100Hz hum in preamp

[untune]

Hi all,

I recently completed building a preamp that I put together with help from the lovely people here. It has been in the works for a couple of years, on and off, and I've built an enclosure and boxed it all up. It performs much better than I had hoped for a first project (bar a couple of very small issues) but the most obvious problem affecting it is a low level 100hz hum.

I can't post any pics until later, but I'm wondering what the obvious diagnosis for this might be.  The last uploaded schematic on here is https://groupdiy.com/index.php?action=dlattach;topic=64477.0;attach=47843;image but it has been revised a bit since then. The HT/ filtering is still unchanged I believe, although the voltages there are wrong and I think it's about 303v at the EL84, 270v ish at the second 6AU6 then 250v ish at the first 6AU6.

Power is rectified via a 6X4, and fed from it's own 6.3v heater tap on the power transformer.

There are two channels - each one is roughly similar, and goes into the EL84 via a switch.  Changing gain does not change the hum level, it is constant.  There is a 1k trim pot between the output of the OT which naturally does drop the signal, hum included, to silence.

I have a roughly 250 ohm pot to balance the heaters on the main 6.3v tap, with an artificial center tap to ground.  This made little to no audible difference to the hum level.  I also tried referencing this to the cathode of the EL84 with no real difference.  I don't know if putting another humdinger on the 6X4 heater and creating an artificial center tap for that too would be beneficial?

Other first guesses are that the HT needs more filtering (cant really do much with the 22uF on the rectifier though) or that the output transformer may be a little too close to the power transformer, although it's closer to the choke.  Or perhaps there's vibration fron the power tranny through the chassis, albeit very low.

It's not a massive level of hum, the volume has to be cranked to hear it, but I can detect it just behind mic signals and it's really annoying - if I can get it down another 6 or 10dB (or more!) I reckon it would be fine.

Any ideas or things to check?

Cheers!

 

[ruffrecords]

I am never 100% sure what to do with heaters for tube rectifiers but it is important to manage the heater cathode voltage. So if these heaters are left floating they are definitely not managed. I recommend connecting on side of the 6X4 heaters direct to its own cathode. I don't think this is the source of your hum problem but it is probably worth doing anyway.

If the hum is definitely 100Hz then it implies it is an HT rail problem. There seems to be adequate decoupling for the first two stages but the output stage V3 sees only roughly smoothed HT. As the output is single ended with no NFB the HT ripple will appear directly in the output attenuated by the pot divider consisting of the 5K1 anode resistor and the ra of the triode wired output tube (which in this case probably around 2K). Not much you can do about this but you can verify if this is the source by shorting the grid of the output tube to ground so you hear only the hum induced in that stage. If the hum stays much the same you have identified the source.

Cheers

Ian

 

[untune]

Ian, thanks for the suggestions - that gives me a good starting point to work from.  I have a shot of the spectrum showing that while there is some 50Hz, the 100Hz is an order of magnitude above it in level - I'm working in Birmingham today and it's on my home computer so I'll upload it this evening so you can take a look.

 

[abbey road d enfer]

Most of the times, noise/hum problems cannot be seen in the schemo.
Number one cause is improper ground circulation.
Your published schemo is a perfect example of hierachical ground, where ground follows signal. Have you respected that in your build? In particular the node that connects the secondary's center tap and the reservoir caps should not be allowed to connect to the ground west of R14-C7.
Ther is also the possibility that the choke radiates into the audio xfmrs; you need to check its orientation and distance.

 

[untune]

Abbey Road - great point, the grounding scheme might have gotten confused during breaks on working on it. From memory... The input transformer ground, mic input ground and line input connect to the base of the V1 cathode cap and resistor. The instrument input (schemo is out of date, there are two now) and local grounds are connected at the V2 cap/resistor base. I believe these grounds do run back to the base of the 47uF cap after the choke.  Transformer secondary CT etc connect to the base of the reservoir cap and that all runs to a point on the chassis near the IEC mains earth connection. So should I be grounding the input stages in another way?

The chassis layout is quite tight so positioning could be having an influence... I may see if the OT can be rotated out of the way. The last coupling cap between EL84, a rather large polypropylene IIRC, is quite close to the choke.  Could that be contributing?

 

[abbey road d enfer]

Abbey Road - great point, the grounding scheme might have gotten confused during breaks on working on it. From memory... The input transformer ground, mic input ground and line input connect to the base of the V1 cathode cap and resistor. The instrument input (schemo is out of date, there are two now) and local grounds are connected at the V2 cap/resistor base. I believe these grounds do run back to the base of the 47uF cap after the choke.  Transformer secondary CT etc connect to the base of the reservoir cap and that all runs to a point on the chassis near the IEC mains earth connection. So should I be grounding the input stages in another way?

That seems correct.

The last coupling cap between EL84, a rather large polypropylene IIRC, is quite close to the choke.  Could that be contributing?

Not likely.

 

[untune]

Got a bit waylaid with other chores tonight, but I did manage to update the schematic to what I think is accurate to the current circuit.

I listened earlier and it almost seemed like noise had dropped a couple of dB.  I'm wondering if I'm just being too picky now, as the volume does need a boost to hear the hum... but then it's still above the noise of the valve.

I've not tried any of the suggestions yet, but will tweak it tomorrow and report back

 

[untune]

Noise when tested tonight

 

[PRR]

That B+ filter is just inadequate for that amplifier and use.

With some assumptions for data you have not provided, I get about 43mA load and PSUD figures 68mV p-p ripple at the second cap. This is 24mV RMS. The resistor-loaded triode has PSRR about 3:1, so 8mV at plate coupling cap. OT is 4:1, so 2mV at secondary. This figures as not even 60dB below nominal Line Level. That's not good.

The 6H choke is 3.7k at 100Hz. We would *like* its reactance to be greater than the load. Load is about 300V/0.043A or around 6.5k. This really should be a 12H to 50H choke. However you are at 2/3rd choke current rating, so a higher inductance probably means a bigger choke, and you are already out of space.

Changing 2nd cap to 470uFd gets you to 3mV of 100Hz, which is better than 24mV, but not complete happiness.

 

[PRR]

Large-uFd small-case electrolytics are sometimes an answer.

A 3 stage filter gets you to 1.2mVp-p, 0.5mV RMS, ripple, which is a lot better than 24mV. Output S/H improves from 60dB to 90dB.

 

[abbey road d enfer]

I listened earlier and it almost seemed like noise had dropped a couple of dB.  I'm wondering if I'm just being too picky now, as the volume does need a boost to hear the hum... but then it's still above the noise of the valve.

Troubleshooting 101 says you have to determine where noise builds up. Is there still hum when the volume pots are turned down? Is there a sweet spot where hum disappears? Is hum similar in DI, mic or Line?
Indeed, PRR's analysis indicates that hum build-up is in the output stage. Answering these questions should corroborate it.

 

[Tubetec]

Ive done several pre amps based on a single  EL84 output stage , they always required hundreds of uF(or thousands)  ,Rc as well as Lc smoothing and very carefull placement of iron components to get hum down to an acceptable level . I tend to put Rc sections after the rectifier then follow with LC closer to the load , that way the amount of ac ripple seen by the choke is reduced ,which in turn should means it should  radiate less  . Looks like you have a fair amount of 50hz induced hum and then the residual 100hz in the supply itself . The nature of a balanced output stage means much noiseier supplies work fine , single ended though at that kind of current draw , almost impossible to do quietly without seperating Psu and signal chassis in my experience . If you can spare some ht volts try an Rc network between rectifier and choke , 100 or 200  ohms wont loose you too much ,then up the cap post choke to 470 680 or even 1200 uf . I used three small chokes and about 1500uf in one of my SE EL84 pre's psu's ,its very quiet ,but the psu weighs several kg's  :-[ 

Tube recifiers live longer and happier with cathode tied to heater also , I also found dc heating of the EL84 to be beneficial in my builds .
Its tough to back engineer when you already have your layout etc , I guess just do the best you can ,learn from it , and capitalise on the knowledge for your next build .

Might be worth making the vu meter switchable also ,as across the output transformer they do tend to add some distortion .

 

[Tubetec]

Opps missed the updated schem

 

[untune]

PRR: Many thanks for the input and the calculations, I appreciate that Finding caps that size is proving to be difficult, most I can get of the ones I'm already using is 220uF @ 500v.  Will have to do some hunting round.

Abbey Road: Hum remains constant with everything minimised, and everything maxed out.  The first stage valve is a little noisier than the second and may contribute a tiny amount, but nothing major.  It seems to be pretty much conistent in all stages, and the final trim pot between the OT secondary drops everything to silence including any signal/noise from the preamps valves and the hum (as expected.)  I will try a couple of the things Ian suggested tonight, such as tying rectifier heater to cathode and shorting the EL84 grid.

Tubetec: That's very useful info, thanks! PRR's calculations point in the same direction... I may actually have some leeway to move the OT further from the choke and PSU.  I used pre-drilled chasis holes that an old OT was mounted to without giving much thought to position, so I may be able to shift it further away, if not by much.  I actually originally designed this for slightly lower voltages so I don't think dropping the HT a bit could affect things too much - I jut settled on what it measured at since it seemed ok. 

Can I put an extra RC network before the choke?  My understanding was that I can't play with that 22uF or increase it without harming the 6X4.  I did leave myself a bit of room when I laid the board out to separate the power section from the rest, so I think I can fit some bigger caps in without much issue.  Finding axial caps the right size and rating might be a challenge though...

Quick clarification on the rectifier heater/cathode - am I just tieing either pin 3 or 4 to pin 7 at the rectifier valve base, or should it be done elsewhere?

Cheers all

 

[abbey road d enfer]

Hum remains constant with everything minimised, and everything maxed out. 

OK, so that confirms PRR's diagnosis. I believe you could choose to operate the EL84 at a lower current, by increasing the cathode resistor. I would think you don't really need the massive headroom. I would think increasing the cathode resistor to 470r, combined with an additional RC in the B+ line would bring serious improvement.

 

[untune]

OK, so that confirms PRR's diagnosis. I believe you could choose to operate the EL84 at a lower current, by increasing the cathode resistor. I would think you don't really need the massive headroom. I would think increasing the cathode resistor to 470r, combined with an additional RC in the B+ line would bring serious improvement.

Great suggestion, I'll try that too.  I'll need to go through the resistors from varous iterations, I may have tried a 470R at one point.

I think I do also have 1x 47uF and 2x 22uF caps left; I doubt that they're big enough for another stage based on previous replies, but they might be useful in testing it

Cheers!

 

[scott2000]

Any pics of this build??....

 

[untune]

Any pics of this build??....

I will get a few up tonight - is it to get a view of layout and proximity etc?

 

[scott2000]

Just Curious .... ;D

Also,  would elevating the heaters referenced to a voltage off the B+ be an option?? 

I have also had good luck with a couple of .47uf caps virtual ct to a reference voltage ...... more so than a hum pot...


I don't think treating the 6x4 heater is beneficial in any way in regards to your one question about that....... ???

this is after all the suggestions obviously....

I did have a question about the pots and if it would be beneficial to have a  coupling cap after the wiper????

 

[untune]

Just Curious .... ;D

Hi Scott, I've just taken a few pics, you can view the album here http://imgur.com/a/ZWlv2gV

Cheers