Turtlerock Forum

Wow - seems ages since I last posted here. Still, silence means nothing I had important to say, I guess.

Anyhoo, tinkering with building 2 channels of NYDave's MILA P2P value pre. The idea is to meld together some JLM interface kits (DI, GoBewteen Plus, VU) and a cutdown version of the regulated power supply from the Gyraf G9 to run the tube pre amplification, Xformer Lundhals in and Edcor's out.

I've got my P2P layout looking pretty (ugly), power resisters all down one end and grid resistors at the tube sockets, separated DC 12.6V heater, etc. Just wondering about the effort vs buying a Sebatron or building a Hamptone, so any war stories greatly appreciated.

Sorry - I should have posted the schematic so you know what I'm talking about.

http://musiciansgig.com/mila/MILA1-Updated.pdf

OK, first question. After all, the joy of DIY is learning the whys as much as the hows ....

NYDave says that signal ground and chassis ground should be connected at one point only: at the power input.

If I use the term "earthed" for things like chassis, control bodies, transformer cores, connector bodies etc that are not part of any electrical circuit, all these I understand should connect to the earth lug of the 240V supply. The intention is that they never conduct, they are there for safety so that if something goes wrong you don't get fried adjusting the volume knob.

The active and neutral lugs of the 240V supply go to a transformer for isolation and voltage step up/down/split as required. In the Gyraf scheme, this would create one 15V AC supply to a voltage tripler for 48V, another 15v AC that gets rectifed for a 12.6V heater (and in this concept a 12v meter circuit), and a rectified 300V B+ tube supply. The "0V" signal in the two rectified circuits and the (effective) "0V" arm of the 15VAC circuit all get tied together - which I understand, creating a common "0V" reference for the schematic.

Do I understand the schematic correctly in saying that the "0V" signal should be connected to the chassis? If so, why? Won't this cause current through the IEC ground circuit and trip my ground leak detector? Or does it simply reference circuit 0V to ground potential? If the latter, can "0V" be floated without referencing ground?

All learnings greatly appreciated.

without any reference to this unit in particular as I haven't looked at the schematic, however on general principles:

Do I understand the schematic correctly in saying that the "0V" signal should be connected to the chassis?
YES, at one point

of far greater importance, if this unit is not being built on a proven pcb but rather "point to point " then how all the 0V points come together is the key to low noise and stable performance.

If so, why?
sometimes it improves circuit stability, but is not always necessary. Some devices are hard grounded, some soft grounded through a resistor, say 10R, some RF grounded with just a capacitor, say 0.1uF, some open circuit. There is no simple answer to why .... it is a design aspect


Won't this cause current through the IEC ground circuit and trip my ground leak detector? Or does it simply reference circuit 0V to ground potential? If the latter, can "0V" be floated without referencing ground?

NO, current won't flow through your mains earth, it is not part of the circuit, unless your mains transformers fail, even so your " ground leak detector " doesn't actually trip by looking at the current flowing down the earth wire.

Rob

Rob the Genius Tech wrote:sometimes it improves circuit stability, but is not always necessary. Some devices are hard grounded, some soft grounded through a resistor, say 10R, some RF grounded with just a capacitor, say 0.1uF, some open circuit. There is no simple answer to why .... it is a design aspect

Thanks - I think the "open circuit" approach is what I have usually experienced, and so the hard ground approach got me wondering. From your response, I was also able to track down this site for more detailed information -

http://sound.westhost.com/earthing.htm

Rob again wrote:NO, current won't flow through your mains earth, it is not part of the circuit, unless your mains transformers fail, even so your " ground leak detector " doesn't actually trip by looking at the current flowing down the earth wire.

For completeness, earth leakage detects current differentials between neutral and active wires, correct? But this won't be a problem because neutral and earth are themselves connected at the switchboard - they reference the same potential. Have I got this right?

Got my tubes for this project today. The schematic says 12AV7, so I've sourced some NOS Mullard E180CC, which perhaps need a little more careful attention to noise than the stock AV's but are apparently worth it sonically.

My question to the great wise ones of the forum. Dave's schematic calls for a 10K:600Ω output transformer, but that's just so the pre can drive output loads down to 600 ohms.

I don't need that - this will go directly into my DAW's line inputs (10K load). My understanding is that I can use a simple 1:1 on this basis like the DYNE's (600Ω:600Ω) I scored from the great transformer reallocation. Am I right? And more broadly, I know that transformer ratios are based on windings and that reflected impedance is the square of the ratio - but do transformer impedance ratings have any actual bearing or are they simply another way of stating ratios?

So is a 10K:10K transformer different from a 600Ω:600Ω and if so, in what way?

10K:600R output transformer will do two things - transform the relatively high output impedance of the tube amplifier to a low impedance suitable to drive the outside world and to shed 12dB of level. While you may only want to drive ~ 10K inputs, as a general rule to do so without loss you want a output impedance a factor of 10 lower, ie 1K or lets call it 600R, indeed a factor of 100 is typical, ie 100R. The unit is also driving cable ... maybe not much but again a high output impedance is not suitable for driving long cables without HF loss or slew rate distortion.
The difference between a 10K:10K and 600R:600R transformer is the insertion loss when loaded, both broad level and frequency dependent level.
You really do want a 10K:600R transformer for this job.

(slaps forehead)

Once again the concept of impedance matching rises up to bite me. Its the classic "of course" once Rob's explained it, but that's due to your teaching ability, mate.

So in fact impedance matching is a misleading phrase - you don't equate the impedances at all, you look for a factor of at least 1:10 output impedance to input impedance. I have read that before but it obviously did not sink in.

Cinemag has a good 10K:600, so unless anyone has a pair lying around ..... ? Anyone have listening experiences with those cheap US Edcors (I've seen forum comments)?

Hi Chris,

I can highly recommend Harbuch transformers. They should have an off the shelf 10k:600 in stock and I think it would cost less than a Cinemag... plus you'll be supporting an Australian manufacturer.

They can also do one off custom transformers to your specification at a very reasonable cost.

9/40 Leighton Pl
Hornsby
NSW 2077
(02) 9476 5854

http://harbuch.com.au/doc/INPUTOUTPUT_TRANS_May2010.pdf

Just checked their catalogue and what you need is an AT-111.

Although the schematic shows the Edcor has a centre tapped primary and secondary it's irrelevent in this circuit.

Hope this helps

As Always +1 for Harbuch, support local, they're as good if not better than many on the market and you're supporting an Australian manufacturer

I got an email from Harbuch a few weeks ago, saying they dont do tube related transformers anymore.

I was looking at purchasing a quantity of: AT124, AT2004 & AT2007

Peters response was:

Hi Michael

We no longer stock or manufacture Valve devices, sorry.

Kind Regards
Peter Terlich
Manager/Director

I don't think 10k:600 would be considered exclusively a tube transformer, more for general line impedance matching.

But it is a shame they're not supplying a full range of transformers anymore... maybe there is not enough demand.

Hope they're still doing custom winds.

If you cant get Harbuch's maybe you want to consider Carnhills, from Audiomaintenance in the UK, a VTB2291 10K:600 is about $AUD30, planning to buy some myself, let me know if you do, we can combine order and share the shipping costs.

a VTB2291 10K:600 is about $AUD30

That's hard to beat... genuine Carnhill?

I guess that's one of the benefits of a strong Aussie dollar.

Hey Junction

Harbuch Transformers wrote:Have reviewed your situation and the very involved manufacturing process of the AT2004.

The original AT2004 had two interleafed Primaries and a Secondary.
Version two has three interleafed Primaries and two interleafed Secondaries.
Version three (current version) has four interleafed Primaries and three interleafed Secondaries.

Now I understand their difficulty to manufacture and the associated cost, but their coupling characteristics are outstanding.

If I could get the price of the AT2004 down to say $60, would you be interested in going Harbuch instead?

If I could get the price of the AT2004 down to say $60, would you be interested in going Harbuch instead?

Hey Chris, thanks, but i have got to order a bunch of stuff anyway from audiomaintenance and I have used Carnhills before in some Neve-elish pres, which sound awesome, they have a 15K:600 tranny also for $AUD30, so think I'll go down that route.

you need to check a few things. If it is a tube output transformer then 15K to 600R may be a resoanable ratio. But can you get say +20 dBu out of the 600 R output? Most tube output transformers are quite large if they are designed to output decent levels. You need to consider the levels on the 15K side and ensure your transformer can deal with it at the lowest frequencies of interest. ( distortion increases as freq goes down )

Thanks Rob, yes I understand, cant find specific specs for the Carnhills, but given the description, one would assume we would get a high output level, as follows;

CA-18-VTB2380 - Transformer: Audio Output (15K:600R Ungapped - Flying Lead Version)
Carnhill Transformer VTB2380
Audio signal (High Output ) - Ungapped Core - Flying Leads
"15k:600R" step down transformer
A "Broadcast/Pro-Audio" quality signal output transformer capable of delivering high level output at low distortion.
Dual primary and secondary windings provide the option of different input/output inpedances to suit source and load requirements.
£20.30 each (Ex.VAT)

Cheers

Front Panel for the MILA's. Or should I say Röhren 180cc Studioverstärkeren.



Two channels, XLR and DI in, XLR out, 48V, phase and pad, saturation and volume controls plus "golden glow" VU meter.

chris p wrote: Or should I say Röhren 180cc Studioverstärkeren.

Dost es maken guten sounden ? Nice panelwork.

Mine is not ready yet for ignition

Junction wrote:

chris p wrote: Or should I say Röhren 180cc Studioverstärkeren.

Dost es maken guten sounden ? Nice panelwork.

Mine is not ready yet for ignition

NYDMila-DaBomb.jpg

I must say, as an ex Powder Monkey in the RAAF who is now into audio, that excites me on two levels...

Damn Junction, that's pretty.

Not ready for ignition yet. I'm still just beginning, really. But I do have all the bits and pieces, including the old Lundhal 10:1 input trannies and some NOS Mullard E180cc tubes for the amplifier stages. Power supply will be IC regulated linear.

All built into an old instrument case Rick had lying around when I last visited him.

BTW, moving to Canberra over the next 6 months (and commuting in the meantime). Look forward to catching up with any CBR t'rockers out there.