MILA-1 decent build pictures

[Kingston]

Kingston, do you have a front panel express file for this build, or does anyone for that matter?  Thanks, ChrisP

I don't use front panel express and never will. These are autocad directly to the CNC guy.

 

[JamesW]

Kingston

I have just finished my MILA-1 with cmmi-10c input transformer. Did you need a zobel network in your build?

 

[Dr_J]

... wow this looks like a dream machine...  ;D ;D ;D
I would like immediately with building one...  
And I would like to do a turret board now after seeing this - great fun!

ROCK-ON!

 

[Kingston]

Kingston

I have just finished my MILA-1 with cmmi-10c input transformer. Did you need a zobel network in your build?

I used whatever was specified in the transformer datasheet. Don't remember what. No one ever really needs a zobel network. It's just a way to smooth out ringing of square wave edges. Makes spec sheet types happy. Performance does not suffer and sound does not change in any meaningful way.

 

[JamesW]

I had reason to call CineMag and they said don't use any zobel. I am now going through tube selection and testing including 5965 type.

 

[Kingston]

I had reason to call CineMag and they said don't use any zobel.

now that I look at the datasheet, all measurements were done using that 100k resistor as drawn. clue enough.

 

[mitsos]

Quick question for people who may have built this.  How sensitive is it to layout stupidities?  I built one in an old case not really paying much attention to layout, just wanted to get it working first.  I have unregulated, but heavily filtered DC for B+ and heaters, ripple is little to nonexistent.  But I am getting weird harmonic spikes on output, especially with decreasing feedback.  IIRC it was at 250Hz, 360+Hz, ~500Hz, or so. I would have expected 120Hz to be highest since we're on 60Hz mains but I didn't see any.  250Hz was the highest, with the rest progressively lower as F went up. 

I'm going to do a more proper layout, but I was in a hurry to test this circuit and did it P2P to the tube sockets, which has worked just fine with other circuits but not this one, so I'm wondering if others who may have built this may have noticed if it was more prone to noise from badly planned layouts.

thanks!

 

[MeToo2]

Quick question for people who may have built this.  How sensitive is it to layout stupidities?  I built one in an old case not really paying much attention to layout, just wanted to get it working first.  I have unregulated, but heavily filtered DC for B+ and heaters, ripple is little to nonexistent.  But I am getting weird harmonic spikes on output, especially with decreasing feedback.  IIRC it was at 250Hz, 360+Hz, ~500Hz, or so. I would have expected 120Hz to be highest since we're on 60Hz mains but I didn't see any.  250Hz was the highest, with the rest progressively lower as F went up. 

I'm going to do a more proper layout, but I was in a hurry to test this circuit and did it P2P to the tube sockets, which has worked just fine with other circuits but not this one, so I'm wondering if others who may have built this may have noticed if it was more prone to noise from badly planned layouts.

thanks!

Kingston has published his layout for this circuit, and he obviously put a lot of thought into it. It seems to me you're looking for problems with the B+ when the more likely problem is in the ground layout (B-). It could also be the location of the grid stoppers (R6 R10 R14 R15) which should be as close to the grid pin as possible. HF oscillation can also cause artifacts in the audio band.

 

[mitsos]

Thanks for the tips. That's a good point on the ground, it's exactly what happened with a comp I recapped recently. Couldn't figure out where the noise was coming from, then I found a cold solder joint in the ground path  I'm going to look into the ground layout when I get home. The case I used is from an old Ampex 601 that I acquired, and I had been sitting on the parts for a MILA for 3-4 years now, so that's why the rush to put it together.  IIRC I "grounded" each section to the negative lead of its B+ filter cap, and then take each of those to the ground of the main filter caps, but maybe there is a better way?

I'll look into the grid stoppers, but I'm pretty sure those are soldered directly to the sockets.

thanks again

 

[Kingston]

MILA-1 is perfectly ordinary when it comes to layout requirements. Same things apply as every other p2p tube project. Start from hierarchical ground. Work through the stages in priority order. If you manage ground just like NYD drew them in the schematic it's bombproof. Try to avoid rats nests. Shield all critical wires, like the feedback switch or the gain pot, all wires to transformers. Then shield a couple more just to be on the safe side. I find that a big bus bar keeps things clean. If you have VAC heaters you need to keep them very far from the critical tube bits. I wouldn't do VAC heaters in this very high gain circuit. Too hard to get them perfectly noiseless.

My layout that has no noise issues, other than the tubes themselves if you happen to come across bad ones. But the design is not very critical about this like some mic circuits. I don't remember where I posted the layout so here it is again: http://www.michaelkingston.fi/files/mila-1_turret_layout_kingston.pdf

If you have a passive B+ supply and VAC heaters then some issues might arise that I wouldn't know about. Both were regulated on mine.

 

[mitsos]

hey thanks, kingston. My interent went down last night, but I looked at it. I'm going to start with shielding some more wiring (so far inputs (XLR/DI) to trafo and/or valve is shielded, but not the gain/feedback or output wires.

DC heaters with CRCRC filter.  B+ is similar but has an L stage, something like CLCRCRC.

Not sure I'll get to it today, if not then for sure over the weekend. Goal is to work on it between projects and have it fixed by Sunday night.

As for the bus bar, I have some 1.5 or 2mm solid copper wire around here, is the idea the same as hierarchical grounding, linking local grounds to their stage's capacitor, and then taking this cap's negative lead to the bus bar? 

 

[Kingston]

As for the bus bar, I have some 1.5 or 2mm solid copper wire around here, is the idea the same as hierarchical grounding, linking local grounds to their stage's capacitor, and then taking this cap's negative lead to the bus bar?

Yes this is a similar idea, but you don't even have to get tricky with the cap leads. Just follow the schematic. I really just meant I like the idea of a bus bar from layout design point of view. It forces a very clean layout that's easy to look at. And eventually repair and mod of course!

 

[Dylan W]

Is 250V rating sufficient for C6? I can't find any 500V PIO in that large a size (except for the gigantic bathtub caps).

 

[Kingston]

Is 250V rating sufficient for C6? I can't find any 500V PIO in that large a size (except for the gigantic bathtub caps).

If it's a russian PIO, then yes. They manage something like double the rated constant max voltage.

I had to get gigantic caps because there was no lower voltage 4.7uF PIO available anywhere.

 

[bancho]

Great build Kingston! You actually made me going to build one
I have some questions about caps... does it make any sense to combine pio and teflon caps in the same build. Teflons are said to be not so soft as pio (?)... I'm asking because I can't get 4.7uF teflon... others yes.
How much character does C1 (100pF) contributes? ... could it be vacuum capacitor? >:]  ...
(below is the schematics I have... do you have the same? what/where are the mods?)
Thanks!

 

[Kingston]

I don't know about cap sound anymore. The whole thing is mostly placebo and very difficult to measure so I pretty much refuse to comment on that area anymore. I only used PIO because they will practically last forever. I have no clue about teflon, if they are any good. I also don't believe a word those audiophiles write about caps around the internet.

C1 is part of the input transformer zobel network. Check your transformer datasheet how to do yours.

 

[buildafriend]

Sorry to bump such an oldie. I'm just curious what power supply schematic and power transformer you used. This is a beautiful build.

Thanks regardless

- Jon P

 

[buildafriend]

Quick question for people who may have built this.  How sensitive is it to layout stupidities?  I built one in an old case not really paying much attention to layout, just wanted to get it working first.  I have unregulated, but heavily filtered DC for B+ and heaters, ripple is little to nonexistent.  But I am getting weird harmonic spikes on output, especially with decreasing feedback.  IIRC it was at 250Hz, 360+Hz, ~500Hz, or so. I would have expected 120Hz to be highest since we're on 60Hz mains but I didn't see any.  250Hz was the highest, with the rest progressively lower as F went up. 

I'm going to do a more proper layout, but I was in a hurry to test this circuit and did it P2P to the tube sockets, which has worked just fine with other circuits but not this one, so I'm wondering if others who may have built this may have noticed if it was more prone to noise from badly planned layouts.

thanks!

I had a bit of a noise battle with mine.. my issue ended up being my power transformer but I'm still thinking about switching to a regulated power supply design. I've had issues with putting two NYD units in one box with the one bottle and the MILA but when I think about it neither are an easy pair to be cramming into one box. I also have to consider that these were some of the first tube pre amps I have built so I expected to be fighting an up hill battle. My advice is to order a high quality power transformer and consider using a regulated supply. I would also count on Kingston having a pretty good layout, hes a smart guy. I hope your build did/does work out. post pics!

 

[mitsos]

Thanks, I worked on this again a couple of weeks ago, and it got a bit better, but it's not done.  I redid the layout and used shielded cable almost everywhere.  It's going in an old Ampex 601, using the original power and input trafos. I noticed that the switch picked up noise, but when I would move it further from the power supply (and closer to the gain pot) I lost a lot of signal level, so I'm guessing some kind of negative feedback was going on.  At that point, my friend messed with it a bit and it started blowing fuses, so now I have to find a short before I fire it up again.  ;D  I'm going to look at it again after the holidays and will post a pic when it's all done.

Quick question for people who may have built this.  How sensitive is it to layout stupidities?  I built one in an old case not really paying much attention to layout, just wanted to get it working first.  I have unregulated, but heavily filtered DC for B+ and heaters, ripple is little to nonexistent.  But I am getting weird harmonic spikes on output, especially with decreasing feedback.  IIRC it was at 250Hz, 360+Hz, ~500Hz, or so. I would have expected 120Hz to be highest since we're on 60Hz mains but I didn't see any.  250Hz was the highest, with the rest progressively lower as F went up. 

I'm going to do a more proper layout, but I was in a hurry to test this circuit and did it P2P to the tube sockets, which has worked just fine with other circuits but not this one, so I'm wondering if others who may have built this may have noticed if it was more prone to noise from badly planned layouts.

thanks!

I had a bit of a noise battle with mine.. my issue ended up being my power transformer but I'm still thinking about switching to a regulated power supply design. I've had issues with putting two NYD units in one box with the one bottle and the MILA but when I think about it neither are an easy pair to be cramming into one box. I also have to consider that these were some of the first tube pre amps I have built so I expected to be fighting an up hill battle. My advice is to order a high quality power transformer and consider using a regulated supply. I would also count on Kingston having a pretty good layout, hes a smart guy. I hope your build did/does work out. post pics!

 

[My3gger]

I built quite a few point to point preamps, some of them had high gain, some not, etc. Most of them don't have shielded input trafo's primary, only twisted, secondary shielded in cases when wires are longer than ~5cm, otherwise only twisted. Attenuator pot (interstage) is also shielded. Used 1mm tinned copper wire for ground bus, ground mostly follow schematic. I really try to have it as on schematic, where that is not possible wires cross at right angles (avoid having close B+ and grid..), components are on tag board or turret board (this is much nicer and logical). Found out that layout is not that problematic if runs are short, logical and there are no ground loops.
Position and type of power transformer makes much more difference in hum, i prefer toroids for this reason. All connection to it are twisted, specially primary which goes to the front plate. B+ psu doesn't have to be regulated, but it might be easier for unexperienced to do so because others already calculated filtering, layout, etc. My heaters are always regulated, but i could get away with unregulated because trafos are custom and easier to work with. Also found out that psu's ground has to be connected from the end of it, not near diodes and other parts with switching spikes.
I would say take care of grounds, this is very important. Phantom should have separated ground which connects to ground at XLR input only. Check some Ian bell's posts about it, he explains it very clear. Circuit's ground should only join at "main ground", i use big screw with washers. Beside circuit's ground B+'s and heater's ground also connect only there. Ground pin on IEC connector is connected to chasis with very short and fat wire.
All this is much easier than it seems. Of course making only one channel for the first time saves a lot of pain. There is pretty nice paper about grounding, it's called "Grounding Valvewizar". Plan seems a bit paranoid, each stage does not need all grounds connected ot it's decoupling cap. Actually, i get better results by Kingston's way, follow schematic with short runs, except i don't shield almost anything except pots. Maybe it is so because of short runs.
My first p2p preamp was very, very quiet, i'm sure it has something to do with galvanic isolation from transformers and some thought...