Preamp Power SMPS

[defaced]

The SLO circuit hisses, that's the nature of how it's designed.  Bypass the anode resistors on the first two stages with 100p to 470p caps.  If you go too large these caps, you will hear the high end get rolled off.  There is a TON of info out there on the SLO.  It is the grand daddy of 4 gain stage amps.  You should be able to find more info on this and other hiss solutions out on the web. 

Moving the preamp filter caps won't effect hiss. That *could* effect ground noise (50/60 cycle hum) if it were an issue because you'd likely be forced to re-sequence the ground scheme in the circuit. 

[iccaros]

I like it..

As my esteemed colleague said, hi gain amps hiss..........

you have to kill the hiss before its amplified in the next stage..

A+ on the SMPS by the way.. 

[craigmillard]

Hi Guys,

Cheers for the thoughts!! Been bugging me for a week or so now trying bits and bobs to decrease the noise!
Do you think the noise in the clip is just standard hiss? Just seems like a helluva lot

I will give the resistor bypasses a go and see if it helps some tonight. Was thinking of trying it but seen so many people have so much less hiss than this with a standard build:(

Anyways, thanks for the A+ iccaros:) It works a treat and i now have 1 board that can rectify, smooth, elevate the heaters and supply a B+ up to 400v!:) Once i confirm this build as fully working i will knock up a build doc for the power board so everyone else can give it a go!

[defaced]

Do you think the noise in the clip is just standard hiss? Just seems like a helluva lot icon_biggrin

I can't tell what's going on in the clip with regard to the hiss.  It sounds like it, and really the whole clip, changes volume drastically.  Do you have a compressor in there, or are you throttling the knobs, or...?  There's times when the hiss to guitar sound ratio sounds right, and others where it's sounds totally fugged up.  If you can post a clip that doesn't have the drastic volume changes, I can give a better opinion on if that's like what you'd get from a SLO or not.  A clip that is standard playing volume for you that has some guitar playing and some hiss will be good.  Also, does the hiss change when you short the input to ground?

I am very much interested in a 400v SMPS.  Definitely looking forward to seeing what you've done. 

[craigmillard]

Well i'v been digging away at this for a while now and have managed to get the noise floor down a bit. I had incorrectly put a 220k resistor on the third tube plate instead of 100k and swapped out alot of carbon for metal which made a big diff.

But still have some hiss and buzz and high gain and believe it to be coming from the heaters.
Here is the board im using:


And the toner transfer if anyone is interested:
https://www.dropbox.com/s/ji3iyeozoi4k3p9/Power%20Board.pdf

I will create a PDF of this project once im happy but want to be sure the hiss/hum isn't from the heaters!!

Can i ask if you guys could check over the board again and spot any problems??

One thing i do notice is if i kill the power to the amp the hum/hiss goes while the power drains down for a few seconds and can still play. Would this indicate the heaters?
I'm considering splitting the board up to separate the AC rectifiers from DC parts and also giving some more distance between the inductor and everything else ads it does put out some interference!?

What you all think:)

Cheers
Craig

Edit: Just realized i haven't explained the board!

I have a transformer with 2*12 secondaries which connects to the transformers clamps. The transformer is due to the need for elevated heaters in my design.
Clamp HV-2 connects to J1-1 and supplies the voltage to the SMPS to be multiplied.
Clamp HV1-2 connects to J1-2 and supplies a steady 12v for the max1771 v+
J3-2 clamp connects to elevation-2 which allows for elevated heaters from the b+ line
J9 & J10 are the B+ outputs to be connected to the preamp circuit.
ToHeaters 1 & 2 connect to 2 series strings of 2 * 6n2p-ev tubes, each toheater output is around 12.6v
J3-1, J2-1 & 2, HV1 all connect to ground.

I hope that makes sense:)

[defaced]

Schematic? Staring at boards without a schematic is not fun, and not very beneficial. 

If your heaters are AC, you'd get AC line hum.  If they're DC, you could get rectifier transient noise, or double the line frequency hum.  Going by the board, it looks like you have regulated DC heaters, but can't tell for sure. 

[craigmillard]

Hi defaced,

Thanks for taking a look!!!
Here is the schematic of it:


It is regulated DC heaters on the right of the board

What causes rectifier transient noise, or double the line frequency hum then?

Cheers again

edit: here are some pics of inside the amp too, not the tidiest as i have done alot of poking around!!

[defaced]

I think your problem is how you strung the B+ nodes for the tubes; they're all in parallel.  That's great with a pedal because the power supply is relatively clean, but tube amps traditionally have them in series so that the more sensitive stages get the cleanest voltage because they've been passed through so many RC filter stages; your project is more akin to a tube amp than a pedal.  I'd dig up some high gain tube amp schematics and see how they did the power supply filtering sequencing, most of them will be in series.  Also, traditionally RC filter combos for the supply nodes are setup to be effective down below 1Hz to clean up line ripple and provide full frequency decoupling between tube stages.  10k/22u, 4.7k/47u, and ratios thereof are common.  Lastly, you're only using one 100n cap between the switcher and the supply nodes.  I'd be suspect if that's getting all the high frequency junk out, so adding a cap in parallel with the filter caps may be needed.  I'd also be suspect of it's voltage rating, your B+ nodes are 450v caps and your 100n cap is 250v per the schematic. 

Do you have a scope?  If so, I'd scope the B+ line to see what if any garbage is on it and I'd scope the heaters to see if there's ripple on it.  Also, do you have a standard HV tube transformer handy?  You can sub that in for the SMPS to see if your problem is the junk in the B+ or the heaters.  IMO, your issue isn't the heaters unless it's a grounding problem where the DC grounding points for the heater regulators are riding on garbage from the B+ ground.  Heaters get a lot of flack for causing noise when in reality, a properly built project will have nearly no heater hum issues; they're a scape goat for poor building and layout. 

What causes rectifier transient noise, or double the line frequency hum then?

Rectifier diodes when they switch on can cause voltage spikes.  Strapping small (pf value) caps across them, or using ultra fast diodes can remedy this.  Double the line frequency is residual ripple on the supply from not enough capacitance being used to filter it.  If you try to regulate a voltage that isn't smooth enough, it'll be very unhappy.  4700u might leave too much ripple for proper regulation, you'd need to a scope to be sure. 

[craigmillard]

Cheers mike!:)

I tried to keep the b+ nodes similar to the soldano x88r schematic which parallels the stages:
http://www.prowessamplifiers.com/schematics/post/soldano_x88r.pdf
But i haven't included the 40uf before the paralleled stages!! Also i didn't use 10uF cap but used 33uF 450v caps instead as what i had.
So you reckon i should try adding a 100n cap off each b+ node?
The 100n cap that is already there is a 400v cap, should i add an electro here too?
Shall add some more uf to the heater supply too maybe double the 4700uF thats already there?

Just looking for stuff to try really:) Very appreciative of your help though!!!
Unfortunately i don't have a scope!:( too expensive for little me, would be a handy addition though! So i cant view where the ripple/junk is, makes it all trial and error.

[defaced]

So you reckon i should try adding a 100n cap off each b+ node?
The 100n cap that is already there is a 400v cap, should i add an electro here too? You may need to copy the CRC filter used in the XR88 schematic, a simple cap might not be enough
Shall add some more uf to the heater supply too maybe double the 4700uF thats already there?

Excellent ideas, all of them.  Do them and report back.  The next step I'm thinking (and I don't want to go this route if we can get around it) is a choke in your B+ supply.  You're welcome for the help.  Debugging this sort of stuff is fun for me.  

[craigmillard]

Cheers mike! will give them all a go and see the outcome:)

just outta curiosity, how should b+ rc filters be calculated?

[defaced]

Using a program like Duncan's power supply simulator.  There are pencil and paper ways of doing it, but I've never bothered to learn them (yet) because I can use a program like that one and see V and I for every supply node.  There's one problem with your build, you are using a SMPS, so you're off in lala land which means you get to figure it out the hard way.  

[craigmillard]

lol just how i like it, lala land!

[craigmillard]

I have tried all the above and they all helped a little bit, i have also restrung the B+ into series string which also helpped but there was still a little buzz!

Then i realized i have hardly any filtering after the rectifier for the smps! So I put a RC filter on it and the hum is gone!:)!!!!! The preamp is almost silent at full volume!

Wicked! Thanks for the help Mike, deffo pointed me in the right directions, i was thinking it was the heaters all along!

Anyway, one last question if you could help:) The RC filter, how do i work out the wattage required for the resistor and what voltage it should be? The rectifier puts out about 18v and i think the SMPS can pull upto 300ma-1a. I tried 50ohm 1watt for the test but it started to burn out the resistor! Im useless at the theory!

[PRR]

> can pull upto 300ma-1a. I tried 50ohm 1watt

300mA is 0.3 Amperes and 0.3 Amperes times 50 Ohms is 15 Volts.

1 Ampere times 50 Ohms is 50 Volts.

You don't _have_ 50 Volts, you hardly have 15 Volts, and you sure don't want to put-in 18 Volts and get out 3 Volts.

You don't want to blow-off ALL (or most of) your Volts.

But just for interest: if most of the 18V is dropped in 50 ohms, the current is 18V/50r or 0.360 Amperes. The Power is 18V*0.360A or 6.48 Watts. Yes, a 2W part will smoke.

Pick a smaller number of ohms to drop less Volts.

1 Ohm is a smaller number (and easy to calculate).

0.3 Amperes through 1 Ohm drops 0.3 Volts.

1 Ampere through 1 Ohm drops 1 Volt.

Starting from 18 Volts, you drop to 17.7 Volts or 17.0 Volts, which is a "small" loss of your 18 Volts.

> hardly any filtering after the rectifier for the smps!

If you use a 50/60Hz power transformer and rectifier, and expect pure-clean DC, you _must_ put a big storage cap after the rectifier. Look at all power supplies you can find. Power transformer, rectifier, 100uFd to 10,000uFd capacitor, _then_ to tubes (or switcher).

As a rough-ballpark, use 1uFd for each 1mA of load. (Or 1,000uFd for each 1 Ampere of load.) Tube power amplifiers pulling 100mA at 350V use 40uFd. (So do tube preamps pulling 10mA at 300V-- preamps need cleaner power and 40uFd is sorta-standard.) A transistor power amplifier pulling 2 Amperes will use at least 2,000uFd (and for other reasons often two 4,000uFd in series).

You have a preamp and need clean power. However your SMPS appears to regulate, so the raw power does not have to be dead clean. That does not mean you can feed it UN-filtered "DC" from AC source. You want 470uFd to 1,000uFd after your rectifer.

[JimSoprano]

Hi,

I'm looking for a SMPS which can be used with three 12ax7 tubes (6 triodes), any ideas?

Thanks,

Jim