8 Channel, 3U NEVE 1073 completed

[boomfred]

Are you writing about that question ?

As you have an external PSU you should have a separate conductor for 0V and another one for Chassis , from the external PSU into the mic amp.
Do you have this way?

Because I answered that, which I thought did respond yes :

"Then I have a 4 wire cable running to the rack case (case, 0v, 24v, 48v)."



Let me try to rewrite my question.

0v and case are separated in PSU (with the 10 ohm resistor and the capacitor dedicated).
0v and audio ground are connected on the board.
If I connect pin 1 of the XLR Input (audio ground) to the case, I will therefore make a connection between case and 0v (through the direct connection on the input transformer)

Or did I miss something ?

 

[Whoops]

Do you have the PSU PCB on plastic or Metal standoffs?

You didn't reply to this...

 

[Whoops]

0v and case are separated in PSU (with the 10 ohm resistor and the capacitor dedicated).
0v and audio ground are connected on the board.
If I connect pin 1 of the XLR Input (audio ground) to the case, I will therefore make a connection between case and 0v (through the direct connection on the input transformer)

Or did I miss something ?

XLR Pin1 is not Audio Ground (0V), PIN1 is "Shield", a case is "Shield" also. All Shields connect directly to "Mains Earth"
Treat Pin1 has an extension of the Case

I know the terms can be misleading, because incorrect terms are used everywhere, even in PCB silkscreening and in build manuals. "Audio Ground" is misleading, better would be to call it just "0V reference"
"Ground" can also be misleading because people use it for "0V reference" and for "Mains Earth" sometimes

Please read this document as this is really well explained:
https://www.ranecommercial.com/legacy/note165.html


important thing separate everything which is 0V reference and Mains Earth.
Separate 0V reference and all "Shields" that should be connected to case and then "Mains Earth" directely
0V reference and Mains Earth only connect at one point and one point only, through a 10r resistor in parallel with a 100nf Cap.
That happens in the JLM PSU pcb, 0V is feed to a 10r + 100nf cap, then there's a trace that goes into one of the PCB holes, and using a metal stand off it connects to Case, thats how it's connected.
Then PSU case should be connected to the IEC Earth Pin

On the Preamps rack case end, don't connect any 0V reference to case keep them separated as they are already connected in the JLM PSU

Hope this clears it out

 

[boomfred]

Sorry, I thought I respond but probably rewrited my answer so it kind disappeared.

I use the stand-off provided with the kit, and they’re metal ones.

Thanks for your explanation.
Although, using the guide I did connect XLR 1 to « shield » on the board. And that « shield » is connected to 0v on the board.

 

[Whoops]

Sorry, I thought I respond but probably rewrited my answer so it kind disappeared.

I use the stand-off provided with the kit, and they’re metal ones.

Thanks for your explanation.
Although, using the guide I did connect XLR 1 to « shield » on the board. And that « shield » is connected to 0v on the board.

I know what you mean, I dont agree with that connection on the guide.
Pin6 is the transformer internal shield
I would remove the connection between Pin6 and Pin10.
Connect XLR Pin1 to Pin6 on the transformer, but dont connect Pin6 to Pin10

 

[boomfred]

Well I can try that : grounding inputs to the case, « as usual » and leave pin 10 of the transformer to the board (and 0v).

 

[Whoops]

Well I can try that : grounding inputs to the case, « as usual » and leave pin 10 of the transformer to the board (and 0v).

Yes,
That connection between pin6 and pin10 forms a ground loop

 

[boomfred]

Wasn’t that.
I keep on searching.

 

[ruffrecords]

Yep, you’re right. I fixed that. It’s way better, but I still have a pretty high noise floor (around -36dB with the switch all the way up).

Whether that is bad or not depends on what you mean by dB and exactly how much gain there is and if there is anything plugged into the input XLR.

The first test you should do is plug in an XLR with pins 2 and 3 shorted together and crank the gain up. A 1073 has 80dB gain at maximum gain. Its EIN is about -126dBu which means at 80dB gain the noise at the output will be -46dBu. Laws of physics dictate it will not get much better than that.

Cheers

Ian

 

[boomfred]

Oh, you got a point there, sir.

Turns out I did run some tests today, with a ribbon mic and a static one, with the switch and all the way up,  and I manage to have a noise floor around 40dB (according to my DAW).
I did some comparison with more modern stuff (Audient ASP, noise floor around 60dB) but I didn’t consider the amount of gain available.

 

[ruffrecords]

Oh, you got a point there, sir.

Turns out I did run some tests today, with a ribbon mic and a static one, with the switch and all the way up,  and I manage to have a noise floor around 40dB (according to my DAW).
I did some comparison with more modern stuff (Audient ASP, noise floor around 60dB) but I didn’t consider the amount of gain available.

The big problem for everyone using a DAW with an external pre is that you often have no idea how dBFS inside the box relates to dBu in the real world outside the box. Interfaces are all over the place in this respect, often poorly specified and quite probably with hardware and or software adjustment of input sensitivity. The only way to get round this is to feed a known 0dBU signal into an interface so you can calibrate it.

Seems to me this might be useful gadget for DAW owners - maybe a little groupDIY project?

Cheers

ian

 

[boomfred]

Here’s, finally, mine. Perfectly working.
Thanks everyone here for the support.

(I finished a D-U87 too, it was waiting for its tranny).

 

[Bakula]

I was hoping to not have to post a question here any time soon, but 2 of the EZ1290 boards I’ve got have a questionable pad on them and I want to make sure I’ve got this right before I do anything with it.

The transistor side of the 330pf poly cap looks like the solder mask just filled over the pads. Seems obvious one side of a cap wouldn’t be unsoldered, but since it’s the same on more than one board I figured I should make sure it’s just a mistake and not some revision I’m not aware of. The other boards I have already populated don’t have the same issue. All versions I have are v2.6, but from different sources.

I’ve made it through about 80 pages of this thread and don’t remember reading anything about it, but it’s a lot to take in so I may have missed it.

If it’s just a mistake on the board I’ll scratch it out and move on. Pictures attached.

Thanks in advance.

 

[Bakula]

Other side:

 

[Whoops]

It seems a production mistake, I would just scratch the green paint and expose the copper so you can solder there.

You should let Martin know that you received those 2 pcb's with that defect.

 

[Bakula]

Thanks Whoops, apologies I didn’t see your reply until just now.

You’re right it was a defect, which I assumed but since it was the same on 2 boards I wanted to confirm.

I have a few boards from different sources, so I contacted the shops and they contacted Martin.

So anyone purchasing boards for this currently, be aware and scratch off the pad. Easy fix.

 

[Clakejr]

How do you go about securing both of  the transformers to a 2u case? Is there an adapter or certain bolts used?

 

[Whoops]

How do you go about securing both of  the transformers to a 2u case? Is there an adapter or certain bolts used?

You can easilly do your own adapter if needed.

Both the input and the output transformer have placements for bolts that you can attached to the metal case

 

[Bakula]

After finally completing a channel, I've got some off transistor values and the preamp sound is very quiet, thin, and crackly.

On first power on, I tested only the heat sink on the 3055 and then the chassis ground fell out and I heard oscillation from the output trafo itself. After quickly powering down and securing the chassis connection, oscillation stopped.

Transistors 4-6 values seem ok, but a few values on 1-3 are consistently off. I've checked all values, and replaced and secured just about everything in that section of the board. I've also swapped it out for another completed board with similar results. Most parts are off the BOM or matched value/quality. A small handful of electrolytics are just random off brands I had laying around but have used in other audio builds with no issue.

Transistor values (C, B, E):

TR1  4.9    .93  .34
TR2  22.5    4.9  4.3
TR3  22.5    4.3    3.6
TR4    3.88  2.44  1.85
TR5  12.68  3.8    3.2
TR6  21    12.68  12

I'm not sure what else to check without just flat out guessing. I've read the thread through a couple of times but it's a lot to remember. I searched for similar issues but haven't found a solution. Have I blown a trafo? I wouldn't think a couple of seconds of oscillation would do that.

Any guidance would be appreciated. Thanks!

Update/Edit: I realised the jumper was missing in the picture I attached (but was not missing on the other boards I checked), but essentially same results once I soldered it in. I've checked 2 other boards and swapped the 9049 and still same results. The transistor values are a bit different but still the same pins are off and the same amount on each board. I'm stuck.

 

[Bakula]

Picture of board and trafos for reference.