Neumann VCA101 Layout/Gerbers available

Help Support GroupDIY:

amplexus

Well-known member
Joined
Jan 11, 2018
Messages
132
Not sure if anyone is interested in this or might need it- but I've had occasion to need to rebuild a VCA101 to repair some U473's for a client. As part of the process I've ended up de-potting the NFG module, stripping it down, documenting the layout and reproducing the PCB layout as closely as practical. Mostly I did this so I would have an accurate layout to rebuild the boards after stripping down- so right now they're a bit 'sloppy' from a neatness point of view- but the layout is correct to the original PCB with one or two very marginal component orientation changes.

If this would be useful to anyone I'll clean up the screen layer, add some notes and post the gerber files for download.
 

Attachments

  • Screen Shot 2022-01-10 at 4.28.11 PM.png
    Screen Shot 2022-01-10 at 4.28.11 PM.png
    293.2 KB · Views: 71
  • NeumannVCA101.zip
    21.8 KB · Views: 13

MidnightArrakis

Well-known member
Joined
Sep 25, 2015
Messages
179
Not sure if anyone is interested in this or might need it- but I've had occasion to need to rebuild a VCA101 to repair some U473's for a client. As part of the process I've ended up de-potting the NFG module, stripping it down, documenting the layout and reproducing the PCB layout as closely as practical. Mostly I did this so I would have an accurate layout to rebuild the boards after stripping down- so right now they're a bit 'sloppy' from a neatness point of view- but the layout is correct to the original PCB with one or two very marginal component orientation changes.

If this would be useful to anyone I'll clean up the screen layer, add some notes and post the gerber files for download.
There is an "acute-angle" track between the silkscreens of "C2" and "R5". This type of routing track is also known as an "acid trap" during the PCB fabrication etching process. I would move the "R5" silkscreen over to the other side of "Q3", so then the track going to the emitter of Q3 can connect orthogonally to the horizontal track immediately above it like on Q4.

Then, there seems to be some kind of "double-routing" over by the emitters of both Q2 and Q7 that looks like some kind of "foopah" in routing. You may wish to clean that up.

As a general rule.....it is usually best - NOT - to have such close minimum track-to-pad spacing as I see here if you don't actually need it. Having a little bit more space just kind of helps out not only during the fabrication process, but also during the assembly process, especially if these boards are hand-soldered.

There are a couple of other "almost/kind of acute-angle routes" over by the top pad of C4 and just above the collector of Q9. In short, any place where you have a 90-degree right-angle routing track, the inside corner of said track can trap acid during the etching process and it causes the copper of the track to be ever-so-slightly delaminated from the FR-4 material. Over time, this delamination can cause the track itself to work loose from the board and then one day it will split, causing that portion of the circuit to "open" and fail. So, "right-angle" routing is generally considered to be a "no-no" in PCB layouts.

Just my 2-cents worth!!!

/
 

amplexus

Well-known member
Joined
Jan 11, 2018
Messages
132
There is an "acute-angle" track between the silkscreens of "C2" and "R5". This type of routing track is also known as an "acid trap" during the PCB fabrication etching process. I would move the "R5" silkscreen over to the other side of "Q3", so then the track going to the emitter of Q3 can connect orthogonally to the horizontal track immediately above it like on Q4.

Then, there seems to be some kind of "double-routing" over by the emitters of both Q2 and Q7 that looks like some kind of "foopah" in routing. You may wish to clean that up.

As a general rule.....it is usually best - NOT - to have such close minimum track-to-pad spacing as I see here if you don't actually need it. Having a little bit more space just kind of helps out not only during the fabrication process, but also during the assembly process, especially if these boards are hand-soldered.

There are a couple of other "almost/kind of acute-angle routes" over by the top pad of C4 and just above the collector of Q9. In short, any place where you have a 90-degree right-angle routing track, the inside corner of said track can trap acid during the etching process and it causes the copper of the track to be ever-so-slightly delaminated from the FR-4 material. Over time, this delamination can cause the track itself to work loose from the board and then one day it will split, causing that portion of the circuit to "open" and fail. So, "right-angle" routing is generally considered to be a "no-no" in PCB layouts.

Just my 2-cents worth!!!

/
Shit, missed that! I generally am pretty good at spotting any right angle traces (I know it's bad layout karma!), thanks for pointing those out. I haven't fabbed any of these at all... I really just did this as a roadmap so when i tore the old board down I'd have the component locations so this is like, power drafting based on the actual PCB (now that I think on it, the actual PCB may have had some right angle traces so I might have just copied it directly)

I hear your feedback on the track to pad spacing. This is set a 0.2mm- if i go any larger I'm getting "pad to close to pad" errors on my transistor footprints. Tracks probably aren't a big deal to keep clearer, i likely don't NEED .5mm traces- again, just straight copied the original there.
 

gyraf

Well-known member
Staff member
GDIY Supporter
Moderator
Joined
Jun 4, 2004
Messages
10,609
Location
Aarhus, Denmark
Thanks!

Do we know what transistors to match carefully?

It looks to me like Q3-4-9-10 would very much like to be same-same - and perhaps also Q2-7

/Jakob E.
 

amplexus

Well-known member
Joined
Jan 11, 2018
Messages
132
Thanks!

Do we know what transistors to match carefully?

It looks to me like Q3-4-9-10 would very much like to be same-same - and perhaps also Q2-7

/Jakob E.
Yes. The original schematic has them marked with an asterisk. The note is in german, but even my non-existent german skills made it clear that it meant matched.

also I believe R9 and R10 were supposed to be “matched” as well- but i think they just wanted a 1% part there. Still, can’t hurt to go better. I used 0.1% dale CMF when i did the rebuild.

Will confirm when i get to the shop today.
 

amplexus

Well-known member
Joined
Jan 11, 2018
Messages
132
Then, there seems to be some kind of "double-routing" over by the emitters of both Q2 and Q7 that looks like some kind of "foopah" in routing. You may wish to clean that up.

This is actually how the trace looks on the original pcb. I was initially more interested in having a “picture” of the original layout so I threw it in there. I can’t see how it would be functional so I agree, there’s no reason I can see to keep it that way in a reproduction.
 

amplexus

Well-known member
Joined
Jan 11, 2018
Messages
132
Thanks!

Do we know what transistors to match carefully?

It looks to me like Q3-4-9-10 would very much like to be same-same - and perhaps also Q2-7

/Jakob E.
Ok so Q3, 4, 9, 10 all want to be matched. R10 and R11 need to be 1% or better.

In addition, on the original Q3, 4, 5, 9, 10 had a small aluminium bar glued between them prior to being potted (obviously)- so clearly they wanted good thermal coupling on them. In my rebuild I did the same, but with thermal compound and clamping pressure. In addition C4 was pressed right against the body of U1, and soldered directly to pin one of U1, and of R3- it did not have pads on the pcb. My suspicion based on how tight everything was is that may have been more for space saving than anything else. C4 was a ceramic cap in the original.
 

amplexus

Well-known member
Joined
Jan 11, 2018
Messages
132
ok, here's a bit cleaner revision. Fixed up the acute angles and weird trace from the original. Ignore the flags, they're just courtyard overlaps but the components shouldn't overlap at all in real life.

Added notes to the schematic re: matching and original components.

The only thing I haven't landed on is a good modern sub for the BF414/Q6. I think a 3904 should work. When I was doing the rebuild I found that anything higher than a gain of around 40-60 would just pin the gain reduction on the 473's. I was able to reuse the BF414 as it was in perfect working order so I didn't have to dig into it too far.



Screen Shot 2022-01-26 at 1.40.44 PM.png
 

Attachments

  • VCA101_Rev1.pdf
    72.2 KB · Views: 24

amplexus

Well-known member
Joined
Jan 11, 2018
Messages
132
Hey all. Here's the Gerbers for the last rev. Updated the OP to attach them there for easy finding as well.
 

Attachments

  • NeumannVCA101.zip
    21.8 KB · Views: 20

firerunner

New member
Joined
Sep 20, 2017
Messages
1
Thanks a lot for posting this! I was looking to remake the VCA101 in smt but I think I'll start by building one of these up. Cheers!
 
Top