Tube DI Box

[IBarfoot]

Gang,
Ok first I think the switch for input impedance is going to be changed to input PAD.
This works well, SNR ~-120dB, frequency response is pretty much flat actually stupidly flat to below 10Hz and above 22Khz. I can't even get the damn thing to go below -3dB on the low end and well the dScope III can do it but it's way out on the top end as well. I will have to use a meter and do this with my function generator but it looked to be like -3dB above 80Khz.
Time to box it up and see how my crappy mechanical skills are
G.

Very sweet looking indeed, nice work.

 

[MidnightArrakis]

Nope, just a weird search engine that does not return the 2U unit when you feed it "Galaxy Maggiorato GX288 230 x 280 mm"

[just a weird search engine that does not return the 2U unit when you feed it "Galaxy Maggiorato GX288 230 x 280 mm"] -- Try using the search engine called --- startpage.com --- for your searches. I have found it to work extremely well and it also DOES NOT track your search usage or history. While it does use a basic GOOGLE search engine, it is nonetheless, still an independent search engine. I did a "Copy & Paste" of your search term -- Galaxy Maggiorato GX288 230 x 280 mm -- and got this as an immediate response:



If you are also still using the -- Vivaldi -- browser I had once suggested in another thread you had made a response back to me several months ago, I have since learned that the -- Vivaldi -- browser has been rated as one of the fastest and most thorough browsers in use on the Internet. I really like how it operates, how it organizes its tabs, displays items on my monitor and much more!!!

GOOD LUCK!!!

/

 

[Wavelength]

Ok the plan:
1) I am going vertical with the input and thru. Amphenol has a nice MONO one at Mouser.
2) dump the input Z and the output Z. The output Z effected the response of the Jensen output, but not the CineMag.
3) Move the XLR Right, keep the pin 1 ground, most cases it does nothing but just in case.
4) With the picked up area between the vertical TS and the XLR put in a Grayhill probably 12 stepped input level. Thoughts on the curve at 1M 1,1,1,2,2,2,3,3,3,6,6,6dB? I have tried a ton of 1M volume controls and they all roll off the top too much. Just had an argument with Bourns engineers when I used one in a 500 series module I am working on.
Thoughts?
Thanks,
Gordon

 

[Wavelength]

Ota, someone want to check my math here?
Step 1, Attenuation = 22 dB, Rx = 920567 ohms, Ry = 79433 ohms, Resistor = 79433 ohms. 82K5 I set this so Zin = 1M
Step 2, Attenuation = 20 dB, Rx = 900000 ohms, Ry = 100000 ohms, Resistor = 20567 ohms. 20K5
Step 3, Attenuation = 18 dB, Rx = 874107 ohms, Ry = 125893 ohms, Resistor = 25893 ohms. 26K1
Step 4, Attenuation = 16 dB, Rx = 841511 ohms, Ry = 158489 ohms, Resistor = 32596 ohms. 32K4
Step 5, Attenuation = 14 dB, Rx = 800474 ohms, Ry = 199526 ohms, Resistor = 41037 ohms. 41K2
Step 6, Attenuation = 12 dB, Rx = 748811 ohms, Ry = 251189 ohms, Resistor = 51663 ohms. 51K1
Step 7, Attenuation = 10 dB, Rx = 683772 ohms, Ry = 316228 ohms, Resistor = 65039 ohms. 64K9
Step 8, Attenuation = 8 dB, Rx = 601893 ohms, Ry = 398107 ohms, Resistor = 81879 ohms. 82K
Step 9, Attenuation = 6 dB, Rx = 498813 ohms, Ry = 501187 ohms, Resistor = 103080 ohms. 102K
Step 10, Attenuation = 4 dB, Rx = 369043 ohms, Ry = 630957 ohms, Resistor = 129770 ohms.130K
Step 11, Attenuation = 2 dB, Rx = 205672 ohms, Ry = 794328 ohms, Resistor = 163371 ohms.162K
Step 12, Attenuation = 0 dB, Rx = 0 ohms, Ry = 1000000 ohms, Resistor = 205672 ohms. 205K
Note: My Grayhill footprint works with both a 6x2 and 12x1, so pine 14 is not used here that is for 6x2. Thanks Gordon

 

[ruffrecords]

The values look fine to me. Couple of small points. To avoid the possibility temporary disconnection of the grid when switching I would be tempted to add a permanent 2M2 from grid to ground. This will slightly alter the calculated resistor values.

I would suggest that there is little need to opt for expensive E48 series resistor values. The nearest E24 values will be fine. The small deviations from exactly 2dB steps will be undetectable and exact 2dB steps are not really necessary.

Cheers

Ian

 

[Wavelength]

Ian,
Thanks will do on the shunt to ground, actually thought of that myself.
I was going to put resistor values on the PCB, then I figured everyone would probably do their own thing anyway. I just search Digikey for 1-5% in companies and types I thought were acceptable and matched from there.
I also added an LED in the back to indicate power. The front IEC has one of those switch on led active.
Thanks,
Gordon

 

[Wavelength]

New Back Panel:

 

[ruffrecords]

Cube Tube - nice.

Cheers

Ian

 

[MidnightArrakis]

The only thing I am not sure about is whether the 6in by 6in by 4in version will be big enough for both a mains transformer and a Carhill output transformer and the tube PCB and the connectors and controls. I will probably make one out of cardboard and try real parts in it. I could do it in 3D CAD but I don't have enough lifetime left to get up the learning curve

Cheers

Ian

[I am not sure about is whether the 6in by 6in by 4in version will be big enough for both a mains transformer and a Carhill output transformer and the tube PCB and the connectors and controls] -- If you can at least provide me with links to the various datasheets of all of the major components used within this project, I could create a 3D CAD-model -- mockup -- of how things may possibly fit within your enclosure of choosing. And/or.....I could do the same exercise with one of the enclosure vendors I am personally familiar with who also provide CAD-models of their enclosures as well.

In a nearly identical scenario such as this.....I am currently working on a project with a large U.S. defense contractor where I am having to determine what the size of a PCB will be by importing 3D CAD-models of all of the connectors, relays, switches, test points, mounting hardware, etc. onto a mock 3D PCB first and then also doing a components placement within a mechanical design program. In addition and to also add to the "FUN" of this exercise, two (2) of the D-Sub connectors need to be placed on opposite sides of the PCB at an opposing 25-degree angle. This is so the mating cable assemblies will be able to plug straight-in to an associated piece of equipment and the cables themselves won't need to be bent as the PCB itself is going to be rotating at a rate of 240-RPMs!!!

My doing all of this is just to determine how large the PCB will need to be in order to easily accommodate everything as that is an unknown detail as of right now. As I have been working on this project, I feel like I am "designing a PCB -- BEFORE -- it is actually designed"!!! So, this is the first time that I am designing a PCB by using a mechanical design program before using a PCB-design program!!! OH!!! MY!!! How times have changed!!!.....

In any case.....send me whatever you already have and I can create a mockup for you, OK???

/

 

[Wavelength]

Cube Tube - nice.

Cheers

Ian

Ian,
Thanks, question that has bothered me for years. Output transformers like to be grounded. One designer said... oh you need to ground the center tap of the secondary for the best results. Then you can offer XLR Pin 1 = ground, open or some cap/resistor combo.

The early spec said that the differential pair should not have any link to ground.

Your thoughts?

~~~

The hammond is 4x4x6 inches, they have a 4x4x8 as well. The Carnhill is a bit large. I do have one of these new switching power supplies in. Specs are 6.3@4A, 200V or 270V @ 140ma. Probably a little overkill for a single tube but they are pretty small (4x2.25x1.25) and somewhat shielded.

 

[merlin]

Output transformers like to be grounded.

 

[ruffrecords]

[I am not sure about is whether the 6in by 6in by 4in version will be big enough for both a mains transformer and a Carhill output transformer and the tube PCB and the connectors and controls] -- If you can at least provide me with links to the various datasheets of all of the major components used within this project, I could create a 3D CAD-model -- mockup -- of how things may possibly fit within your enclosure of choosing. And/or.....I could do the same exercise with one of the enclosure vendors I am personally familiar with who also provide CAD-models of their enclosures as well.

In a nearly identical scenario such as this.....I am currently working on a project with a large U.S. defense contractor where I am having to determine what the size of a PCB will be by importing 3D CAD-models of all of the connectors, relays, switches, test points, mounting hardware, etc. onto a mock 3D PCB first and then also doing a components placement within a mechanical design program.

In any case.....send me whatever you already have and I can create a mockup for you, OK???

/

I will do that. Many thanks for offering to do this.

Cheers

ian

 

[ruffrecords]

Ian,
Thanks, question that has bothered me for years. Output transformers like to be grounded. One designer said... oh you need to ground the center tap of the secondary for the best results. Then you can offer XLR Pin 1 = ground, open or some cap/resistor combo.

The early spec said that the differential pair should not have any link to ground.

Your thoughts?

~~~

A balanced signal exists only between the hot and the cold pins. A balanced signal is NOT a pair of differential signals referenced to a third pin. The often to used third pin is there only connect a screen around the other two. It is not signal carrying connection. It is therefore unnecessary to make any connection the centre tap of an input or output transformer.

Cheers

Ian

 

[Wavelength]

A balanced signal exists only between the hot and the cold pins. A balanced signal is NOT a pair of differential signals referenced to a third pin. The often to used third pin is there only connect a screen around the other two. It is not signal carrying connection. It is therefore unnecessary to make any connection the centre tap of an input or output transformer.

Cheers

Ian

Ian,
Thanks same thing David said!
Gordon

 

[GussyLoveridge]

I did a project with one of these enclosures this year and I really like working with it. It felt robust and the metal work was not too arduous. It looks wonderful too. Maybe too much space for this project?

https://www.hammfg.com/electronics/small-case/general-purpose/518

 

[Wavelength]

In a nearly identical scenario such as this.....I am currently working on a project with a large U.S. defense contractor where I am having to determine what the size of a PCB will be by importing 3D CAD-models of all of the connectors, relays, switches, test points, mounting hardware, etc. onto a mock 3D PCB first and then also doing a components placement within a mechanical design program.

/

I had to do that for a Bluetooth Audio device I designed. The solid cadd guy sent me this crazy layout with a Toslink, USB, 3.5mm audio jack, 8 PCB holders and I had to place everything so it fit right.
Altium makes it easy to inport the DXF file but really trying to put a boat load of components and keep them to one side was a bit daunting.

 

[Wavelength]

Ok here is the proto done with the new panel that made the power transformer fit just right. I also got the PCB's and stuff in for the final version with the stepped attenuator and other small changes. Will make that up after I finish up 2x guitar amps for Barenaked Ladies.