AMS Neve 1073 Sweeps

[Aaronrash]

Hi everyone,

Sorry I have been super busy and haven't been on here much. My inbox on here is full with over 50 request now

for the etch files.

Here's the deal... They took me 2 weeks to draw by hand. Tons of work went into making them. I wish it would

have been easy or I could have just asked someone for them but I had to spend over 2 weeks making them

flawless. All the others floating around have errors.

After talking to my buddy who happens to be an attorney, he pretty much told me I'm asking for a lawsuit if I share these so I'm sorry but they will not be shared, for my own protection as well as my respect for AMS Neve

If they were my designs I wouldn't care but they do not belong to me therefor I have no right to post another companies designs or share them. If you want the original layouts you will have to draw them like I did.

It is pretty obvious they are exclusive to AMS Neve... If they were not im sure Herritage and BAE would be using the same layout.

Sorry and thanks for understanding.

The easiest way to pull them off is to find high res pictures on the net and warp them in Photoshop to make them flat and oriented correctly.

After that just start drawing. It takes a long time but that is the best way to do it without a real module sitting in front of you.

With that out the way, I have a little update about my builds. I had the output secondaries wired incorrectly! I discovered it actually sounds really cool but I went ahead and corrected the errors and they measure pretty much identical, if not a little better than my AMS 1084.

681Ohm termination, Sweeps included. They sound Sweeeeet!

 

[Aaronrash]

My build sweep

 

[bruno2000]

Thanks for your work and info!  I applaud your decision to release the etch files.
What level is the input signal, and what gain setting are you using?
Thanks!
Best,
Bruno2000

 

[Aaronrash]

The measurements are mic input sweeps in the "HI" (1200) position. Unity gain.  (20dB)

So anyone that builds these. This is the response you're gonna want to aim for.

Also when drilling be sure to wear a mask, not only for the board dust but for the copper dust too. Copper is extremely toxic so be careful and have fun!

I also want to stress to anyone considering building these that it is a ton of work. I had no idea at first (after building the EZ1073's) that this would be much different but it is on a completely different level.

I actually don't know how in the world BAE and AMS makes much a profit doing this. I have a new found respect for them after doing it myself.

 

[API]

Would you care to post a picture of your build?
I am doing something similar and wanted to check how you laid yours out.

Also, i have a full set of 1084 boards from an old groupbuy, i would suspect the files from those should be around somewhere.
Think it was a French guy named Vincent that made those.

 

[Aaronrash]

Hey everyone,

If anyone is interested, I'm selling my AMS 1084 $,3,050 USD

 

[Aaronrash]

I have some amazing news for EZ 1073 owners.

This week during repairing my broken unit I decided to tinker around a bit. I have come up with a modification that will get your EZ1073 to the spec of my builds and the AMS 1073/1073.

Now I can't hear a difference between all my units. So here's what I did

First things first, I am using AMS output transformers in my EZ units so I can't speak for standard carhills or PCB mount versions but this made my units sound tremendously better.

Step 1.
Forget wiring the input transformer on the PCB mount. Wire it up and mount it to the frame like the AMS modules. Also I discovered that the VTB9045 doesn't have internal continuity for the ground, the AMS TF10003 do so there's a difference for you right there. Make sure you wire directly to/off the impedance switch.

Step 2.
Forget wiring the output transformer to the pads on the board. Wire it like a real module. This means wire directly off C12 and C14 to the primary. Off the secondary wire it with shielded wire per Neve schematic directly to the phase switch. Go directly off the phase switch to the output XLR.

Step 3.
Get rid of the -10dB attenuation and make it a full functional fader like a VK Rack or a real console. This means put a jumper on R134 and replace R135 with a 10K. This turns that 10K fader into a 5K fader with -00dB- to 0dB (Full fader attenuation)

Step 4.
Strap a 681 on the XLR output, this is what AMS Neve does as well as vintage king.

If you do all these things properly you should get a frequency response like the on I attached. This is my EZ 1073. It sounds so much better now. It measures identically to my 1084

Hope it helps

 

[Aaronrash]

It's actually a bit scary how close these look now.

AMS1084 on the right, EZ1073 on the left. The distorion is the same on both too

 

[Aaronrash]

Here's the EZ Unit

 

[ding]

Thats very interesting. I am just finishing my ez1073 and I am curious about a few things you mentioned that I am confused about.

1. Can you expand on this…

Also I discovered that the VTB9045 doesn't have internal continuity for the ground, the AMS TF10003 do so there's a difference for you right there. Make sure you wire directly to/off the impedance switch.

2. Are you strapping the secondaries of the mic transformer into -/+ of C28? Or are you strapping an 180p off of the secondaries and wiring to R145/ground?

3. Are you still putting the 1k5 and 0.01 cap on the secondary of the output transformer like the original schematic? I also noticed the ez1073 has an extra 620R on the output.

4. What is a 681 and what do you mean by this…

Strap a 681 on the XLR output, this is what AMS Neve does as well as vintage king.

5. One more thing what size standoffs are you using?

Thanks Aaron.

 

[Aaronrash]

Yes,

Use your multimeter and figure out witch pad of C28 goes to ground and you will run your shield of the cable to that( pin 10 on the secondary) of the mic tx, the other pad will be from pin 7 on the input zobel and that will go to the other pad of C28. The way I did it though is just strapped it on the transformer like the original modules instead of dealing with more PCB tracks.

The secondary of the output will still have the zobel network 10nF and 1K5, not sure how this will work though if you have the PCB mount transformer.

681 is a 681 ohm resistor across pin 2 and 3 of the output XLR, all this does is terminate the output.

The VTB9045 isn't internally connected. For instance, take your multimeter and test for continuity between the screw by 6 and then the other side, 11. There will be no continuity... The AMS versions do have continuity so they are in fact different. I've confirmed this by testing all my AMS transformers.

I got lazy and only drilled one hole to mount the mic input to chassis and took me a while to figure out pin 6 wasn't grounded because I had the screw on the other side of the tx mounting the transformer to chassis. If it would have been an AMS transformer it would have been grounded. Make sense?

So instead of drilling another hole I just ran a jump for both sides of the transformer, this ensures pin 6 is getting to chassis ground

 

[Aaronrash]

Also make sure you take all the PCB mount zobel stuff off the board

 

[haystack]

Yes, this is amazing news Aaron. How did you find all this?

I have built mine and not sure if I can get the input trans off the pcb now.

Also what does step 3 refer to? As a bit of a newb to this, would you be able to describe this
eg what is the -10dB attenuation section?

I would definitely like to have a go at some of the alterations.

Thanks

Hay

 

[haystack]

It may be worth having a go at steps 2, 3 and 4 and see how that sounds?

Hay

 

[ding]

Aaron, I've been thinking about this all night and some things got me feeling unsure about whats going on here. I have no doubt you have skills and you are a great fabricator so please take my comments with no disrespect I am just trying to understand. Here is a list of what you did.

1. replaced key components in audio path with spec components including poly's, tants, vishay lytics. (got this, makes sense)
2. used AMS output transformer. (got this, makes sense)

At this point you felt there was something not right about the sound/sweeps and it did not match the original so you took on a munumental quest (applause) to figure it out. Here is what you felt fixed that sound.

1. mounted input tx to chassis and wired it by hand the same way its wired on the pcb. granted you gounded pin 6 and 11 but that would be a non-issue unless you mount one screw to the chassis. (I am not sure this would change the sound?)
2. mounted the output tx to chassis and wired it by hand the same way its wired on the pcb with one exemption, you used a 681ohm resistor after the phase switch and the ez uses a 620 ohm resistor before the phase switch. (I am not sure how this would change the sound?)
3. turn the attenuation pot into a full fader attenuation by jumping 134 and replacing 135 with 10k. (this should not change the sound but awesome idea)

I got to say with all due respect that this has me scratching my head a little as to the difference in sweeps from before and after. I've looked at the schemos on both, keeping in mind what you are suggesting and I got nothing to explain your results. I certainly do not doubt them. What software are you using to make those sweeps? I would not mind running a few sweeps thru the one I am currently working on.

Thanks

 

[simonsez]

Mounted the transformer into chassis can make sound better?  :-\

 

[Ilya]

I believe this has something to do with long shilded wiring that is not available on the PCB. And wire gauge.

 

[Aaronrash]

All sweeps are done in REW.

With my mods, I am bypassing as much PCB tracks as possible, that is the idea here. We are getting real capacitance with the wires mounted this way. Look at the original Neve 1073/1084 schematic. Everywhere you see a little (O) that's also a shield to ground connection. These connections are different if you wire to the PCB. They get to the same place but take a different route. and they certainly don't contain an isolation shield.

Avedis as well as vintage king techs tought me that the LO1166 was originally designed to be used with a 600Ohm termination. 681 is mathematically correct for the LO1166. You can use a 620/619 if you want as well.

All this means nothing if you're not using the AMS output transformer or at least the Carnhill LO1166 version. If you're using the PCB mount transformer then don't even bother. It's a really strange transformer and doesn't sound the same.

Now of course it matters that the input transformer is grounded. That's why it's in a metal can to shield it. If it's not getting ground then of course that's a problem. I found it odd that the Green cans are different from the Blue cans. Almost like AMS used they're output transformer as a distraction from the real difference, the input tranny.

AMS Neve personnel also told me (We can not speak on the Blue can transformers but our Green can transformers (TF10003) are of original Neve specification.

You can try this out if you want. I have found that it opened up the sound allot after doing this. I have been working with the EZ1073 for almost a year now and it has never sounded anywhere close until I did this. Just my 2 cents of experimentation, not all things especially in audio have a explanation.

Always ears first. The sweeps just confirm what I'm hearing and keep me on the right track.

But obviously these changes changed the response drastically. One thing I've noticed with all the AMS Modules is you want the graph to look like its going uphill just a bit... Just a slight degree upward. Also without the termination you want to see a little hump around 40K before the rolloff. The units do this now. Just like my AMS modules, they didn't before

 

[Aaronrash]

I've attached a picture of the high end on my AMS Neve 1084 and my EZ1073 unterminated so you can see the little hump I'm talking about. You want to see this on your units without the termination. Like I said, my EZ unit's wouldn't do this before. Not sure why.

 

[ravachol]

I'm fascinated, bewildered and miffed....

All this talk over the past few months about the original PCB layouts with solder bath, the micro-inducting mojo of the Canford wire vs. the ez1073 ground plane: all this was beside the point?

That's incredible -- shocking, really. Well, maybe not; I remember somewhat doubting these things -- but I was swept away!

Aaron, what's puzzling me now is that I'm still trying to see what's really new about these mods. Back in January, you already had the input and output transformers mounted on the chassis. What changed along the way?