6.3VAC heaters to filtered/regulated 6.3VDC. best solution?

Thank you both of you. That helps me immensely.

OK, I've analysed the unit further.

The unit was being tested using an RME sound card balanced IO. The whole system is connected to the same mains output, so I'm guessing the ground loops aren't coming from there. If I measure the RME IO self noise, there's no hum, just flat low noise floor, about 10-20dB below the preamp.

I also tested whether the noise was coming from the input side, using a resistor configuration suggested by NYD earlier (I think it was using 200ohm resistor between pin 2 and 3). Hum still there. Both channels have exactly an equal amount of hum by the way, ever since moving the PSU out of the rack.

The funny thing I noticed is when I screw the lid on the rack, the hum actually increases about 5-10dB! Almost as if some field was being reflected inside the unit.

All XLR Pin1's are connected to chassis as close to the XLR connectors as I could manage. This is also the only chassis point of the whole unit.

The layout is unorthodox due to the fact I wanted to test if moving the capacitors around would have any effect on the noise, same with the current connection of all audio transformers. Maybe I should post some pictures, although right now it looks a bit of an abomination due to the troubleshoot configuration changes.

[quote author="Kingston"]...
The funny thing I noticed is when I screw the lid on the rack, the hum actually increases about 5-10dB! Almost as if some field was being reflected inside the unit. ...
[/quote]

Sounds as if the lid attachment is enhancing magnetic field pickup. What does the device do out of the rack?

I've only used the unit on several tables around the house, not racked up with any other equipment. It was meant to be a kind of mobile solution anyway, with the DI inputs etc.

Moving it around doesn't seem to have any effect on the hum.

I've used 6.3v Vregulators in parallel with low value resistors in series with their outputs to drive heaters. Works very well with no noise if properly decoupled. They MUST be heatsunk very well or you will get strange output noise once the device starts to reach it's overheat temp. this goes for most regulators.

EDIT: i have also had the same problem with the lid being ON the box. I found that radiated noise from the system coupled onto the wiring. Moving the wiring around the box and finding a position near the tube itself fixed this and dropped my noise floor 15db. To this day I still don't know what the problem is but as long as the wiring stays(zipties..) i have no trouble.

check your wiring at the output. maybe the primaries of the output transformer are connected wrong.

I also tested whether the noise was coming from the input side, using a resistor configuration suggested by NYD earlier (I think it was using 200ohm resistor between pin 2 and 3).

Click to expand...

To do the short circuit input test, you need to connect pins 1, 2 and 3 of the XLR together. Just a resistor between 2 and 3 still leaves the inputs free to float up and down together (common-mode).

The funny thing I noticed is when I screw the lid on the rack, the hum actually increases about 5-10dB! Almost as if some field was being reflected inside the unit.

Click to expand...

Is the lid made of sheet steel or of aluminium alloy? What happens if you hold the lid nearly in place without it making contact with the rest of the chassis - is the hum increase still there or does it only cut in when the lid and the chassis make contact?

[quote author="Boswell"]To do the short circuit input test, you need to connect pins 1, 2 and 3 of the XLR together. Just a resistor between 2 and 3 still leaves the inputs free to float up and down together (common-mode).[/quote]

I've been busy with other things so I haven't been able to trouble shoot for a while.

Just shorted the input. hum goes down 10 dB or so, but turning on the 20dB input attenuator *increases* it. weirdness.

Curiously though switching on the 20dB output attenuator sets down the general input noise (if there's any), but *also* the hum down 20dB, when there's connection made from, say from a soundcard.

This was with full gain.

if at minimum gain, nothing happens to the hum with shorted inputs.

Conclusion is that there's some input hum, but that the biggest problem still lies elsewhere.

I also tested wether the power switch/wiring caused the noise, but there was no difference in shutdown behaviour with either yanking off the cable, or using the switch.

Curiously the lid nowadays makes no difference to the hum, wether on or off the unit. The case is made of sheet steel, but front panel is 3mm aluminum.

Try using a different PSU or try using this PSU on something else that normally works OK. This is a ground loop, make sure if you use a CF that the heater voltage is properly floated.

analag

When I was having hum trouble with my 1176, I ended up lifting the ground at the XLR input, and my hum was gone. That could be a source of a ground loop in your circuit.

I redid the grounding throughout the preamp. It's now in proper order, like in the schematics. I tried all kinds of chassis grounding schemes and the one that seemed to work best was where PSU filter cap serves as commons reference, the whole routing goes thru both of the channels in correct order. The other end of all ground (both channels and I/O) meet at I/O chassis point.

Other schemes were less successfull, including chassis point at PSU. Star grounding seems like an awful waste of wire and didn't bother with it yet.

I managed to get the zero gain RMS level to -103dB, but there's still hum present (at -90dB) and at maximum gain RMS is a around -55dB (the hum spike is at -33dB and quite audible.) Am I asking too much from the preamp? I think not.

I'm at my wits end now officially. After I try DC heaters heaters the only thing left to do is redoing the whole bloody layout. Damn! Too bad I don't really know where to start with that, since I already did the best I can.

Both channels still have a perfectly equal hum problem, and nothing seems to help. Quite difficult being one of the only DIYers in town as I don't have the PSUs and equipment to try all the tricks suggested in this thread.

The behaviour of the output attenuator is what you would expect, but the input one is not.

Can you post a schematic of the pre-amp anywhere?

miniscule info? yeah.
hard to read?
yes.
but I want it off my HD and onto the website, so...

http://vacuumbrain.com/The_Lab/TA/Grounding/gndloop_a.jpg
http://vacuumbrain.com/The_Lab/TA/Grounding/gndloop_b.jpg
http://vacuumbrain.com/The_Lab/TA/Grounding/gndloop_c.jpg
http://vacuumbrain.com/The_Lab/TA/Grounding/gndloop_d.jpg


BTW, I read that there is a ton of good grounding info etc in the app notes for 78 series regs.
National Semi I believe.

humm is noise, i guess...

http://vacuumbrain.com/The_Lab/TA/Grounding/noise_a.jpg
http://vacuumbrain.com/The_Lab/TA/Grounding/noise_b.jpg
http://vacuumbrain.com/The_Lab/TA/Grounding/noise_c.jpg
http://vacuumbrain.com/The_Lab/TA/Grounding/noise_d.jpg
http://vacuumbrain.com/The_Lab/TA/Grounding/noise_e.jpg
http://vacuumbrain.com/The_Lab/TA/Grounding/noise_f.jpg

Ahh, interesting trick in there!
Sync your scope off the line, and look at the hum.
If it's drifting, then it is coming from another source.
Be careful on the sync hookup!

I build my own regulatars witha TO-220 transistor, three 1000Mfd caps and one 2200Mfd cap. Then I get a slow turn on time and no hum. I get a little less, about 10'55V to 11.35Vdc and I have found that most tubes need a little less DC than AC to get the same emissions.

I can post the circuit. real simple. I use a 12V/24vAC wall-wart, a bridge and reg for DC filiments, then a RS trans backwards for HV.

Then I have no UL problem when building rack gear.

The 6.3VAC workd as the centet tap of the 6.3V winding is grounded by the supply so the AC is "balanced out" in the chassis. Look ata Fender Git-Amp. The whole PS is on one side where it is "hot" with AC. The pre tubes are way at the other end where the chassis is "cool" with little AC.

Look at examples of grounding and chassis current flow in older gear.

I went to DC for totally quiet circuits.

gEO... :green: