Small valve mixing console

[C.B. - Boudio]

That'll be in a couple of weeks, because of exams ;-)

 

[C.B. - Boudio]

So, the last revision of the first channel is almost done!


The rudimentary EQ under construction.


The Grand Scheme of Things. (nevermind the crappy paint job on the right, it's just some primer to protect the bracket from rusting).



Back part, with connections and the PF86 stage.
In the back, there's a small 220/12-0-12 transformer I use instead of a real Edcor one (which are being ordered), secondary terminated with a 27k resistor. There's also a switching jack: when a jack is inserted, the transformer secondary gets decoupled and a 10M grid resistor comes into circuit.
The switch right behind the pot, with the small print attached, is a +-20db pad.
Every signal wire is shielded.

I have to test it all yet, and I'm very curious as to how it'll work out!

 

[C.B. - Boudio]

So, some tests have followed, and there is amplification! At least, up to the aux send - it doesn't seem to get past the EQ...
A quite distorted signal because of it being a 0,7V line source, but that's to be expected. So, the instrument input seems fine, the mic input still needs some work (probably either the transformer, the phase, the pad, or the switching of the jack - or a combination of those...). At least, up to the aux send - it doesn't seem to get past the EQ...
The g3 NFB at the PF86 is a tad disappointing, so I just connected g3 to the cathode and completely did away with the NFB there (for now).
The gain control (2M rheostat between ground and the g2 decoupling cap of the PF86) doesn't do much; should I install a lower value pot, or stop trying there and go for another system altogether?
There was a lot of oscillation going on, mainly about 100Hz I think (even after rebuilding the B+ supply, which should be really quite clean now), and using NFB from the cathode of PC(F)802 to g1 of P(C)F802 via a 220k resistor got almost entirily rid of that.

 

[C.B. - Boudio]

I'm almost there! I seem to reach decent gain now (which used to be the biggest problem), and noise is surprisingly low. I don't get remotely the impressive noise and distortion figures of, say, an ITB product, but I am quite pleased. Of course, there is still a lot of tweaking possible.
I'm still looking for a way of gain control. At the moment, I use a 1K pot as V1 cathode resistor with the bypass cap connected to the wiper (recipe for scratchiness, I am aware), which doesn't really work.
The EQ section needs some revision (especially capacitor value-wise), and my PL95 cathode follower stage could be improved a bit, too.
Next week, I won't have time to do anything at all, but I'll try to upload the revised schematic at the end of the week. Won't win a beauty contest, with resistors paralleled until I get a defintive value, for example, but, again, it finally is coming to life!

 

[C.B. - Boudio]

Transformers arrived! Four Edcor PC 600:10k's input, and two Edcor PCW 10k:150 outputs (I decided that I won't do the PL95 cathode follower in the master section upon closer inspection).
They are unshielded, so I thought I'd try something myself. I had these sheets of tin (or whatever it is, it's metal but you can get through it with scissors), so I used a piece of those and made it look like this:







I know it looks quite rough, but it's more like a proof of concept. I made sure that the edge of the tin isn't touching any pins.
But will this make any notable difference? It will be more or less in a corner of a sheet metal enclosure, removed from power supplies, and (probably) not too close to a high level signal (about 5 centimeters from a simple shielded unbalanced line level at worst).

 

[C.B. - Boudio]

I redesigned the EQ section. Including it would mean drilling some more holes, but it's probably worth it, functionality-wise.


However, I'm not sure if it will work all that well this way - I'm a bit afraid I made some stupid mistakes.
I also don't know how to calculate its gain loss.
I can't really provide makeup gain, since that would mean bringing in two amplifying stages extra, because I want to keep the phases the same at every in-and output, and if I'll move the EQ between the two pentodes, I'm afraid it will bring even more hiss and the Send will be after the EQ, and I don't know if I want that.
Could someone please shed some light on this?

 

[MidnightArrakis]

I redesigned the EQ section. Including it would mean drilling some more holes, but it's probably worth it, functionality-wise.

However, I'm not sure if it will work all that well this way - I'm a bit afraid I made some stupid mistakes.
I also don't know how to calculate its gain loss.
I can't really provide makeup gain, since that would mean bringing in two amplifying stages extra, because I want to keep the phases the same at every in-and output, and if I'll move the EQ between the two pentodes, I'm afraid it will bring even more hiss and the Send will be after the EQ, and I don't know if I want that.
Could someone please shed some light on this?

>> I believe you could benefit from using a proper schematic design program!!!

It's a whole lot easier to change/delete/edit/move components and connections around on-screen than it is in having to completely redraw a circuit over and over again using a pen and paper. Just sayin'.....

Here.....try this.....IT'S FREE!!!.....

https://www.kicad.org/download/

/

 

[C.B. - Boudio]

@MidnightArrakis I get your point, but I actually adore (re-)drawing schematics by hand. Most of the time, it's easier to understand for myself, too.
And it's some time less spent on looking at a screen, which is also really handy.

 

[MidnightArrakis]

@MidnightArrakis I get your point, but I actually adore (re-)drawing schematics by hand. Most of the time, it's easier to understand for myself, too.
And it's some time less spent on looking at a screen, which is also really handy.

[I actually adore (re-)drawing schematics by hand] -- And.....I get your point as well!!! I have spent many, many, many years literally standing at a "drafting table" manually hand-drawing schematics for the various companies I have worked with long, long ago. "During those days of yore".....those of us who created those ancient hand-drawn schematics drew our schematics as being "works of art". We all took great care with our symbols placement, how to cleanly and clearly draw our connection lines and junction points and also precisely placing all of the component -- REFERENCE DESIGNATORS -- and component values. So, I know of what you speak!!!

However.....if you are going to insist upon sharing your hand-drawn schematics by means of (probably) cellphone photos, then by all means, at least have your camera lens facing your schematics straight-down in order to create a "flat" surface (i.e., not angled or skewed), while also having the schematic evenly well-lit. The image that I see upon clicking on your schematic JPG file is far too fuzzy to even discern that it is a schematic!!!

In other words.....while I can certainly and personally appreciate your enthusiasm with creating schematics as I once did over 40-years ago, things do finally get to a point.....especially when dealing and working with electronic technologies.....that you "need to get with it" and join in with the rest of the world by using today's CAD-technologies for designing and documenting electronic circuits. As I have just mentioned, I used to hand-draw schematics and I also used to -- manually hand-tape -- PCB's way back then!!! However, after creating schematics and designing even complex multi-layer PCB's using today's modern CAD-tools, I wouldn't want to ever go back to my "manual days". TOO TEDIOUS!!! An ART??? DEFINITELY!!! --- A PAIN IN THE A$$??? YOU BET!!! -- (Can you even imagine designing a 16-layer PCB for a U.S. fighter jet -- BY HAND!!! -- using 20-sheets of 22" x 34" Mylar-film stacked on top of one another??? YEAH!!!.....THAT'S how things used to be done)!!!.....

>> Just my inflationary devalued 2-cents worth.....

/

 

[ruffrecords]

Any passive EQ, like the one in your diagram, will have an insertion loss equal to the value of the maximum boost. So you need extra gain to make up for this. There is not getting round this I am afraid. It's basic physics.

Cheers

Ian

 

[C.B. - Boudio]

Any passive EQ, like the one in your diagram, will have an insertion loss equal to the value of the maximum boost. So you need extra gain to make up for this. There is not getting round this I am afraid. It's basic physics.

Cheers

Ian

My thought was that I could amplify the signal up to, say, about 90dB, and so I could keep a maximum output gain of 60-70dB with the pots centered/in bypass and the EQ only being buffered afterwards by that PL95 cathode follower.

 

[ruffrecords]

My thought was that I could amplify the signal up to, say, about 90dB, and so I could keep a maximum output gain of 60-70dB with the pots centered/in bypass and the EQ only being buffered afterwards by that PL95 cathode follower.

You need to manage gain throughout the signal chain in order to manage both the noise performance and headroom. If you need up to 70dB of gain and your EQ also has a 20dB loss then you do need a total of 90dB of gain. But it is important to spread this gain throughout the chain to minimise the chance of clipping and to keep noise to a minimum.

Cheers

Ian

 

[C.B. - Boudio]

That makes sense. Then I do have to place it in between the PF86 and PCF802 I guess...
Time for another redesign then!

 

[C.B. - Boudio]

This is my first try of using KiCad. I wouldn't do it for fun like hand-drawing them, but it does have some usefulness.

I thought of making the EQ active, in the NFB of the second pentode - I think that I'll replace the PCF802 with another PF86 for some more gain, and maybe add a PC92 to buffer the echo send.
C14 then connects to both the PC92 and the PL95, in this version.

SW5 is normally-closed of course.
SW3 is a presence switch.

I hope it's a bit clear.

Please do give your opinions!

 

[ruffrecords]

EQ in the NFB of a gain stage is fine but you cannot just place a regular standard EQ network between output and input.

The EF86 is essentially an inverting op amp with the + input tied to ground. Check out basic op amp theory for ways to use this configuration for EQ.

Cheers

Ian

 

[C.B. - Boudio]

I know I could have used something like a Baxandall, but that doesn't seem to close the loop at DC, which I did try to achieve in my design, which is one of the reasons why it looks the way it does.

 

[ricardo]

difficult to make firm recommendations but my guess would be cost would be your number one consideration. In which case you have perhaps three choices of supplier: OEP (UK) , Edcor (USA) and UTM (Poland). assuming your prefer an EU source the you could try the UTM3510 for the output mand the UTM2555 for the input - but I have not used either of these myself

Ian, I'm sad you haven't mentioned Sowters who used to be near you.

Last time I heard, they'd been taken over by rif raf from Huntingdon. Has Ipswich really run out of Virgins, Unobtainium and solid BS?

 

[MidnightArrakis]

This is my first try of using KiCad. I wouldn't do it for fun like hand-drawing them, but it does have some usefulness.

I hope it's a bit clear.

Please do give your opinions!

[Please do give your opinions!] -- While I certainly do realize that my aging eyes are nowhere nearly as sharp as they once were during my much younger days, nonetheless.....the manner in which how I see things are connected-up here looks to me as though XLR Pins-2 and 3 are connected together!!! Am I not seeing things correctly? Are you creating an "unbalanced" circuit???.....




>> NOTE: A problem with the KiCAD XLR schematic symbols is.....they don't show any difference between "Female" and "Male" connectors, such as say.....OrCAD does.


/

 

[ruffrecords]

I know I could have used something like a Baxandall, but that doesn't seem to close the loop at DC, which I did try to achieve in my design, which is one of the reasons why it looks the way it does.

It is nigh on impossible to close the loop at dc using a single tube due to the large voltage difference between the output and input terminal

Your design closes the loop at dc via R15 direct from anode to grid which is liable to turn the tube hard on.

Cheers

Ian

 

[ruffrecords]

Ian, I'm sad you haven't mentioned Sowters who used to be near you.

Last time I heard, they'd been taken over by rif raf from Huntingdon. Has Ipswich really run out of Virgins, Unobtainium and solid BS?

Well I did preface my reply by stating I guessed the OP was looking for inexpensive transformers otherwise I would have mentioned Sowter, Jensen and Cinemag. Andd before you ask, I did not mention Lundahl simply because I have no experience of them.


Cheers

Ian