Reducing hum in SE amp by phase cancellation, with a differential op amp?

[cboysen]

Hi Guys!

I recently built a 5f2A replica, and it sounds fabulous. But it also hums quite a bit. This is due to lack of supply filtering at the adequitely sized OT, and no phase canceling as per SE design. I did what PPR suggested, added an extra RC stage. It works, but it also takes away some of the swirl and bouncy feel of low filtering. Ultimately, it's a compromise between great vintage tone or great noise reduction.

So, I'm sorry for trying to have my cake and eat it too, but is there any way to reduce the hum further, without stiffening the amp?

I have had this rather odd idea for a solution for some time. Would it be possible to make a small board with a differential op amp -  That is supplying the Vin- of the differential op amp with a low-voltage signal, divided from first HT supply node. Then for the Vin+, take the same low voltage HT supply signal and filter it with a big capacitor. The Opamp would then differentiate between a smooth DC of say 5V at Vin+, and rippled, poorly filtered 5V at Vin-. The output would then, only be the ripple voltage of a few milivolt.

Now from here, could you send this signal into the cathode of V1B, where the phase is flipped and ultimately end up as a mirror image of the ripple voltage at the output, in which case they cancel each other out, kinda like in a push/pull amp?

is this doable, or waay off?

Kind regards
Boysen

[trobbins]

Worth a try Boysen, if you are that keen.

I presume the main objection for you is the hum during idle or very low signal levels?  And any hum that exists during louder volumes is not really an objection, or not overtly noticeable?

You may want to try a simpler 'error signal' for starters, which is just AC coupling the B+ ripple voltage down to your error correction opamp, and seeing if you can open-loop null or adequately suppress the ripple voltage in the speaker output signal.  You may want to bandpass, or at least low-pass the error signal so that it tries to correct mainly the fundamental frequency of the ripple.  Given class A operation, that ripple error signal shouldn't change dramatically between idle and full volume playing (as compared to a PP stage).

If you get that far, then you could do a frequency sweep at different volume settings, to see how much interaction there was with the ripple reduction circuitry in, versus out.

[PRR]

> supplying the Vin- of the differential op amp with ....voltage HT supply signal ...only be the ripple voltage....into the cathode of V1B, ...flipped .... ....  cancel each other out

Whew!! The ripple is BIG. Almost any two points can be a "differential". Leak some ripple to the driver cathode. I do not know what "???" is. It probably wants a DC-blocking cap, and a large resistor? Maybe 0.05u and 1Meg? Much experimentation is needed because if you come down from infinity a little too far, it will still buzz, just anti-phase, which sounds the same.

I suspect the idea is doomed by the radical kinks in impedance in speaker and OT in the 100-600Hz range. You can null one harmonic or another, but not all of them, or not so it works day to day.

[jjasilli]

I have a hard time believing the ripple is so high.

My hotrodded VibroChamp is dead quiet, no hum.  Schematic attached.


It should be possible to ID the source of hum & reduce it in a conventional manner.

[Latole]

I have a hard time believing the ripple is so high.

My hotrodded VibroChamp is dead quiet, no hum.  Schematic attached.


It should be possible to ID the source of hum & reduce it in a conventional manner.

I can dowload you schematic but I can't read it.

Any single end amp hum more than a push pull amp.

[Williamblake]

What about a second Champ with reversed input und reversed speaker?

[cboysen]

Thanks for the response!

Worth a try Boysen, if you are that keen.

I presume the main objection for you is the hum during idle or very low signal levels?  And any hum that exists during louder volumes is not really an objection, or not overtly noticeable?

Thanks for the tips, I'll look into that! And yes, with baby on the way, I'm kept to fairly quiet levels at the moment, and one of the reasons for building a 5f2a. The hum is about quiet speaking volume, so nothing too bad. But this is at the cost of some 'vintage dynamics'.

> supplying the Vin- of the differential op amp with ....voltage HT supply signal ...only be the ripple voltage....into the cathode of V1B, ...flipped .... ....  cancel each other out

The reason for choosing a differential Opamp, was to avoid any added harmonics from the non-linear loading of a capacitor in conjuction with the rectifier, causing the voltage divided signal to be distorted. But I suspect I might have been overthinking this.

> It probably wants a DC-blocking cap, and a large resistor? Maybe 0.05u and 1Meg? Much experimentation is needed because if you come down from infinity a little too far, it will still buzz, just anti-phase, which sounds the same.

I suspect the idea is doomed by the radical kinks in impedance in speaker and OT in the 100-600Hz range. You can null one harmonic or another, but not all of them, or not so it works day to day.

I have thought about the Capacitor decoupling too, but quickly dismissed the idea, for the reasons above. Anyway, I ran your idea through LTspice, and it might have some promise. (see attached picture)

You might be right though, this is pretty much a long shot, and probably pretty doomed.

What about a second Champ with reversed input und reversed speaker?

Haha! I see your point, adding to the probability that this might be quite a failed attempt.  :)

Kind regards
Boysen

EDIT: The odd ripple, was fixed by increasing the LTspice resolution. I've attached a second photo to this post, for comparison.

[Latole]

Look filter caps circuit do not work.

[cboysen]

I said the capacitor distorted the voltage - This was an LTspice artifact. I've deleted previous post

Here's the correct picture.

kind regards
Boysen

[Latole]

I said the capacitor distorted the voltage - This was an LTspice artifact. I've deleted previous post

Here's the correct picture.

kind regards
Boysen

I had same issue, it was the filter cap

[cboysen]

I had same issue, it was the filter cap

The build is completely new, and I've changed the filter caps several times to adjust the hum vs tone. While a bad filter cap will cause hum and other problems, what I'm trying to describe here ( and sorry if it's a litte vague), is a natural problem with SE amps. Not a problem with any bad wiring, component etc. Just an inherent "flaw" if you will.

The capacitor I mention, is C5 in the picture. With LTspice not a high enough sample rate, it made it look like the capacitor was doing funny business to the circuit, and since it completely correlated with some overthought electrical hypothesis, I didn't think more of it.

Anyway, I will try to make the circuit later today, and report back :)

kind regards
Boysen

[Latole]

I never built a amp with C5 m R6 and R7 and all of them are quiet.

Your saw teeth waveform look a bad filter cap,
10 mfd caps look very small to me

[cboysen]

I never built a amp with C5 m R6 and R7 and all of them are quiet.

Your saw teeth waveform look a bad filter cap,
10 mfd caps look very small to me

Latole - This isn't a scope of my amp, it's a scope of a computer simulation in LTspice.
Notice the ripple is from 385.2 to 387V, thats a ripple of 2 volt, at the first node. Thats a ripple percent around 0.5%.
Also 0.1uF is more than enough to capture below 50hz of AC output.


Anyway.. I just tried to make it work, and it doesn't. Of course nothing is like on paper, and I should have guessed you can't just cancel one signal with another that easily. Oh well - We learn something everyday.

Kind regards
Boysen

[Latole]

2mv ripple !! My bad ; everything look good, Sorry

[jjasilli]

I have a hard time believing the ripple is so high.

My hotrodded VibroChamp is dead quiet, no hum.  Schematic attached.

It should be possible to ID the source of hum & reduce it in a conventional manner.

I can dowload you schematic but I can't read it.

Any single end amp hum more than a push pull amp.

1.   Google Express Schematic.  Free download.  Lot's of people on this Forum use it.

2.  Any single end amp hum more than a push pull amp.  This is plainly false.  It is indisputable that some PP amps hum more than some SE amps.  Also, if an SE amp hums at a volume below the threshold of human hearing, then it doesn't matter if it has more hum than a PP amp.  Again, my SE amp is dead quiet at full volume, with no signal input.  To paraphrase Marx (Groucho), who should I believe, you or my own ears? 

3.  I agree that there may be a fault in cboysen's amp.  Or, the problem may arise from outside the amp: guitar; cable; ambient noise such as RF; fluorescent lighting; dimmer switches; etc. 

At this point I would identify the frequency of the hum.  Presumably either 60Hz or 120Hz.  This would narrow down the search. 

[brewdude]

In the first post you mentioned that you added an extra RC filter stage.  Have you considered instead using an CLC filter pi filter?

[cboysen]

3.  I agree that there may be a fault in cboysen's amp.  Or, the problem may arise from outside the amp: guitar; cable; ambient noise such as RF; fluorescent lighting; dimmer switches; etc. 

Okay, this is awkward..

So, yeah, after replacing caps about 3-4 times a few weeks ago, I actually did make a mistake. My arrogance has gotten the better of me, 'cause after a few cups of coffee, it was staring me right in the face.. I had forgotten to put a capacitor on the OT node after swapping caps. I put a 10uF in that spot, and now the hum is next to nothing. The amp is a little stiff, but I suspect the new speaker still needs to be broken in..

So..  I guess that concludes this thread..   Sorry!!

Kind regards
Boysen

[Latole]

 
[/quote]

   To paraphrase Marx (Groucho), who should I believe, you or my own ears? 

 
[/quote]

That is what I tough !

[jjasilli]

3.  I agree that there may be a fault in cboysen's amp.  Or, the problem may arise from outside the amp: guitar; cable; ambient noise such as RF; fluorescent lighting; dimmer switches; etc. 

Okay, this is awkward..

So, yeah, after replacing caps about 3-4 times a few weeks ago, I actually did make a mistake. My arrogance has gotten the better of me, 'cause after a few cups of coffee, it was staring me right in the face.. I had forgotten to put a capacitor on the OT node after swapping caps. I put a 10uF in that spot, and now the hum is next to nothing. The amp is a little stiff, but I suspect the new speaker still needs to be broken in..

So..  I guess that concludes this thread..   Sorry!!

Kind regards
Boysen

Happens to everybody.  Helps to have a sounding board!