Feeler: EZ Tube Lunch Box

[ruffrecords]

Regarding gain-staging and 6db compensation:

If I understand things correct,
all eq- and comp-modules needs +6db compensation possible for the lineoutXfm -6db.

True if they use a 2:1 output transformer. The new stand alone EQ will use a 1:1 transformer but of course it will only drive a 10K load. You could use a 1:1 transformer on the modules that normally use a 2:! transformer as long as you only drive 10K loads.

That means non-true bypass and getting the 6db from makupgain-stage.

Not sure what you mean by non-true bypass.

In my case, going micXfm > pre > eq > comp > outXfm
the eq could be true bypass, and the comp non-true (with bypass utilizing +6db in the last gain-stage).
That is for the units to operate in 0db.

You could switch the transformer primary from the comp output to the EQ output to bypass the comp. Here's a link to a channel block diagram I did for Holger a long time ago that uses 600:600 output transformers and has a switchable insert and selectable direct out.

http://www.ianbell.ukfsn.org/EzTubeMixer/docs/EzTubeMixer/HolgersChannel2.jpeg

Bottom line is the lunch box just provides a box into which you can fit a number of building blocks. There are all sorts of ways they could be connected together.

Cheers

Ian

 

[G-Sun]

Thanks Ian!

True bypass: Hard bypass is the correct term here I guess, signal in being switched directly to signal out.

So, I'll need a 600:600/1:1 line out transformer for my eq/comp then?
Something like Edcore WSM600/600 out and WSM10K/10K for possible line in?
(or better)

You could use a 1:1 transformer on the modules that normally use a 2:! transformer as long as you only drive 10K loads.

I'm not sure if I understand load in/out good enough. 10k is ok, but 5k is not? Is that correct?
So, I cannot take the output of an ezComp with an 600:600 and split it to two devices?

 

[ruffrecords]

Thanks Ian!

True bypass: Hard bypass is the correct term here I guess, signal in being switched directly to signal out.

OK, understood.

So, I'll need a 600:600/1:1 line out transformer for my eq/comp then?
Something like Edcore WSM600/600 out and WSM10K/10K for possible line in?
(or better)

Definitely something better. Some of the Edcor WSM and XSM output transformers have rather low primary inductances. This is fine if you drive them from a very low source impedance like an op amp but they will lose bass when driven from a tube amp with a rather higher source impedance. You should preferably use an output transformer with a higher inductance like the Carnhill VTB2281. For the 10K:10K input transformer you might be better off with something like the OEP Z3003.

You could use a 1:1 transformer on the modules that normally use a 2:! transformer as long as you only drive 10K loads.

I'm not sure if I understand load in/out good enough. 10k is ok, but 5k is not? Is that correct?
So, I cannot take the output of an ezComp with an 600:600 and split it to two devices?
[/quote]

The regular EZTube amplifiers will happily drive 2400 ohms. That is exactly why I normally use a 2K4:600 transformer to drive 600 ohm loads. However, not many people use actual 600 ohms loads these days. They are mostly 10K bridging. Since the amplifiers will drive 2400 ohms you can safely drive four 10K loads at once via a 600:600 transformer.

Cheers

Ian

 

[G-Sun]

So, I'll need a 600:600/1:1 line out transformer for my eq/comp then?
Something like Edcore WSM600/600 out and WSM10K/10K for possible line in?
(or better)

Definitely something better. Some of the Edcor WSM and XSM output transformers have rather low primary inductances. This is fine if you drive them from a very low source impedance like an op amp but they will lose bass when driven from a tube amp with a rather higher source impedance. You should preferably use an output transformer with a higher inductance like the Carnhill VTB2281. For the 10K:10K input transformer you might be better off with something like the OEP Z3003.
..

The regular EZTube amplifiers will happily drive 2400 ohms. That is exactly why I normally use a 2K4:600 transformer to drive 600 ohm loads. However, not many people use actual 600 ohms loads these days. They are mostly 10K bridging. Since the amplifiers will drive 2400 ohms you can safely drive four 10K loads at once via a 600:600 transformer.

Thanks!
Ok, Carnhill VTB2281
http://www.audiomaintenance.com/acatalog/CA-18-VTB2281_extended_info.html
seems budget-friendly. Good sounding? Will you provide pcb-mount for it?

Ok, then I better understand the load-question. Guess 600:600 will be perfectly fine for most users I belive. And it will then give better gain-flow for the eq/comp-modules, right?

 

[ruffrecords]

Thanks!
Ok, Carnhill VTB2281
http://www.audiomaintenance.com/acatalog/CA-18-VTB2281_extended_info.html
seems budget-friendly. Good sounding? Will you provide pcb-mount for it?

Very good sounding (one of the Carnhill factories used to make transformers for Neve).

It is a BIG transformer - no way you would PCB mount it!!! In the lunch box is would be bolted to the rear panel next to the input and output XLRs and then wired to the backplane. Here is a sketch I drew a few years back showing the basic idea:

http://www.ianbell.ukfsn.org/EzTubeMixer/docs/EzTubeMixer/Mechanics/MIcPreMechAssyShortscaled.jpeg

Cheers

Ian

Ok, then I better understand the load-question. Guess 600:600 will be perfectly fine for most users I belive. And it will then give better gain-flow for the eq/comp-modules, right?
[/quote]

 

[G-Sun]

Very good sounding (one of the Carnhill factories used to make transformers for Neve).

It is a BIG transformer - no way you would PCB mount it!!! In the lunch box is would be bolted to the rear panel next to the input and output XLRs and then wired to the backplane. Here is a sketch I drew a few years back showing the basic idea:

http://www.ianbell.ukfsn.org/EzTubeMixer/docs/EzTubeMixer/Mechanics/MIcPreMechAssyShortscaled.jpeg

Ok, I see. Well, wired to the backpane is fine (I thought the pins were for pcb-mount, but that's not the case then). Looks like my outXfm

Can I use the CA-18-VTB9045 for mic-inputXfm ?
http://www.audiomaintenance.com/acatalog/CA-18-VTB9045_extended_info.html
(less vendors/shops are good for keeping costs down.)
Or VTB 9045M
http://www.audiomaintenance.com/acatalog/CA-18-VTB9045M_extended_info.html
?
(Depends a little on the mounting options I guess. No big deal using the Sowter 1948)

 

[ruffrecords]

Ok, I see. Well, wired to the backpane is fine (I thought the pins were for pcb-mount, but that's not the case then). Looks like my outXfm

It has always been that way from the beginning of the EZTube Mixer development. WHen you get one of these transformers in your hands you will immediately realise why it is not PCB mounted.

Can I use the CA-18-VTB9045 for mic-inputXfm ?
http://www.audiomaintenance.com/acatalog/CA-18-VTB9045_extended_info.html
(less vendors/shops are good for keeping costs down.)
Or VTB 9045M
http://www.audiomaintenance.com/acatalog/CA-18-VTB9045M_extended_info.html
?
(Depends a little on the mounting options I guess. No big deal using the Sowter 1948)

Both these transformers have a 1:2 ratio. For my designs you need a 1:10 ratio transformer so unfortunately these transformers are not suitable.

Cheers

ian

 

[G-Sun]

It has always been that way from the beginning of the EZTube Mixer development. WHen you get one of these transformers in your hands you will immediately realise why it is not PCB mounted.

Understand

Both these transformers have a 1:2 ratio. For my designs you need a 1:10 ratio transformer so unfortunately these transformers are not suitable.

Ok, thanks!

 

[G-Sun]

I'm still not sure what to do with the second gain-stage of the classic or the eurocard.

I've chosen the Carnhill VTB2281 600:600/1:1 as my output line transformer,
and this gives me
Classic: +50db(one stage) / +76db (two stages)
Eurocard: +60db / +90db??

Now, +55db clean gain is usually max of my needs. For the odd occasion I have a Fethead.
More clean gain is not a problem, but I need to manage high level mic input and 0db line level.
Is that a problem with any of the configurations?

To, me it seems ideal to be able to switch the second classic gain stage
between configured as
- mu follower (for a high gain fairly clean sound)
and
- as SRPPs (for a lower gain rather dirtier sound, as a tube distortion unit)
perhaps with a in/out (bypassing the gain stage)
and a pot for adjusting the gain.
The first gain-stage would need a little more control than just a pad,
maybe a pot, or a 5-way switch?

Just, thinking laud here, not knowing the practical side of things

 

[ruffrecords]

The second stage of the Eurocard was always intended to be a gain make up for a passive EQ. It has enough gain available that it can also add 10dB of gain to the channel fader and make up for the 6dB loss if you use a 2K4:600 output transfrormer. So, the normal signal flow would be:

1. Mic/Line pre to OUT1 (up to 60dB gain)
2. OUT1 (via optional 600:600 and 10K:10K insert/direct out) to 10K channel fader
3. Channel fader to passive EQ in
4. Passive EQ out to IN2 (gain set to make up EQ insertion loss plus possibly 10dB in fader plus transformer loss)
5. OUT2 to output transformer (2K4:600 or 600:600)

This is a basic EZMixer channel with up to 70dB gain from mic input. Hi level mic inputs are not a problem. The minimum gain  from mic input to OUT is just 7dB with the 20dB pad switched in.

0dB line level inputs are not a problem. The mic/line switch, in the line position, feeds the line input via a 33dB pad to the mic transformer, so if you set the mic gain switch to 33dB the gain is unity.

The Classic board contains two identical fixed gain amplifiers. Configured as mu followers the gain is close to 28dB. Configured as SRPPs the gain is a bit less at around 26dB. This means with a 20dB transformer, the first stage can have a gain of up to 48dB with a 600:600 output transformer. If this is sufficient gain then you can build two mic pres with one board. For a gain control you would nee dto insert a  100K or 220K pot between the transformer secondary and IN1 or IN2. This is a high impedance unbalanced connection so you would definitely need to use screened cable and keep it as short as possible.

If 48dB is not enough gain then you can use a level pot after the first stage and feed it to the second stage giving up to 75dB gain. Swtching the second stage between mu and SRPP is not easy. To do it properly you need to change both cathode resistors to obtain the right dc condictions and short out the 4K7 in the middle. Switching dc will certainly cause loud audible clicks. A simple alternative is to selectively bypass the 4K7 resistor with a cap and resistor (switched or a pot). This idea was discussed hear a few years ago. I do not know if anyone ever tried it.

I think a 20dB pad will be sufficient for most mic sources. With the pad in circuit the mu follower has 28dB of gain. It will certainly output +28dB into a 10K load (but with a bit of distortion) but you would need to input 0dBu to make it do that. for a 0dBu output the input would need to be -28dBu. If you regularly have sources greater than this level then you should condier designing a switched pad with maybe 20 ad 30dB positions.

Cheers

Ian

 

[ruffrecords]

New version of the 6 module motherboard arrived yesterday. These are ideal for Lunch Box builders intending to use an external power supply. I will build one to make sure they are right and then offer them in the White Market.

Also, today the new batch of V3 Classic boards arrived.

Cheers

Ian

 

[G-Sun]

The second stage of the Eurocard was always intended to be a gain make up for a passive EQ. It has enough gain available that it can also add 10dB of gain to the channel fader and make up for the 6dB loss if you use a 2K4:600 output transfrormer. So, the normal signal flow would be:

1. Mic/Line pre to OUT1 (up to 60dB gain)
2. OUT1 (via optional 600:600 and 10K:10K insert/direct out) to 10K channel fader
3. Channel fader to passive EQ in
4. Passive EQ out to IN2 (gain set to make up EQ insertion loss plus possibly 10dB in fader plus transformer loss)
5. OUT2 to output transformer (2K4:600 or 600:600)

This is a basic EZMixer channel with up to 70dB gain from mic input. Hi level mic inputs are not a problem. The minimum gain  from mic input to OUT is just 7dB with the 20dB pad switched in.

Thanks! Yes, Ian,
I've seemed to mixed up general design thoughts with my personal build.
The second gain stage is quite usabel for me as makup for the eq and I don't need the extra gain-stage on the eq pcb.

Maybe I should think putting in a simplified channel fader for the eurocard preamp?
So, a SPTT switch with -10db, 0db, +10db maybe.
It's space for that, right?
Or would there be little benefit in a -10db setting?

I'll maybe have to change a resistor (eq makup) if building the preamp first, then adding eq/comp later,
but that's not a biggie.

Is the signal flow described above different for the classic?

 

[G-Sun]

But then,
how should an eq bypass-switch be implemented?

Will you provide bypass-switch for the Helios, ReddEq and Pulteq 3band?
Is bypass for those modules only possible when utilizing the on-board gain-stage?

 

[ruffrecords]

Thanks! Yes, Ian,
I've seemed to mixed up general design thoughts with my personal build.
The second gain stage is quite usable for me as make-up for the eq and I don't need the extra gain-stage on the eq pcb.

THat's right. Here's a sketch I drew a few years ago showing a typical EZMixer signal flow :

http://www.ianbell.ukfsn.org/EzTubeMixer/docs/EzTubeMixer/EZtubeMixerSigFlowsm.jpeg

Maybe I should think putting in a simplified channel fader for the eurocard preamp?
So, a SPTT switch with -10db, 0db, +10db maybe.
It's space for that, right?
Or would there be little benefit in a -10db setting?

I am sure you can fond room on the front panel for a switch. The normal use of a fader is to fade or make small adjustments to level. You can vary the mic pre gain in 3dB steps so I am not sure of the value of 10dB steps. SOme preamps have a level trim pot with a 10dB or so range. That might work. It's up to you.


I'll maybe have to change a resistor (eq makup) if building the preamp first, then adding eq/comp later,
but that's not a biggie.

Is the signal flow described above different for the classic?
[/quote]

Yes, the classic does not have as much available gain. The EZMixer has 60dB mic gain plus as much as another 40dB from the second amp so there's a total of 100dB available.  With that amount of gain  you can afford to lose 20dB insertion loss in a passive EQ and still have 70dB gain overall. The Classic has maximum of 76dB  using both stages. If you lose 20dB in a passive EQ you overall gain drops to 56dB. For many applications that will be fine. If you limit the EQ to cut only (as in the HPF) then there is no insertion loss so you get the full 76dB gain. Swings and roundabouts.

Cheers

Ian

 

[ruffrecords]

But then,
how should an eq bypass-switch be implemented?

Will you provide bypass-switch for the Helios, ReddEq and Pulteq 3band?
Is bypass for those modules only possible when utilizing the on-board gain-stage?

The normal way to implement bypass is to switch in a pad which attenuates the same amount as the insertion loss of the EQ. Here is a diagram that shows how it is done in the poor man's EQP1A:

http://www.ianbell.ukfsn.org/EzTubeMixer/docs/PMEQP1A/EQSystem.png

The principle is the same for any EQ. The pad is built into the PCB on the 3U REDDEQ board and the original Pultec 3 band PCB. The Helios is slightly different in that the pad is built in to the original circuit but the principle is the same. In all cases you need to wire up a suitable switch and fit it to your front panel.

Just to put this into context, the EZTubeMixer boards are basic building blocks that you can build into almost anything you like - a bit like Lego.

Cheers

Ian

 

[G-Sun]

The normal way to implement bypass is to switch in a pad which attenuates the same amount as the insertion loss of the EQ. Here is a diagram that shows how it is done in the poor man's EQP1A:

http://www.ianbell.ukfsn.org/EzTubeMixer/docs/PMEQP1A/EQSystem.png

Thanks,
so it's a DPDT switch?
The pad-section I don't understand well. 10k fixed, 1k pot (trim for adjustment?) and a 470ohm fixed to ground?

Anyway, starting to get the concept

 

[G-Sun]

THat's right. Here's a sketch I drew a few years ago showing a typical EZMixer signal flow :

http://www.ianbell.ukfsn.org/EzTubeMixer/docs/EzTubeMixer/EZtubeMixerSigFlowsm.jpeg
...

I am sure you can fond room on the front panel for a switch. The normal use of a fader is to fade or make small adjustments to level. You can vary the mic pre gain in 3dB steps so I am not sure of the value of 10dB steps. SOme preamps have a level trim pot with a 10dB or so range. That might work. It's up to you.

...
The EZMixer has 60dB mic gain plus as much as another 40dB from the second amp so there's a total of 100dB available.  With that amount of gain  you can afford to lose 20dB insertion loss in a passive EQ and still have 70dB gain overall. The Classic has maximum of 76dB  using both stages. If you lose 20dB in a passive EQ you overall gain drops to 56dB. For many applications that will be fine. If you limit the EQ to cut only (as in the HPF) then there is no insertion loss so you get the full 76dB gain. Swings and roundabouts.

Thanks!
Nice diagram.
So, the pad-switch is for mic only? Good to know

So, I have 100db gain with the eurocard (20+40+40)
The eq is -20db (3 band pulteq boost/cut)
I'll go for +70db for my preamp.
My out transformer will be 0db.
That means setting a fixed extra resistor for -10db?
Then I need a good value for line in pad
And perhaps an extra pad somewhere, labeled 0db/+20db?
And/or a fine-gain (pot) -10/-5db to +10/+5db?

If the compressor will have input gain-pot,
the 3db steps for the preamp is not a problem.
Easy setting line in to 0db is important,
as is handling  of normal/high-level mic input.

 

[ruffrecords]

Thanks!
Nice diagram.
So, the pad-switch is for mic only? Good to know

Correct

So, I have 100db gain with the eurocard (20+40+40)
The eq is -20db (3 band pulteq boost/cut)
I'll go for +70db for my preamp.
My out transformer will be 0db.
That means setting a fixed extra resistor for -10db?

You use the pre-set pot on the PCB for the second amp to set its overall gain

Then I need a good value for line in pad

The line in pad is set at 33dB. The 'line' ident' on the gain switch is at the 33dB position so this gives unity gain for the line input.

And perhaps an extra pad somewhere, labeled 0db/+20db?

Not sure why you would need this

And/or a fine-gain (pot) -10/-5db to +10/+5db?

this would sit between the two amps like a narrow range level control

If the compressor will have input gain-pot,
the 3db steps for the preamp is not a problem.

And it is not a problem anyway as you have the fine gain control.

Easy setting line in to 0db is important,

See above

as is handling  of normal/high-level mic input.

The reason for the stepped gain control is to allow this. The minimum gain of the amp is 7dB plus the 20dB from the mic transformer so a -27dBu input would be raised to 0dBu. If you mic signal is higher than this then you use the 20dB mic pad so now an input of -7dBu is raised to 0dBu. Since the mic pre max output is +30dBu, you could in theory have a mic signal level of as much as +23dBu before the amp clips. I suspect the transformer would saturate before that.

Cheers

Ian

 

[ruffrecords]

I finally got round to starting the assembly of the rear panel. I do like those Neutrik XLRs with the threaded holes -so much easier to assemble and the rear mounting version look so much better. I made a minor error - I got the line in XLRs a little to close to the MIC ones - had to file a bit off to make them fit. Also, the fitting of the XLRs would look neater if the holes were countersunk. Need to update the front panel layout. Thanks again to Frank Rollen for making/engraving  the panel.

Cheers

Ian

 

[rmaier]

Almost done my PSU. Question: where can I pull power for the relays for four Classic preamps?
Edit: could I simply use an off-the-shelf 12vdc supply?