Voltage Multipliers (for our tube projects...)

[johngovan1234]

This one Sir. The voltage keeps on flactuating. It doesnt stabilize at certain voltage. Almost dropping 1mV every 2 seconds. My startup is 270vdc at the B+. Then it goes to 263vdc when the tube heats up. Then it will keep on fluctuating a milivolt every 2 seconds. Is this normal?

[thomasha]

Low frequency motorboating?

From your first post it sounded like it was oscillating between 270V and 180V, that would be really bad. Probably some faulty part, or missing resistor in the feedback network.

But milivolts sounds OK. There are PSUs with more ripple than that.

[vigilante397]

I use basically the same SMPS, my most common mistake is to leave off the 220k feedback resistor. Check to make sure you did everything right, and if so, like thomasha said wandering millivolts at a time is not a huge deal.

[duck_arse]

what transistor did you use, BC54x? photos of your build?

[anotherjim]

It might be normal if it's only a few mV -  but any electro cap on the HT is under the stress of high voltage. It could be you have one or more that isn't 100% formed, or a bit leaky anyway. I would leave it on power to soak for a while. An hour or so.

[Kevin Mitchell]

Thought I'd point out that these boost regulators are now commonly available on eBay for pennies. Described as:
"8~32V to 45~390V DC-DC Boost Converter ZVS Step-up Booster Module Supply"
Although it says it's good for 40W output I have found that 25W is a more reliable maximum for continuous operation. Above this the transformer gets hot.

The listing had been taking down. But I wanted to chime in on this. I've bought two of them. Didn't realise they installed one of the HV capacitors backwards and it blew up in my face. Took a whole day to air out the workshop and for things to smell right again. Be weary when using the cheap boosters.
This was the image from the listing and what I had received;

[Rob Strand]

Be weary when using the cheap boosters.
This was the image from the listing and what I had received;

The EI cores are going leak out more field than a shielded inductor.
However, they are probably better than an unshielded bobbin core.

ZVS looks like something to try.

I came across this interesting boost converter variant over the last 6 months or so,

It uses an autotransformer instead of a straight inductor.  This lets you get better efficiency when the voltage conversion ratio is large, like it is for a tube supply.    It's perhaps possible to adapt that idea to the 555 circuit.

The down side is you have to make you own coil.   Parts for making your own shield inductor aren't so widely available.

[Kevin Mitchell]

I'm set with HV pedal designs for a little while. I nearly finished my Tube-O-Vibe and was sorting out the on-board SMPS after success with the cheap booster - or the one that didn't have a backwards capacitor. I'll dig back into it eventually when the PTSD wears off.

[DaveLT]

That cheap booster uses a ZVS with a flyback so it doesn't use a straight inductor... however its voltage jumps about a lot!

[vigilante397]

So... what's the deal with the 680k 1W resistors? Even if it were just one that should be way less than 1W of dissipation, a 1/4W should be fine there.

[Rob Strand]

So... what's the deal with the 680k 1W resistors? Even if it were just one that should be way less than 1W of dissipation, a 1/4W should be fine there.

I suspect the reason has more to do with voltage rating and reliability.   A 1W resistor typically might get a 400 to 500V DC rating.   A single resistor roughly meets the cap voltage of 450V (I don't have the original article so I don't know what max voltage they are targeting.)    Two 1W resistors would give good reliability.    However, three might be more to do with with the magazine's paranoia of people sub'ing parts and getting sub-standard parts in the open market.   Like if the resistors only met 250V then you probably would want three of them.   All speculation on my part but those types of things do go through my mind when I see HV stuff, especially when the device could be left on permanently and the resistor is permanently stressed. 

In 240V mains countries it's pretty common for switch-mode power supplies to not start-up when they get old (3 to 5 years).   This is usually due to resistors going high-resistance because they are subject to high-voltage for long periods.   (Nowadays the caps fail before that happens  )

[vigilante397]

I suspect the reason has more to do with voltage rating and reliability.

So what you're saying is that if one of my 200V supplies ever failed, the 1/10W 150V rated resistors could possibly be to blame?

I'm shocked! Well maybe not that shocked.

[Rob Strand]

So what you're saying is that if one of my 200V supplies ever failed, the 1/10W 150V rated resistors could possibly be to blame?

Probably not a good thing in the long term.

For that feedback resistor, I suspect the fault you would see is the voltage would rise-up over time.   The cause is the rising resistance.   You probably wouldn't notice for a while but at some point the voltage would get to a point where the circuit behaves weird or it could over-voltage the filter cap and pop it.  For an effects pedal that isn't on 24hrs a day it could take a while to fail.    As the voltage goes up it stresses the resistor more and more.  In theory it could arc through the resistor (on a small scale).   Apparently you can get micro arcing on resistors and perhaps outright arcing but I have not seen this myself in the low hundreds of volts.

[merlinb]

So... what's the deal with the 680k 1W resistors? Even if it were just one that should be way less than 1W of dissipation, a 1/4W should be fine there.

Yes 1/4W is fine there, since the voltage is also shared evenly between them.