Adjusting this schematic for a 300V output

[Jack Lynch]

Hi!

I would like to adapt the design found here for a variable 300V max output, instead of around 180–220V:

I understand that the output voltage is adjusted by the resistors R1 and R2, per the MAX1771 datasheet:

and that in this case a potentiometer is used to vary the voltage of R1. However, I'm not sure how the above equations are used to arrive at the voltage bounds specified in the schematic. Could someone explain? Using those equations, what value of R2 should I substitute in for the one used in the schematic to arrive at my desired maximum voltage of 300V?

Thanks! 

[Terry S]

It looks like you just pick an R1 and solve for R2. If you pick R1 at 10K, then R2 == 2 megohms.

300/1.5= 200.

200-1 = 199

10000*199 = 1990000

R2 = 1990000 or 2 megohms.

Terry

[John Rehwinkel]

I would like to adapt the design found here for a variable 300V max output, instead of around 180–220V:

Unfortunately, these flyback supplies can only boost a voltage so far in a single step, and getting 300V out isn't really going to work with that circuit topology.  The easiest modification is to run it with 30V in (with a separate supply for the MAX1771).  The second easiest modification is to add a voltage doubler on the output.

I understand that the output voltage is adjusted by the resistors R1 and R2, per the MAX1771 datasheet:

and that in this case a potentiometer is used to vary the voltage of R1. However, I'm not sure how the above equations are used to arrive at the voltage bounds specified in the schematic. Could someone explain? Using those equations, what value of R2 should I substitute in for the one used in the schematic to arrive at my desired maximum voltage of 300V?

With a little algebra, the equations from the data sheet become:

Vout = Vref(R2 + R1) / R1

In the circuit, the upper limit would be reached with VR1 at minimum, so the lower resistor would be 10k and the upper resistor would be 1.5M, so Vout = 226.5V.  The lower limit would be with VR1 at maximum, which would make the lower resistor 15k, making Vout = 151.5V.

Making R2 higher might lead to noise problems, so raising the output voltage is probably better done by reducing R1.  Reducing R1 to 7.5k would give a maximum voltage of 301.5V, but again if you did that with the existing circuit, it would need to run on a higher input voltage (if you tried to get 300V out from 12V in, you'd probably end up just burning some parts).

The note in the schematic is true: board layout is critical, not just for high efficiency, but to get it to work at all.  For beginning builders, I recommend the more robust 555 based circuit, even though it's less elegant and efficient, it's more tolerant to layout issues.  The fancier MAX1771 and MC34063 circuits are much less forgiving.

- John

[Jack Lynch]

Thanks! I learned my lesson about layout when I didn't heed the warnings and printed a supply recently with entirely unpredictable behavior. 😂

Both of those options sound doable.

Could you link to the 555 circuit you mentioned?

Thanks for your help!

[John Rehwinkel]

Thanks! I learned my lesson about layout when I didn't heed the warnings and printed a supply recently with entirely unpredictable behavior. 😂

Both of those options sound doable.

Could you link to the 555 circuit you mentioned?

There are a bunch of designs in the group files, the 555 ones are here:

https://drive.google.com/drive/folders/0BzH-L4OXjfbRTURWQlhqTXM5aXc

Also, Mike Moorees (AKA threeneurons) has several nice designs, including 555 based and with voltage multipliers for higher voltage here:

https://onedrive.live.com/?id=F9DB37B8211CE831%21142&cid=F9DB37B8211CE831

- John

[Jack Lynch]

Thanks! It might be better to use one of these.

If I were to double the voltage after the initial jump to 150–220, would that essentially require just another boost converter? Or is there some simpler method of DC voltage multiplication?

Thanks again!

[Best You]

I design a HV DC-DC PSU, you can see the datasheet as below. Max output 390V@100mA

https://drive.google.com/open?id=0B4j8u6jWqs0QV3dJaVpXUk9jMmFQNUhCOGdYbWRnRk9mVi1F

Jack Lynch於 2017年7月13日星期四 UTC+8上午3時02分31秒寫道：

[John Rehwinkel]

Thanks! It might be better to use one of these.

If I were to double the voltage after the initial jump to 150–220, would that essentially require just another boost converter? Or is there some simpler method of DC voltage multiplication?

Yes, the voltage multipliers I referred to are just diodes and capacitors connected in such a way as to convert an AC voltage into a higher DC voltage than you would obtain by just rectifying it.  Here's one of Mike Moorrees' designs:

The stuff on the left is the basic supply, and the stuff on the right (C4-C10 and D1-D7) is the voltage multiplier.  In this design, the multiplier has several stages, to allow getting 180V out from 2V in.  I'm guessing you might have 12V in, and wouldn't need as many stages of multiplication.  Note that the feedback network would have to be adjusted as well.  I'd use an adjustable one like this supply has:

- John

[Jack Lynch]

Thanks so much! This is all super informative. And especially edifying for someone like me, who hasn't yet had much experience. I'll be making more high voltage supplies in the future, so everything you've showed me will come in handy more and more as I continue.

I was also looking into boosting the voltage to 30V before using the original schematic, per your first post, and I found these boost converters:
https://www.amazon.com/dp/B01GRIQBRY/ref=asc_df_B01GRIQBRY5072525/?tag=hyprod-20&creative=395033&creativeASIN=B01GRIQBRY&linkCode=df0&hvadid=198076677096&hvpos=1o2&hvnetw=g&hvrand=13981072561463713758&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9006850&hvtargid=pla-319429339171

Do you think those would work? Or would its efficiency or rate impair the operation of the second circuit?

[Jack Lynch]

And would using a 30V input require any modification to the schematic besides probably buying some higher-rated components?

[John Rehwinkel]

While that would work, it seems unnecessarily complex. On the one hand, you're using an entire second boost power supply instead of a couple of capacitors and diodes. On another hand, if you're willing to buy a completed module, it's easier (and probably cheaper) to just buy a 12V-300V converter and be done with it. On the gripping hand, if you're going to need a power supply to run the thing in the first place, why not start with a 30V supply?

That said, the only modification needed to run on 30V in is to add some sort of regulation to provide a lower voltage supply to the control chip (whether it's a 555 or some other one, I don't think any of the usual ones can run on 30V). This can be either a dedicated voltage regulator (I think a switching supply would be silly here, so I'd just use a simple linear regulator that can accept the input voltage and won't dissipate too much heat), or a simple zener shunt regulator.

For the sake of completeness, I'll mention yet another method of getting more voltage step-up: instead of an inductor-based supply, you can build a transformer-based one, which allows you to get voltage multiplication from both the flyback effect and the transformer turns ratio (harking back to another recent discussion on this group). Many commercial supplies work this way, as the engineering effort and production costs of a custom transformer make sense with volume production. For hobbyist use, however, it can be considerably less practical.

Then again, you can get an off-the-shelf mains transformer that will give you enough voltage out to rectify to 300V easily. The advantages are that it's a very simple reliable circuit with few components. The disadvantages are that it's bulky, expensive, somewhat more dangerous, and more difficult to make into a variable voltage supply.

- John

[Jack Lynch]

Thanks so much! I feel like I'm learning all the stuff I should have known before trying to make this 😂

If I get to the point where I'm producing more of the design at some sort of scale, I'll look into using a transformer-based supply instead.

So is the maximum voltage with a 30V input roughly 2.5 times what it would be with a 12V input? I guess I can just look at the equations for a boost converter. Whatever the maximum voltage, wouldn't the high-voltage components have to be rated for that? I think in the linked schematic they're all 250V.

[gregebert]

I'd recommend experimenting with linear supplies first, before experimenting with switchmode (boost, flyback).

Linear supplies are well-behaved, even when they are abused, and all you really need for debugging them is an inexpensive DMM.

For switchmode, a scope is an absolute must, and preferably a digital one. Lots of strange, and often unexpected, things happen when you tweak a switcher.

[Tomasz Kowalczyk]

Another problem with switch mode is PCB layout. I can't find any good guideline about this, but my simple 555 boost converter built on a prototype PCB was easily outputting 40mA at 200V (more than enough for most nixie applications) without radiators, reaching temperatures of 70°C. After I've made a PCB for it, same schematic and similar design it suddenly struggles to output half of that current! 

This Nixie Buffalo power supply looks great. I'm starting a project, which is a portable noval vacuum tube tester, and I'll consider it - I need at least 300V 100mA with regulation powered from a battery (probably a laptop battery).

[Paul Andrews]

This page has good layout guidelines. it includes the reasons, so you can figure out what applies to other designs: http://desmith.net/NMdS/Electronics/NixiePSU.html. The data sheets for some of the converter chips are also good sources of layout design. I'm at the point in my own learning curve where I want to experiment with this stuff, so I'm trying to figure out how to do that whilst not being hit by the layout restrictions. Breadboard inc is clearly out of the question because of stray capacitance, inductance and poor contact resistance. I've built someone else's design on perfboard, and it has behaved very well - it at least produces enough voltage and current to drive what I needed (not nixies). So I am hopeful with some of the lower clock rate designs.

We will see :)

[Tomasz Kowalczyk]

Thanks, it gave me more knowledge about routing around the inductor and feedback trace. While I was using different chip, it still could potentially have problems with improper feedback routing.

Making a circuit on perfboard has one advantage - traces are usually made from wire or a healthy amount of tin, which have greater current capability than usual 35um PCB traces, even if they are 2mm wide. 

I have to redesign my PSU PCB to utilize a solid ground plane, that is also the key - my PCB was one sided (much easier to do at home than two sided). On one sided it is practically impossible to make proper ground routing.

[Dekatron42]

Someone once told me that you could ask for up to 450V output voltage from the small Nixies PSUs (1363 & 1364) that are sold at Tayloredge (www.tayloredge.com) but I've never asked for that myself, but it might be worth checking.

/Martin

[John Rehwinkel]

> Someone once told me that you could ask for up to 450V output voltage from the small Nixies PSUs (1363 & 1364) that are sold at Tayloredge (www.tayloredge.com) but I've never asked for that myself, but it might be worth checking.

Yes, that too uses a voltage multiplier.

I don't have a link to the document, so I'll try appending it.

- John