Please help me de-hum my new pre

[basstard]

Hi there everybody. I'm asking you to be patient with me cos I a total noob. I copied once a schematic and everything went ok, then I made a decision to assemble some circuits together to get the preamp I wanted.
I'm talking guitar here btw.

I use a bogner ecstasy signal preamp [here: http://www.mif.pg.gda.pl/homepages/tom/files/xtcod.gif ] and an AX-84 plate psu with some mods.

For the filaments I use a switching mode psu. I get 9v outta it and then pass it thru 3 x 7805's grounded with 2 diodes each. This gets me 6.29v on the filaments with load connected.
I use a 1000uF cap to ground before the 7805's and a 100uF cap for each 7805 output.

Now, problem is I get a really bad hum, a sort of hi-pitched whistle [I can record it if you want] and I know it comes from the filaments psu cos I put a switch on it and when I turned it off the hum disappeared very soon,,,

If I ground the 7805 casing [which I thought to be the same as the ground pin] to the negative of the first cap of the plate psu I still get 6.29v on the filaments and no hum at all but this is only with the guitar plugged in,,, as soon as I unplug the jack from the female on the amp it hums back again,,, if I just unplug the jack from the guitar, it doesn't hum,,,

I also tried a different psu for the filaments and it did just the same,,,

any suggestions on what to check? this amp sounds very good, just it has this prob it would be a shame and a personal failure if I had to give up,,,

 

[Miniwatt]

I know it comes from the filaments psu cos I put a switch on it and when I turned it off the hum disappeared very soon,,,

your whole amp is pretty much turned off when you turn off the filament..

 

[basstard]

there's a small time in which the tubes are still pretty hot and you can have the signal being still amplified,,, not?

 

[gootee]

Hi basstard,

I am not an expert. But here are a few thoughts:

It would probably really help if you posted a schematic or diagram also showing the power supply and interconnects, even if the SMPS is just shown as a single block. It would also be good if the diagram corresponded to the actual (physical) layout, showing the true topology of your power, ground, and signal conductors, including any cables, chassis connections, etc.

A photo of your unit's internals might also be *extremely* helpful. Contruction techniques and physical layout could be the entire cause of the problem, OR, could already be minimizing the problem, or could be anywhere in between. For now, without more details, we will probably just have to assume that the layout and construction, etc, is 'neutral', which is not helpful.

If you don't have the use of an oscilloscope (Do you?), it might be more-difficult to know what is causing the whistling. SMPS switching noise is typically way above the audio band. So an audible whistling sound is either something else, OR, possibly (but probably not likely), an audible 'artifact' caused by the higher-frequency SMPS spikes. (What is your SMPS switching frequency?)

Without better information about your circuit and its implementation, we can still provide a few suggestions. (But, as mentioned, you will probably get MUCH better help if you provide photos and a wiring diagram.)

So, unless we just get lucky, the suggestions below probably will not solve your problem, since it sounds like it could be either ground-loop-related or impedance/stability-related, or possibly even EMI/loop area-related. But they should at least improve your circuit's performance. And, since they should be relatively easy and cheap to implement, they are probably worth trying, at this point, at least until someone can provide some more-relevant suggestions.

1.) You should probably add 0.1 uF X7R ceramic capacitors between the 7805 input and ground pins, making sure that the caps' bodies are as close as possible to the regulators' input pins and the connections to both pins are as short as possible.

2.) You can also try the same thing between the regulators' output and ground pins.

For both of the above: If you have some other type of 0.1 uF capacitors, you can try using them, as long as you remain aware of the possibility of exciting unwanted high-frequency resonances when paralleling things like film capacitors and electrolytics, and possible instability problems when having too-little ESR across a regulator's output. If necessary, a very small resistance could be added in series with any very-low-impedance capacitor.

(To dramatically-attenuate any SMPS switching-spikes, maybe even down to around the 100 uV p-p range, or less: )

3.) Between the SMPS and each 7805, you should probably have an LC lowpass filter, using a series high-current inductor of, say, 10 uH, with a capacitor to ground, after the inductor (i.e. from between inductor and 7805 input to power ground), of about 1000 to 2200 uF. You could use a larger inductor and a smaller capacitor. But usually the lower inductor values are available with higher max currents.

For the LC filters' inductors, something like a Bourns/J.W. Miller toroidal 10 uH might work well, e.g. 2100HT-100H-RC (10 uH, 9.9 Amps, .012 Ohms, 0.86-inch diameter), available as mouser.com's 542-2100HT-100H-RC for $2.44 ea qty 1 or $1.69 ea qty 10 (For the vertical-mount version, replace ...H-RC with ...V-RC, keeping in mind that EMI radiation is mostly along toroid's axis.) If you need a higher maximum current, their 2300 series 10uH goes to 20 A. For lower max currents, there are probably many other suitable inductor parts.

For the LC filters' capacitors, something like the Nichicon 'UHE' 105 degC low-impedance high-reliability series would probably be very good, e.g. mouser.com's 647-UHE1C222MHD, which is 2200uF, 16V, .027 Ohms @ 100 kHz, 2.23A rms max ripple current (at 100 kHz while at 105 degC), radial, 12.5mm D x 25mm L with 5mm lead-spacing, for $0.78 ea qty 1 or $0.71 ea qty 10.

4.) You should probably also experiment with the routing of the 7805s' grounds. For example, rather than running the SMPS's output ground to a 7805 ground and then to a heater circuit ground near the tube, you should try running the SMPS ground directly to a heater ground and then run a separate conductor from a 7805 ground to the heater ground.

 

[basstard]

gootee, thx a lot for your reply,,, I can figure this thread will be a long one

I need a program to design the schematics of the whole thing to let you know what we're talking about. Do you have any suggestions on what program to use?

I'd really need an oscilloscope and I don't know the switching freq of the smps,,,

Maybe Miniwatt is right tho [and I was talkin' BS then], since it doesn't squeal with the guitar plugged in, I can't test if when I turn off the filaments the squeal goes away or not, cos it ain't there in the first place,,,
with the guitar unplugged instead, it squeals and I can hear the squeal go away when I switch off the filaments but then that could well be that the amp is no longer putting out ANY signal,,,

I must add one thing tho: as you can see in the schematic I posted, there's a switch named 'structure' [puts C4 and R10 to ground],,, when the guitar is plugged and there's no squeal, if I close the switch and put R10 and C4 to ground, the squeal returns back,,, so maybe in my infinite ignorance I was thinking that it was the filaments supply but it's a ground loop,,,

before I post the complete schematics, what are the first things I should check to see if it's a ground loop and what part of the circuit could be triggering it?

many, many thx guys!

 

[gootee]

Hi basstard,

There are probbaly lots of programs that would work well. I happen to use LTspice for almost everything (schematics and simulation). It's extremely easy to use for drawing schematics. It even has pre-made symbols for triodes, tetrodes, and pentodes (in 'misc' folder), and all other more-common components. I also use it for simulation, of course, since that's what its main purpose is. But you wouldn't have to actually simulate, if you just wanted to draw schematics.

After a schematic is drawn, you can just use Tools-->Copy Bitmap to Clipboard, and then use your favorite image-manipulation software (maybe irfanview) to convert it to a suitable GIF file, or jpg, or whatever works.

To download the LTspice installer (.exe) file, right-click the link and select Save Target As:

http://ltspice.linear.com/software/swcadiii.exe

(Don't worry. It doesn't mess with anything else in your system.)

Sorry. No time to comment on anything else, right now. Just started a new job, today. Bedtime.

 

[basstard]

thx a lot mate,,, now I just need time to draw the schematics,,, I got some pics but I'm a bit ashamed to publish them here since I see a lot of incredible works and I'd get banned instantly

 

[basstard]

While I get enough time to put the circuit into its layout scheme, I recorded the noise it makes.

first part it's noise with guitar unplugged and turning gain up and down increases and decreases the frequency of the noise. then middle control raises volume of the noise [both gain and middle have a grounded pin if that means anything]

second part it's how it sounds with the guitar plugged in [no noise] and third part is guitar sound with structure switch closed - noise comes back in,,,

here it is [mind you the noise its not nice to listen to so keep the volume low - oh, and excuse my terrible guitar playin, I'm a bass player ] :
http://www.esnips.com/doc/16bf2079-ec89-4dfd-b212-acd04f9d516f/rog-01

 

[basstard]

Almost killed all hum [touching wood]. Just needed a bit more shielding,,, now, where's that redfaced emoticon?

 

[basstard]

Here's a clip of how it sounds:

http://www.esnips.com/doc/0ba653cf-733d-4d41-9650-8fe489e71cd0/rogmod-02

recorded with my ultracheap soundcard and the stock pickups on my ibanez. oh, and my cr*p playin,,,