NewYorkDave Mila Dual Tube Preamp Help Thread

[babyhead]

Hey-

Im not very sure about the mounting of your electrolytics. With this layout style, I wouldn't use radials unless it wont move. Maybe mount them vertical or just use axles. Those grounds need to be tight.

If you are getting noise with the pot all the way down, it's in the output. Possible culprits:

Cleanliness of V2? It can be wired with 3 leads and heaters and placed closer to it's components.

C8 not fully grounded?

Amplifying PSU noise?

Tube selection? These things are old and need to be sorted. They sometimes need some burn in.

Cold joint? I like to build this way because I can solder hot.


You are close. I find it really exciting and rewarding to do projects on this level.

BH

 

[sethyrish]

Would Beyer TR/BV 352 003 955  be OK for input. I couldn't find any info. I believe they can be used 1:1, 1:3 and 1:5. The Guy selling them also has Lundahl 1550 AMORPHOUS CORE, Jensen JT-16B,Jensen JT-10KB-DPC. More expensive though.

 

[Kingston]

It would work ok with the 1:5 ratio setting. But you would wasting much of the preamp headroom (and gain) with it, never really running "close to the edge", where it starts to sound better (THD). You would be better off with something closer to 1:10, or even higher.

 

[pjdiy]

bh,

I am going to have to wait on the Mila as I am traveling this week.
I won't be back home till after Cinco de Mayo.

I wish I had the time but will keep you guys updated when I get back.
I want to see how quiet I can get this puppy. I also have an extra tube to switch out and see if that makes a difference.
Thanks for the input on the output. Yea a zeroed gain and having the meters jumping around I wanted something more specific to focus on.

PJ

 

[pjdiy]

Hey all,

I did get some time and cleaned up the wiring and move the tubes where
I could minimize the wiring lengths. I substituted one extra tube I had into V1 socket.
I also ran shielded wires to the gain pot, grounding only one side to audio grd.

I'm still getting the same amount of "white noise". The meters are still jumping (around ---60 to -55) with the gain zeroed.

I read about better resistor components (wirewound, metal foil) etc.
Tubes are finicky. My condenser mics sound nice and hot. I am thinking of ordering
some more tubes just to see if they are the problem.

later...
PJ

 

[pjdiy]

I'm back and the extra tubes I ordered came in. I subed two new ones in and there is still the same amount of meter activity with the gain off. I am using an E-MU1212M soundcard that has two setting for choosing between -10 db and +4db levels. When I select the +4db the meter activity goes away. The +4db is for pro equipment right?

- E-MU1212M specs for the Analog Level (software selectable):
- Professional: +4dBu nominal, 20dBu max (balanced)
- Consumer: -10dBV nominal, 6dBV max (unbalanced) 

The Mila is quiet when I record with mics. I do not hear the noise but seeing the meters jump around -60 db with the gain zeroed bothers me. I will post some new pics soon of my cleaned up layout.

Got everything working and quiet too. Here are the pics.
Thanks again for all the help from everyone here at DIY.
http://img40.picoodle.com/img/img40/2/5/17/pajamas/f_DSC00102m_6c5f32b.png
http://img31.picoodle.com/img/img31/2/5/17/pajamas/f_DSC00103m_5d92737.png
http://img33.picoodle.com/img/img33/2/5/17/pajamas/f_DSC00104m_2f77b8f.png
http://img31.picoodle.com/img/img31/2/5/17/pajamas/f_DSC00106m_70a75db.png

 

[pjdiy]

With an  LCD mic what setting are you using for gain and range?

Let's say Gain is numbered from 1 to 10. Do you run the preamp hot like at 6 or higher with the range selected for max and pad off?

What I am asking is it better to operate the preamp at higher settings?

Does this make sense?

 

[Stagefright13]

I use mine with an LDC at range 1 gain about 4-7 depends on mic.

For DI bass I use Range 2 or 3. And somewhere near the same gain.

Range for me is the the 3 position switch. Gain for me is the volume knob.

John

 

[maxheadroom]

maybe a stupid question, but the spezifikation ( http://musiciansgig.com/mila/MILA1-Updated.pdf )
about the OUTPUT:

Source impedance: 58ohms. Minimum load Impedance: 600Ohm

what does source impedance exactly mean?
i thought source impedance (from the input) is 200ohms (according the input transformer, prim. winding)
and output impedance (from the output) is 600ohms (according the output transformer, sec. winding)

 

[greenmanhumming]

I'm not totally sure if this is the right place to discuss the circuit of the MILA rather than the building. If there's a MILA theory thread please redirect me!

I have loads of questions about the schematic.

* why use 100k pot and 1uF cap at C3 and VR1, why not use a 470k or 1M pot and a smaller cap?

* does a 270 ohm grid stopped resistor really do anything much? obviously it must but what is the -3dB frequency for this? (I'm used to guitar amps and 68k grid stoppers)

* has anyone built this toplogy with a different valve in either or both positions? seems to me with appropriate changes for bias etc it could work with 12at7 12au7 or even 5670 in the first pos.

The motivation for asking is that my intended PT is only rated for 1.5A heater and 4x 12av7 for 2 channels of MILA will draw 1.8A  - do 12av7 actually draw 450ma each in reallife or do they vary with diff brands etc? OTOH the HT draw will be much less than the transformer is rated for, so maybe it will cope...

* what is the best 12AV7 to use? I have heard the RCA is the best , and have ordered some of those, but would like to know other peoples experiences.

* why does the MILA use a 1uF cap in the feedback loop, but the NYD one bottle (which is closely related) needs 22uF? I presume its to do with the maximum NFB applied in the one bottle being higher (ie lower gain achieved through NFB)

*has anyone used the MILA without an output transformer? i presume it would be fine into 10k loads since the output tube is paralleled from the onebottle and so is presumably approx half the output Z. is this right?

*i'm thinking of using a lundahl 1538 1:5 input transformer, i guess the lower gain would make up for the higher gain at the other end from not using an output transformer, but is there any disadvantage to running lower levels through the whole circuit? I'm thinking levels vary much more than this anyway with different mics etc.

 

[buschfsu]

doesnt look like there are recent answers to this thread but ill try.  what is the best way to pull 48v dc out of one 250/6.3v xfo.  i can think of dividers on the 250v (heat) then maybe zenners to regulate.  multiplyers on the 6.3v then rectify but what is the most elegant (few parts, no signifciant impact on B+ that would affect audio path) solution?

thanks

 

[greenmanhumming]

i would use the 6.3v if I had to, i don't  see how to do it from the HT without wasting loads of power.

the 48v may require more current that you have spare on the HT too...

I'm running my 48v from a separate external switch mode regulated power supply bought off the shelf as a computer thing... seems like the best solution if you have a trad tube transformer without a suitable winding for 48v

there also the 2 12v transformers back to back thing, you then get your 48v from a voltage tripler off the 12v... have a look at the gyraf G9 circuit... but easier to just buy a separate 48v supply.

JLM audio in australia sells one, for their kits, but i think its a fairly standard computer type supply.

 

[shabtek]

taking a break now...will throw out some guesses etc.
it has been mentioned before that deriving yer fantum from the b+ is a bit riisky; maybe not a worry for chinois condensers but i'd think twice if i had any kraut mikes... there is a chance of b+ showing up at the xlr if something inside lets loose.
I'm not totally sure if this is the right place to discuss the circuit of the MILA rather than the building. If there's a MILA theory thread please redirect me!
* why use 100k pot and 1uF cap at C3 and VR1, why not use a 470k or 1M pot and a smaller cap?
I don't know, exactly, guess better bandwidth and frequency response. output z of 1st stage can drive 100k cleanly.
* does a 270 ohm grid stopped resistor really do anything much? obviously it must but what is the -3dB frequency for this? (I'm used to guitar amps and 68k grid stoppers)maybe, not as much gain as a marshall,not as prone to amplify noise/oscillate as a guitar amp but GS is a habit/potentially helpful* has anyone built this toplogy with a different valve in either or both positions? seems to me with appropriate changes for bias etc it could work with 12at7 12au7 or even 5670 in the first pos.
plugged in different valves, they work/pass clean signal but not optimal
The motivation for asking is that my intended PT is only rated for 1.5A heater and 4x 12av7 for 2 channels of MILA will draw 1.8A  - do 12av7 actually draw 450ma each in reallife or do they vary with diff brands etc? OTOH the HT draw will be much less than the transformer is rated for, so maybe it will cope...try it, play the tolerances, if too much I  then rebias for a 12__7 with lower draw* what is the best 12AV7 to use? I have heard the RCA is the best , and have ordered some of those, but would like to know other peoples experiences.telefunken or nos mullard ...seriously, ones from canadian-made color tvs...whatever
* why does the MILA use a 1uF cap in the feedback loop, but the NYD one bottle (which is closely related) needs 22uF? I presume its to do with the maximum NFB applied in the one bottle being higher (ie lower gain achieved through NFB)
Z. impedancses*has anyone used the MILA without an output transformer? i presume it would be fine into 10k loads since the output tube is paralleled from the onebottle and so is presumably approx half the output Z. is this right?
less than half Zo, will be fine*i'm thinking of using a lundahl 1538 1:5 input transformer, i guess the lower gain would make up for the higher gain at the other end from not using an output transformer, but is there any disadvantage to running lower levels through the whole circuit? I'm thinking levels vary much more than this anyway with different mics etc.
should be fine, noise,s/n may not be optimal, but 'twill work fine...if no need to drive heavy load, this maybe is not the best design/overkill. if yer power trans can do 4 filaments you could build 4  1-bottles. 1+ mike 1+ hiz cascadeable stages for dirt ,

 

[buschfsu]

coupla sourcing finds...

weber has 90 degree brackets pre cut for 2 9 pin tube sockets for 7 bucks- doesn't show up online but....

"If you are planning to order more than just the brackets, you can place an online order for the other items, and in the comments, just write "2x Bracket-SKT9, $7 each" in the comments, and we will add it on.  Let me know."

like in these kits...https://taweber.powweb.com/store/modkits.htm#Cascode

weber also has a 275v/50ma  6.3v/1A power transformer for 12bucks
WLOPO_PT

https://taweber.powweb.com/store/magnetic.htm


i had to use the WRVBPT which is $25 cause im building 2 channels (1.8A total for the heaters) and this guys rated at 3A.

 

[buschfsu]

looking at the G9 phantom circuit.  can i add a diode after D12 and two more caps to make a voltage quadrupler for my 6.3V heater winding?  that would give 50.4v into the regulator vs 60v in the G9.  is that enough overvoltage for this design?

thanks
jason

 

[buschfsu]

rethinking this a bit.  now im going to adjust the dc bias voltage divider to dump 48v dc into the heater ac.  this dc bias reduces the ac hum.  anyway bump the bottom resistor to 170k and you got 48v dc  now take that into two serial 24v zenners and i have 48v regulated.  my xfo can handle small phantom current.  anyone see and issue with this plan?

 

[b3er]

I've never tried using that low of a dc voltage to float the heaters.....75-80V is what I've read.  Mine is at about 70V and I use a separate supply for the phantom.  BTW My Mila sounds great save for a 120hz buzz on the output.  Have tried everything to get rid of it (lifting gnds etc) ,but it won't go away. any suggestions?

pete

 

[Stagefright13]

I just put a separate and tiny board in the ones I made for phantom. Tapped off the AC line in. Didn't have any hum problems from it. But I have a secure and separate 48 volt supply. The one I used from Fivefish worked just fine.

John

 

[buschfsu]

I thought he used 30v or so on the original design.  I could gap the b+ was just trying to save parts.  Anyon know why floating ac heaters reduces hum?

 

[emrr]

maybe a stupid question, but the spezifikation ( http://musiciansgig.com/mila/MILA1-Updated.pdf )
about the OUTPUT:

Source impedance: 58ohms. Minimum load Impedance: 600Ohm

what does source impedance exactly mean?
i thought source impedance (from the input) is 200ohms (according the input transformer, prim. winding)
and output impedance (from the output) is 600ohms (according the output transformer, sec. winding)

Source is the actual output impedance of the preamp.  600 ohms is the minimum load to put on that output, to achieve rated levels and distortion.    The bits quoted have nothing to do with the input.  The input impedance of 200 ohms reflects the suggested output impedance of the mic, which is a typical value.  The actual input impedance is much higher; he quotes 2K.  That will be dependent on the transformer used.  If the pad is engaged, it's around 2.2K, dictated by the pad values.