[CButterworth]

I have been running my DIY Aikido preamp for a week or so now, I have implemented a few methods to reduce ground-loop hum, but I do still have some.  It is around 60Hz and quiet enough for most music, although it is a little intrusive with some classical pieces.

A fellow Aikido builder, Bruce, has suggested that I may not be drawing enough current to allow the chokes to work properly.

The PSU is:

125-0-125 (100 mA) going into a Graetz bridge.
1.5uF cap
10H (102 ohm) choke
60uF cap
10H (102 ohm) choke
60uF cap

I should see around 360VDC, which I do (thereabouts).

Now if I remember correctly, when I leave out the 1.5uF cap (making it choke input), I got something like 260V.  However, I should be around 225V if all is working correctly.  I am using four 6sn7 tubes with a total current-draw of maybe 20mA at the most.  A choke input running on 250V with a 10H choke requires at least 25mA for it to work correctly (according to the Morgan Jones book).  If not enough current is drawn the output voltage can increase up to sqrt2 * V.  As my voltage on choke-input did not get that high, it seems that my current-draw is not that much lower than required.

Does the minimum current-draw still apply for a cap-input filter system?????

Bruce suggested that I use a 50K 5W resistor across my final cap to act as a current-draw.  Assuming 360VDC the current drawn by this resistor would be 360/50K = 0.072A or 72mA.  Is this correct?

The power dissipation of the resistor would have to be at least (360*360)/50K = 2.592W.   Is this correct??

OK, so my transformer is rated to 100mA, so I guess that with 72mA drawn by the resistor, I would be left with up to 28mA for my tubes.  This seems like a slim margin to me.  A 100K resistor would draw 36mA and dissipate 1.296W.  This coupled with the tubes would give a total current draw of around 50mA.

Anyway, does the above sound reasonable?

I would certainly appreciate any opinions, advice, etc.

Thanks,
Charlie

[mgalusha]

Hi Charlie,

I've heard of the Aikido preamp but am otherwise not familiar with it, so take this with a block of salt.

I think you missed a zero in your current calculation, 360/50000 is .0072 or 7.2mA. Your wattage is correct though I calculated it via V*I (360*.0072) = 2.592W.

So if your main tubes are drawing 20mA and you add the 7.2mA from the resistor this gets you over the 25mA you mentioned and doesn't stress out the transformer. I'd probably opt for a higher wattage resistor just for increased reliability. I've blown a 5W resistor in my preamp even though it's dropping less than 1W. The 12 watt Mills would probably work nicely for this.

mike

[JoshK]

I think Mike answered your question well.  I am actually doing a power supply design for my 6sn7 aikido too.  But I haven't worked out as much of the math yet, as I am building the case right now. 

What I was thinking of doing, that would probably work for you too is using a few voltage regulator tubes after the CLC (I am not using a second 10H, just one) which pulls, depending on the regulator, anywhere from 10-30ma.  It also provides a bit of regulation and they look cool.   

The 12B4A linestage threads on diya are chock full of PSU's and long discussions on using voltage regulator tubes (OD3, OC3, OA3, etc).  You stack the appropriate ones on top of eachother until you arrive at the needed voltage. 

One question I have is the 360VDC.  Isn't that a bit high?  I saw the target for the 6SN7 aikido as 300VDC B+.  I was planning to use two OD3's for this (150V each @ 10ma iirc). 

One hint: voltage regulator tubes work that much better when they have a CCS in front of them.  I was planning to use a cascoded pair of DN2540s or a IXYS 10M45 or whatever's clever.  I'll try to find the schematic, that is an example of this.  Its Kevin Carter's Raleigh line stage PSU from years ago (not his newest).  It should work well.  Below is a link to his chapter on the PSU developement.

http://www.raleighaudio.com/chapter_5.htm

About hum.... First the simple stuff, as that is usually what tends to get me.  Did you reference the heaters to half B+ as suggested by Broskie?  Are you running heaters on DC or AC?  Twist all heater wires tightly and make sure all power and signal wires cross at 90º in your layout. 

If it is just your iron that is humming then the hum shouldn't get louder through your speakers when the volume is raised (with no signal going through system).  Check that too.  Put your ears near the iron and near your speakers.

Don't fear, this is a right of passage.  I've built my fair share of humming projects.  The NYAR got to hear my humming 5687 linestage until I worked that out. 

Then when you got these things covered you sketch out your grounding scheme careful and sort out loops and experiment with different schemes until you find the one that breaks all the loops. 

[JoshK]

I'd rather link that upload the file to my gallery, but I couldn't find where I found this...so here is the file.  Again this is the Raleigh Line Stage PSU from '03.  So credit is where credit is due.

Note it uses a hybrid Graetz bridge, diode dampers (which I am using too), .8uf > 8H > R || C || C > R > CCS > tube regulators.  I'd bypass the tube regs with a .056uf cap, russian teflon is good for this off ebay, to shunt the bit of noise the tube regs make.  Then you have some nice glowing tubes in the supply too!  They don't require heaters.

[JoshK]

Let me just make one suggestion.  After having displayed all this talk about tube regs, I'd still suggest you try the resistor first, since that is the simplest way to check if that is what is causing your hum. 

[Occam]

a rhetorical question -

are you dealing with ground loop hum, or hum (ripple) associated with your power supply?
These are two different questions. If it is a question of ripple, there are a number of way of dealing with it. As Josh has provided, there are 'battleship' approaches, which deal not only with ripple, but apply very sophisticated measures via regulation, multipole filters, etc....
Another possible approach to hum is a tuned trap -
http://diyparadise.com/tunedchoke.html

But the issue might truly be one of a ground loop, interaction between components, which is a whole 'nuther story and you need to determine which.
Gotta run to a meeting.....

[CButterworth]

Mike,

Yes, I missed out a 0.  This evening I'll try a 20K 10W resistor.  It should pull something like 18mA and need to dissipate around 7W - if it works, I'll order a higher power resistor.  Given my tubes and resistor choices, I pull around 24mA, so I should get a total current draw of 42mA or thereabouts.

I have managed to reduce the hum quite significantly, and am referencing the heater AC to 1/4 B+ (as suggested by Broskie).  My heater wiring is tight and tucked into the corner of the chassis.  In fact, I still got the same hum, when I totally disconnected the heater transformer (I only did this briefly).

The hum does not increase with volume, it does stay consistent.

As for the B+ voltage, a number of folk on other forums have recommended running 6sn7s up to 380V.  I hope that when I get a greater current draw, I'll get the B+ pulled down to something like 340V.

Thanks for the help,
Charlie