One-Bottle Preamp

[skal1]

At some point yes , at the moment no funds ... thanks

skal1

 

[skal1]

hi guys just wanted to check this is the way to hook up the unbalanced 1/4 jack , take a look

 

[mac]

I have been playing with my NYD one bottle these last two days and am about to make a new layout to reduce hum.

One observation with my current little birds nest made from recycled bits I have lying around - cathode wire and grid wire dont like each other much.

Is there a general rule about cathode and grid wires crossing only at right angles?? trying to learn as much as possible before I build another one with a better layout.

Mac

 

[Kingston]

Cathode wiring is less sensitive in general, in the end that's just your ground reference in this design. Of course a short fat wire (or ground bus) is best here. But grid wires should be kept short indeed. That's a very high impedance point. It will listen to whatever it finds, be it actual grid wire, or static from the world.

There is no general rule, other than to keep wiring as short as possible.

Perhaps you should post a picture of the layout. Like how is your heater wiring and is it AC or DC?

 

[mac]

Heaters are DC. I built a version of gyraf's ps board with a few options, because I wanted to use this box as my ongoing test bed box. Happy to post a pic if you need a laugh!
Early days for me with p2p (tag boards) and tubes...

 

[mac]

One other quick survey question.  Is it poor practice to attach the zobel network components directly to the input transformer? I notice a few layouts have these components on the tag board an then run screened cable to the trafo.
Mac

 

[MeToo2]

One other quick survey question.  Is it poor practice to attach the zobel network components directly to the input transformer? I notice a few layouts have these components on the tag board an then run screened cable to the trafo.
Mac

Don't see why not. As far as I'm concerned if you're going to go the point to point route, then soldering the components as close as possible to where they have effect is actually good practise, not poor. Especially for very high impedance points like grids e.g. soldering grid stoppers and HF filter (miller) caps directly onto the tube socket. I'd rather see a logical point to point layout of components arranged following the signal path and preferably something like in a circle around a tube socket, rather than a nice row of components neatly lined up in parallel with their color codes all in line on a tag board. All of that extra cable just adds extra/unpredictable coupling paths.

If you want to reduce hum then making sure your high current (heater wires) are physically well away from the rest is a good starting point (DC heaters need a power supply which can create it's own switching noise) so they are a prime source of stuff leaking out. Keeping the input transformer physically far from the PSU will help. Physical distance is your friend here. Remember that anything that leaks in at the front end is also amplified along with your signal, so make sure that is clean. And grid connections are a prime source of stuff leaking in. Star grounding is another must.  So it's all pretty logical.  There's some good tips here http://www.freewebs.com/valvewizard/heater.html http://www.freewebs.com/valvewizard2/grounding.html http://www.freewebs.com/valvewizard/gridstopper.html

Check out the LA2a physical layout....  you'll see stuff soldered everywhere. But the basic principles outlined above have clearly been applied http://www.recproaudio.com/diy_pro_audio/diy_files/la2a/la2a_layout.jpg

 

[lassoharp]

Early days for me with p2p (tag boards) and tubes...

If you're using tag board I'd keep a close watch for misformed solder joints.  Very easy to do with tag board.  Maybe a loose ground somewhere?  Also check your PS filtering - is it the same that Dave speced for the original?  SE circuit so will take a fair amount.

 

[mac]

Thanks for the responses guys, and the extra reading  ;D

As I said up the page - I am not too worried about this first layout, as I new it was a rough and ready rig - put together to learn - before having a go at a good and robust layout.

The other thing I am trying to get my head around, is the DC blocking cap in C4 (designated on Dave's design as 2.2uf). What impact does lower / higher capacitance have on the circuit in this position? I understand that it is working with R10 (220k) to block DC (+106v) and to present appropriate impedance to the output, but I have some 0.47uf caps here that I would like to use, and wasn't sure if I should be changing the value of R10 accordingly. Does lowering the capacitance increase the LF rolloff of the output?

Mac

 

[MeToo2]

Thanks for the responses guys, and the extra reading  ;D

As I said up the page - I am not too worried about this first layout, as I new it was a rough and ready rig - put together to learn - before having a go at a good and robust layout.

The other thing I am trying to get my head around, is the DC blocking cap in C4 (designated on Dave's design as 2.2uf). What impact does lower / higher capacitance have on the circuit in this position? I understand that it is working with R10 (220k) to block DC (+106v) and to present appropriate impedance to the output, but I have some 0.47uf caps here that I would like to use, and wasn't sure if I should be changing the value of R10 accordingly. Does lowering the capacitance increase the LF rolloff of the output?

Mac

C4 doesn't just interact with R10. That's the least important resistance to consider. R10 is mainly there to prevent the output floating higher than 0V (due to leakages etc). More importantly, C4 also interacts with Rp (the plate impedance) and don't forget the load impedance, which is not shown on the schematic. Both of these resistances are generally much lower than R10. 0.47uF is almost 5 times smaller than 2.2uF, so the LF roll off will kick in at ~5 times higher the frequency. Impedance of 0.47uF is 8K at 42Hz (low E). It will almost certainly affect the low frequency response and probably make your pre much more load dependent.

 

[lassoharp]

C4 value will change LF rolloff point.  Lower C = higher corner.  With the stock 2.2uF value and no OT (assume you're not using an OT) you'll want higher Z loads - 5K or higher to keep good bass.

With .47 there into a 5K load it looks to corner at around 70Hz - and that's ignoring the effects of the NFB.  Best to sub each value and measure to see what it's actually going to do.



 

 

[bluesbaz]

I've bult these with a shure m68 input transformer (1:10) and it seems i have little control over the gain. The rest of the circuit is the same. Any ideas where I should be looking?
-Thanks

 

[lassoharp]

I've bult these with a shure m68 input transformer (1:10) and it seems i have little control over the gain. The rest of the circuit is the same. Any ideas where I should be looking?
-Thanks

Outside of what the 100K pot offers control wise about all you can do is either use pads before the transformer or add a pot or attenuator on the transformer secondary.

The stock NFB gain control is only speced for a range of 18db.  I think there's been discussion on how to extend that range but there will likely be some compromises.  Anything in particular you need it to do? - go all the way to "zero" ?

 

 

[bluesbaz]

Going to zero would be great! Im having a hard time using this thing on anything at the moment.
-s

 

[lassoharp]

To go to zero you can add a pot or stepped attenuator (if you need a non standard pot value for the sec load) on the input transformer secondary.

For a 1:10 transformer a 100K pot should put you in the ballpark.  It's not the ideal for a mic pre but will work well enough and a good quality pot should not add excess noise.  Attenuating heavily with the pot would likely mean you were miking a loud source so signal to noise figure wouldn't suffer too much.

If you ever plan to build another tube pre the RCA BA-2C and Collins 6Q both have interstage vol which gives the full range of attenuation with no compromises.

 

[mac]

Thanks Lassoharp,
Actually I am thinking of doing one channel stock as per Daves schem, and second channel with an old Trimax transformer on the output. It is a beautiful gigantic old thing TA746 6000:600 centre tapped on the primary.

As you say I shall sub some values and measure - particularly with the second channel (particularly for the value of R10).

Is there any concerns in tying two channels together with one single "ground buss" arrangement?

Mac

 

[Kingston]

Is there any concerns in tying two channels together with one single "ground buss" arrangement? see sketch attached...

Yes. They have to be mirrored. Since you're attempting a bus-ground-to-star arrangement we must assume ground always flows to one direction (resistance). In this project the best choice is from inputs downstream to output stages, then to the PSU, and finally to star. What you absolutely do not want is input stage of one channel ground right next to another channels output stage ground. The heavier currents from the output stage are pretty much guaranteed to pollute the more sensitive input stages.

See here for a tried and tested solution: http://www.michaelkingston.fi/files/mila-1_turret_layout_kingston.pdf

Another variant: http://notinteractive.com/stuff/audio/mila/mila-dual-layout_0.1.png

Or just use two equal layouts for both channels, but take both separately to PSU ground.

 

[lassoharp]

It is a beautiful gigantic old thing TA746 6000:600 centre tapped on the primary.

That sounds like a good candidate for a big push-pull amp . . . . maybe some goodie from the ABC schematics vault?

 

[mac]

Appreciate your comments Kingston...I will have another go at it...

Lassoharp - yes I have struggled to find out much info about the TA746. I dont know the ABC schematic Vault - please do share  ;D

Mac

 

[lassoharp]

http://www.retrovox.com.au/ABCsse.pdf


http://www.retrovox.com.au/datapage.html


Some really cool designs.  There's a thread about the Trimax limiter here in the backfiles.