SMPS for tube pedals

[snailspacejase]

Hi Guys
I'm having a go at building an smps to get decent plate voltage for a tube based pedal.
Ive tried a pre-made nixie circuit from ebay and also built a vero board design kindly posted hrere by Marcos-Munky.

https://postimg.cc/GBMv10YN
https://postimg.cc/NyGRVGym

Both circuits give good voltage but are too noisey to use. I've use a simple audio probe (coupling cap into battery amp) and a simple valve pre amp. So a couple of questions
What sort of resistance do you think would simulate a safe load?
Would I be able to filter out the noise with another inductor/resistor/cap circuit?
Has anyone had any luck with pre-made smps circuits for this application.

Thanks in advance for any help

[GibsonGM]

Welcome to the forum!

I've used both of them with good results. What preamp do you want to run with one of these power supplies?  You can add additional filtering if you need to.

You have tried one or both with a tube amp, and it made noise?  What kind of noise?  Both should be running at a frequency above human hearing.

[snailspacejase]

Thanks.

I have made a version of this (which is from vigilante397's trace of a Kingsley Harlot). Only I've omited the first gain stage and gone straight from my input to plate 1 via a 68k and 1M to ground.
My idea was to drive the pedal a seperate clean boost.

I got a high pitched whine (which my dog disapproved of) when using the valve circuit. I got a lower frequency buzz, maybe more like 200hz, when I tried the audio probe straight to the output of the smps.

[Yazoo]

My experience was repairing a Blackstar HT pedal. The pcb where the SMPS section was was quite charred in places and couldn't be repaired directly ( I tried and failed!). I tried using an eBay SMPS, an NCH6100HV to power the valve. This did work but, like you, I found it was noisy. In the end I built a separate pcb using the original Blackstar HT SMPS design and fitted that and it now works fine  - no noise. Try looking at that if you can't get your existing SMPS to work the way you want.

[GibsonGM]

Or add filtering before either circuit. 100R series, 470u to 1000u to ground, possibly would help.  Even do it in more than 1 section...

[Rob Strand]

My experience was repairing a Blackstar HT pedal. The pcb where the SMPS section was was quite charred in places and couldn't be repaired directly ( I tried and failed!). I tried using an eBay SMPS, an NCH6100HV to power the valve. This did work but, like you, I found it was noisy. In the end I built a separate pcb using the original Blackstar HT SMPS design and fitted that and it now works fine  - no noise. Try looking at that if you can't get your existing SMPS to work the way you want.

As far as I'm aware, the input supply on the Blackstar converter should be 22V *DC* (unregulated).   The input diode has been added to the original circuit for reverse supply protection, not for rectification.    On the Blackstar amps the input supply is AC but it is converted to DC with a rectifier and filter cap before going to the converter.

There's many discussions on this forum about the NE555 converter, as use by Marcos-Munky.

[brett]

What sort of resistance do you think would simulate a safe load?
100 k ohms is typical (12AX7)

Would I be able to filter out the noise with another inductor/resistor/cap circuit?
If you have 2 or 3 stages already, the problem probably isn't in the electronics per se.  Layout is quite important, especially with respect to inductors and the high current lines.  Sub-harmonics can lead to them buzzing like bees, either physically or electronically.

If all else fails, the step-up converters on AliExpress are only $5.

[crane]

Nixie supplies are indeed too noise because they are unregulated.
I saw a nice SMPS HV supply in Kingsley tube pedals and reverse engineered it. Made SMD and THT versions of it.
Not sure I can't get into trouble if I share the schematics but it's basicly boost topology of MC34063 using 220uH coil and IRF730A or IRF840 as a switch.
Super compact pcb. Will check later/tommorow if I can post a pic

[Rob Strand]

Nixie supplies are indeed too noise because they are unregulated.

The NE555 design is regulated to some degree.  The transistor acts as the comparator and voltage reference.

Something not addressed yet is the use of:
- shielded inductors
- caps across the input supply

If you don't use shielded inductors there's a good chance noise will get into the audio.  To some degree you can distance the inductor from the audio wiring/circuit but there's practical limits.    It's possible to do a test to see if the inductor is the cause separating the converter from the audio.   That assumes everything else is good.   You need an input supply cap close to the converter.  Also if the inductor current rating is too low the inductor can saturate and that can promote noise from the inductor.

If you don't have the input cap close the converter, then distancing the converter with long wires can result in other forms of noise from the input wiring.   You need to get everything right.   Keep all the layout tight around the converter:  input caps, output caps, including diodes, inductor, MOSFETs - pretty much the whole thing needs to be tight.   Also don't let any audio run through the noisy converter ground.

There's actually many ways converter noise to get in.

In one of the older threads I went through some details on the switch frequency and switching times.   I seem to remember Marcos-Munky was trying to get more current out of the converter and that meant more attention to the details.

[snailspacejase]

Well thanks very much for the feedback guys. Plenty to think about and try out.
I the meantime I've ordered one of these

[Ben N]

Well thanks very much for the feedback guys. Plenty to think about and try out.
I the meantime I've ordered one of these

Let us know how that turns out.

[Rob Strand]

Well thanks very much for the feedback guys. Plenty to think about and try out.
I the meantime I've ordered one of these

It would be a good idea to keep that circuit away from the audio.   The inductor is an E-core
which aren't magnetically shielded by nature - in fact they can have quite a bit of leakage field
which can get into the audio.

The Kingsley pedal uses a shielded inductor,

[crane]

Ok here it goes - this has been tested and runs 1 .. 2 preamp tubes nicely.

U1   MC34063AD
C1   390pF 0805
C2   2,2uF
C3   450V 8 or 10mm
R1   0805 feedback resistors
R2   need to be ~2,67k
R3   680k 0805 (depends on output voltage)
R4   10k 2W
R5   330R 0805
R6   1A fuse
L2   220uH 12x12mm
IC1   6V regulator 1A TO252
Q1   IRF730A analog SMD (IRF840)
D1   ES1J DO214AC

Here are two pcbs (currently don't have a fully populated one on hand) with a pen next to them (to give the perspective of size)
PCBs have additional 6V linear regulator for filaments that schematic does not show.

[crane]

Well thanks very much for the feedback guys. Plenty to think about and try out.
I the meantime I've ordered one of these

Should work fine but man it is huuuuuuuuuuuuuuuge.
Would be ok for a standalone preamp. For a pedal - too bulky

[GibsonGM]

I find the MC34063 is also available in in 8-pin DIP from Mouser etc, so this doesn't have to be an SMD build.  I'm sure the inductor could be had from there or others, too. 

I've had good luck a few times w/the old 555 Nixie, but it IS sort of...dated....

[crane]

The first version I built was through hole. Will take a picture if I remember some time later. Went to SMT to get smaller size. SMT devices used in this "project" are huge and can be soldered easily.

[theehman]

Ok here it goes - this has been tested and runs 1 .. 2 preamp tubes nicely.

U1   MC34063AD
C1   390pF 0805
C2   2,2uF
C3   450V 8 or 10mm
R1   0805 feedback resistors
R2   need to be ~2,67k
R3   680k 0805 (depends on output voltage)
R4   10k 2W
R5   330R 0805
R6   1A fuse
L2   220uH 12x12mm
IC1   6V regulator 1A TO252
Q1   IRF730A analog SMD (IRF840)
D1   ES1J DO214AC

Here are two pcbs (currently don't have a fully populated one on hand) with a pen next to them (to give the perspective of size)
PCBs have additional 6V linear regulator for filaments that schematic does not show.

Whats c3 value?

[crane]

Whats c3 value?

C2 and C3 are the same - both 2.2uF

[printer2]

Well thanks very much for the feedback guys. Plenty to think about and try out.
I the meantime I've ordered one of these

Should work fine but man it is huuuuuuuuuuuuuuuge.
Would be ok for a standalone preamp. For a pedal - too bulky

I wanted to use it in a larger pedal with three preamp tubes, I cut the fins off the heatsink to save some room. I have yet to start it as, darn it, it still takes up a lot of room.

[GibsonGM]

LOL, I hope it didn't need those fins to dissipate heat well enough