactly the same but the input range changes from 2-16V in to 3-36V and the adjustment range changes from 150-200V out to 150 to 300V out.

jt

[Dekatron42]

Wow, that is really nice! I'm keeping my fingers crossed that it works as expected!

 

/Martin

[taylorjpt]

Got the 1474 (75W) and 1364 (15W) boards this week, going to take advantage of the 3 day weekend to play with power supplies. Also got a mini-lathe last week to cut bobbins for my 8KV project on the UI core using the 1474 as a driver. This should prove to be a fun week end so long as my wife does not think of too many things for me to do!

[Dekatron42]

Just put your wifes tasks at the end of the list, after the 1474 & 1364..... ;)

 

Really looking forward to the results!

 

/Martin

[Adam Jacobs]

... the end of the list?!? I've got items on the end of my to-do list that have been there for years and years. The most frequent way that items on the end of my list resolve themselves is by becoming obsolete/irrelevant. ;)

Nah, the key is to knock out one or two of the items that she wants done and then trick her with flowers and/or chocolate. In the ensuing bafflement/surprise, sneak off to the lab/garage/shed/office and complete the fun tasks. :)

-73 Adam W7QI

--

You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send an email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/92892c7c-b1af-4064-8921-cbcfb972b83c%40googlegroups.com.

[Nicholas Stock]

There is an end to the list? Nobody told me that...i don't believe it...;-)

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/521FA9E6.70405%40gmail.com.

[John Rehwinkel]

> Got the 1474 (75W) and 1364 (15W) boards this week, going to take advantage of the 3 day weekend to play with power supplies.

Excellent plan!

> Also got a mini-lathe last week to cut bobbins for my 8KV project on the UI core using the 1474 as a driver.

Good idea - I'm thinking of getting a mini-lathe myself for various purposes, but hadn't thought of that one. Do you have a coil winder? Or are you toying with the idea of using the lathe as a winder?

> This should prove to be a fun week end so long as my wife does not think of too many things for me to do!

I know how that goes!

- John

[taylorjpt]

I have been known to use the decoy chore, something she can see that makes her think that the rest of the list must be on the plate... you do not fix something hidden by drywall as the decoy chore, no matter how mission critical it is!!!

I have been using this coil winder one for decades, with a motor free-wheel button, turns counter LCD and turns reset button on the front panel. It is a direct drive stepper winder and the movie shows a coat hanger being wound up on a 3/8" rod (I used to design a lot of ribbon wound transformers).

[taylorjpt]

http://www.tayloredge.com/Projects/CoilWinder/coilwinder1.mov

[taylorjpt]

Got a pound of thermally conductive epoxy coming, $300 a pound. I think if I fill the empty spaces with old nails before I pot the modules I can stretch it out! Also thinking of putting in some vertical strips of PGS.

[taylorjpt]

The new board is built up with the final transformer design:  36V in, 306V out @ 100W, 88% efficiency.  Destroyed my old HVPS load bank, building a better one this week end.  My epoxy will be in week after next and I can use the casting tooling from the 1329 module.  Based on the calculations for the epoxy I should be able to hit about 5C differential, module internal to heatsink mounting face at 100W out.  The first potted module will get 6 thermocouples built in to measure the thermal performance.  The module has a 0.288" dia SS stand off molded in with a #6-32 though hole at the center.  Dimensions are 1"x1"x0.563" with 0.063" radiused corners.  The connections are the same format as the 1364, with a new pattern.

Anyone have an application for this much power that can test a prototype?  I'll have 5 of them from the proto run.

jt

[taylorjpt]

One requested, 4 left.

[taylorjpt]

Thermal potting on the 1475 is to remove 10W of heat, while delivering 100W at 150-300V out in a 25.4x25.4x13mm package. My price target is currently $25.

[Dieter Waechter]

That's really good! Well done!
When will these be available, maybe we can use it for another revision of
the IN-18 clock?
Do you have an efficiency curve for 12V input already?
And what is the temperature rise at 12V in / 6.3 Watts out without
additional cooling?
And not to forget: Is it 100% noiseless at all load conditions? (that's the
most important point for us)
I remember I tested some of your old modules (forgot the number(s)) and
these were very noisy.
Thanks
Dieter

[taylorjpt]

The 1474 design is for high power and high voltage (Up to 75W and up to 600V on board, no multiplication). It will be available in about 2 weeks in the 150-300V adjustable version.

100W, 306V: http://ppl.ug/iPAmeWyGAbI/
New load bank: http://ppl.ug/ZDU68hsXQik/

I am also doing a refresh of the 1363/1364 based on the new controller in the same time frame for about double the power (20W) of the original with an extended input range: 5V to 36V. The 2W+ efficiency should be in the high 80s/low 90s. This uses a significantly different control scheme and will be silent at all input/output conditions. The pricing on the refresh will be the same as the original.

[blkadder]

This will be a great product.  I am excited to see it when you make it available.  I have a clock I would like to put together for my daughter, and one of your power supplies would be a great addition to the unit.  Just curious if anyone else has volunteered for the other test units.

Ron

[John Rehwinkel]

> The 1474 design is for high power and high voltage (Up to 75W and up to 600V on board, no multiplication). It will be available in about 2 weeks in the 150-300V adjustable version.

I'm waiting for the 600V version. There were some cute little fixed 600V power supplies on ebay a while back, but I've used all the ones I bought, and haven't seen any more.

> New load bank: http://ppl.ug/ZDU68hsXQik/

What are the objects (that look somewhat like 9V rectangular batteries) in front?

Is that slab of aluminum sufficient heatsink? I suppose you could blow a fan on it, or even bolt on a finned heatsink if need be.

- John

[n1ist]

They look like relays, likely controlled by the thumbwheel switch in front, to set the value of the load bank.
/mike

[taylorjpt]

Yes, relays with snubbers across the contacts to prevent arc damage when the contacts open. I have a USB controller board coming in a couple weeks to remote control the load bank(s) that also has a K thermocouple converter for plate temperature and protection. I like the resistors/relays because it doesn't add a second active circuit while I'm testing. I could use solid state switching, but I like the click-click of relays.

Its a standard 1/2/4/8 load bank using 15 8K/25W resistors (11W per step at 300V in). The plate is a 1/4" 6061 spreader with 4 1/4-20 holes for bolting a heatsink, I currently am testing with a 1" fin block about half the size of the spreader plate greased and just sitting on the plate with a fan on it for the last 12 hours: 34C rise above ambient with a 100W heat load measured at the plate center.

[taylorjpt]

Two proto units claimed, 3 left. One is going for a high power battery (car) operated tube amp that can't stand the hum induced by a standard inverter.

[taylorjpt]

Load bank controller: http://ppl.ug/IN9_UfuijAg/

[taylorjpt]

Ran the first 1474 at 306V out @ 93W (75W design limit) for over a week with convection cooling with the transformer separated from the driver by an inch to limit the power density, should be great with the mounted transformer and potting.

First cut of the1364 refresh: 10-25V in (48V abs max), 303V out @ 25W 82% efficiency and no audible noise. Building up a low power load bank for 0-20W over 16 steps so I can test over more input output conditions. Ready to release beta test samples in a week.

[Dekatron42]

What happens if the load varies or if the load itself introduces hum/noise on the power line? I am thinking of how multiplexing of Nixies and/or loads with varying power draw over time will affect the noise or regulation? I don't think this will be a problem, but thought it would be good to have an answer as real life situations are not the same as using a resistive load bank.

 

/Martin

[taylorjpt]

I have digitally controlled loads specifically to test transient response and phase margin. The relay controlled resistive step loads are simply for generating efficiency curves and measuring thermal performance.

Since the meters integrate everything, you can't use them to measure damping so that has to be done on an oscilloscope: you can infer stability by measuring the frequency of the ripple, but with PFM control schemes, normal operation can look like instability.

[Chris Stalin]

Awesome work John,

Looking forward to the new updated horizontal HVPS.

Take care

On Saturday, August 10, 2013 9:50:57 PM UTC-6, taylorjpt wrote:

Just finished my 50W HVPS design, should be able to push it to 75W at 36V in.
     150-200V out.
     5-36VDC in.  
     200V/0.250A at 24V in.
     Same form factor as 1364.

This was designed to power my nixie Sudoku project:  91 direct drive IN12Bs (2.5mA/tube) with an IR grid input system.

jt

[taylorjpt]

Loadbank controller all programmed, ready to integrate into box: http://ppl.ug/BGzyuOb0FxE/

[Petey]

Hello,

I'm currently designing a compact tube amplifier with a university club.

We've been looking into your 1363/1364 as a possible option to power our amp, but these new designs seem better (as we really need a bit more current).

Could we possible get/buy one of these 'beta test samples'?

Thanks

[taylorjpt]

send me an email, you are tester #3.

[blkadder]

Not sure how to send you an email from here..  ;)

Ron

[taylorjpt]

Ron: j...@tayloredge.com

[taylorjpt]

Ron: j...@tayloredge.com

[a zmijewski]

works!

--
You received this message because you are subscribed to the Google Groups
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to neonixie-l+...@googlegroups.com.
To post to this group, send an email to neoni...@googlegroups.com.
To view this discussion on the web, visit
https://groups.google.com/d/msgid/neonixie-l/5e791e17-6b57-422c-9c3f-c3b6fef3534c%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Chris Stalin]

Hey John,

Curious are the 600V model power supply complete?

Thanks

[Chris Stalin]

Any updates John?

Thanks


On Sunday, September 22, 2013 9:51:18 PM UTC-6, taylorjpt wrote:

Ron: j...@tayloredge.com