100-200 Hz noise

[chocopower]

Hi there.


I have this old tube amp with a selenium rectifier bridge.
i upgrade it with a modern rectifier bridge (is not new. I get it from some old equipment)


After the upgrade, overall voltages rise, but everything is still in safe range.


The problem now is that i have a 100-200 Hz hum noise who start after a few seconds after switch-on and after 2-5 seconds, get down to a more moderate level but still is there.


I check for cold joints, bad tubes...




Any idea?

[jjasilli]

120 Hz indicates HT B+.  Maybe bias supply if thats fixed bias tapped off the B+.  If you change 1 thong in an old PS, might as well change all of it  .

[chocopower]

The amp is cathode bias but it use a voltage divider with a cap to ground to generate a voltage to elevate the heaters center tap.


Could be the cap the culprit?

[sluckey]

If the filter caps are as old as that selenium bridge, then I would replace every one of them. Your line frequency is 50HZ, right?

[chocopower]

Yes.


220v - 50Hz  (Spain)


I´m gonna change the caps...


Thanks

[FranciscoPerez]

Great to see a spanish tube amp enthusiast around here!
+1 on changing psu caps.

[chocopower]

Hi there man!


I´m from Galicia.


Were are you?

[FranciscoPerez]

Based in Valencia, but working in Madrid several days a week.
Nice to know you chocopower 

[HotBluePlates]

... I have this old tube amp with a selenium rectifier bridge.
i upgrade it with a modern rectifier bridge (is not new. I get it from some old equipment) ...

What is the amp? Are you sure the selenium rectifier was a bridge rectifier (with 3 or 4 lugs for soldering wires)?

I only ask because many selenium rectifiers in tube amps are a half-wave rectifier, some are a full-wave rectifier, some are bridges, some are a bridge part but wired as a full-wave rectifier.

[chocopower]

yes sure. 4 wires.


I made that mistake years ago and learned the lesson... 

[chocopower]

One similar to this:

[chocopower]

And i install one similar to this:

[trobbins]

The vintage rectifier was most likely CuO, not selenium.  If you crack it open you would likely find a few cylindrical stacks of small disks. 

You should really try and locate some common diodes like UF4007, and just use them for replacements (even if you have to make up a full bridge).  Using very large current modern diodes/bridges can cause there own noise and hum issues due to capacitance and reverse recovery.

[chocopower]

ok.


I have some 4007 around here.
I just used the bridge because it seems nice and make simple the process, but i didn´t know that using a "too" big one could generate problems.


Tomorrow i´ll install the 4007.


Thanks!!

[kagliostro]

Give a try to the 1N4007 but the better sostitution are UF4007

http://datasheet.octopart.com/UF4007.-Fairchild-Semiconductor-datasheet-35217.pdf

http://www.vishay.com/docs/88503/1n4001.pdf

Franco

[chocopower]

Ok....


diodes and resistors and caps in the B+ line changed. Noise still there.


Right now, the only original cap is the bypass cap for the power tube cathodes.


Schematic attached.


Any idea¿?

[sluckey]

Try a fresh set of tubes, especially the output tubes.

[chocopower]

Yep..


is the last thing i can try. I have not a pair of el84.


Have to say that with the rectifier upgrade voltages rises from 255v to 292v in A point in the B+ rail.


This could damage the old EL84 pair....

[sluckey]

Why did you replace the original rectifier? If you put the original rectifier back in does the noise go away?

[chocopower]

The original rectifier had a huge voltage drop.
Power transformer have 115-115 secondaries and after rectifier i get 255vdc.
With new diodes i get 292vdc.

[chocopower]

more...




in the coupling caps for the power tubes i have 3mV in one and 1.5mV in the other.


Is this leaking too much?

[sluckey]

Is this leaking too much?

no

Did you have hum before you replaced the rectifier?

[chocopower]

no

[chocopower]

More information.


With the light-bulb limiter, star as normal. Light goes down, but after 4-8 second the bulb light increase at the same time the noise came. Bulb not light really high, but seems i have a short....

[shooter]

but seems i have a short

You had used a *block-molded* bridge, then was suggested 4007's, which do you have now?
You might have 1 bad diode in the bridge configuration?  Measure Volts AC from the + of the bridge to ground, how much AC is there?

[sluckey]

What you describe is exactly how a light bulb current limiter should operate when connected to a properly working amp.

[chocopower]

You had used a *block-molded* bridge, then was suggested 4007's, which do you have now?
You might have 1 bad diode in the bridge configuration?  Measure Volts AC from the + of the bridge to ground, how much AC is there?

I changed the block-molded bridge for 4 new ones 4007´s.
Noise is still there.....


Tomorrow i´ll make those measurements.


Everybody, thanks a lot for your help!

[chocopower]

...


I dont know.


I clean up the layout, move some wires and hum is gone.


The good thing, i made a lot of search, after all your recommendations, and learn a lot new stuff in the process.


Thanks a lot!!!!!

[kagliostro]

Nice, all is well that ends well

Please can you say more about the Bern Toaster ?

Ciao

Franco

[chocopower]

Yep..  Sometimes the solution is just do the things with calm and care.


The amp is an old french amp (60´s).
                 
The original design had only a 12ax7 wired in gain stage + cathodyne and a pair of EL84 in cathode bias and U.L.


I added a ef86 gain stage and reorganize the B+ rail. Original amp have the b+ supply for the O.T. after the choke.
I rise the filter cap values from 8uf to more modern values.


Right now im working in tune the high/bass response.


I´m looking for the cleanest amp possible and use pedals for get "more".


I have no way to do pictures right now, but the look is a LOT like a and Orange Tinny Terror, but in grey.


My main concern is the O.T.
It have 4-8-16 outputs, but is not easy to change from one to other. The secondaries are those with a lot of taps and you need to combine them in different ways to get the needed option.

I attach the transformer specs.                             


               

[kagliostro]

I've the same situation on some Geloso and RCF OT (old PS amps)

---

An ef86 in V1 is a strange choice for an amp you want to get the max clear

Which is the reason for this choice, filament current ?

(I've seen the 56k plate resistor)

Franco

[chocopower]

I just love EF86.
I explain me wrong. i love clean, but no so clean.



I already try the same configuration but with EL34 and is my main amp.

[kagliostro]

The 56k plate resistor tame at the right level the esuberance of the EF86, a more quiet tube will be 5879, but isn't so easy to have under hand

(I mean by second hand or at Hamfest here in Italy)

I've a bounch of EF86 (some NOS some used) and only a pair of 5879 that I bought new

Franco

[PRR]

> It have 4-8-16 outputs, but is not easy to change

Bah. They were obsessed with the last dB of fidelity. Wire it like this and switch freely.

There will be some difference of response above 10KHz. And 4-Ohm will measure slightly lower output than 8 and 16. But since we never have speakers *identical* except for impedance (and speaker impedance is always a wide target), who cares?

[chocopower]

PRR, thanks a LOT!