Sunday, July 11, 2021

NEVE: 1073 Channel Amplifier

I went down the Neve rabbit hole with this post, looking at how the channel amplifier works and trying to use one of the preamp stages as a fuzz/overdrive.

Long story short - I couldn't come up with something that I was happy with using the mic preamp stage, so I have walked away from this one for the moment.   Came close, but just not worth spending any more time on.

I will try the other amplifier sections and see if I have better luck.

On the upside, it got me interested in what other preamps were out there, and I had much better luck with the Revox A77, the same as a Trident A-Range mic preamp.   I ended up making one of these and boxing it up, and it's pretty cool.  

- - - - - - - - - 

So what am I trying to do?

  • Modify/abuse a Neve preamp so it can be a fuzz or overdrive
  • Preferably running at 9v, but I can live with something larger using a charge pump (the Neve runs at 24V)
  • Maybe include some EQ from the Neve, but simplify to keep it pedal friendly
  • Avoid transformers.  Yes, I know, everyone says the transformers are a big part of the sound, but I don't care what everyone says cause proper Neve-style transformers are costly and large, and I'm not making an overdrive/fuzz pedal with a really expensive studio grade transformers known for their low distortion
  • Avoid the insanely complicated Sensitivity switch (it's the input gain control, which sweeps all the way from mic to line level)

So what am I not trying to do?

  • Recreate a Neve channel amplifier on vero (madness)
  • Use unobtainable and/or expensive parts
  • Reverse engineer a JHS Colourbox

NEVE 1073 CHANNEL AMPLIFIER SCHEMATIC

You can find a full manual, including schematics, here    This may be better than what you have seen previously, as it is the complete manual with tech specs.

SIGNAL PATH / BLOCK DIAGRAM FOR CHANNEL AMPLIFIERS 1066, 1073, 1078

A few partial schematics floating around on the web do not contain this diagram, which is very easy to understand compared to the next one, which is more commonly found on the net.

NEVE - SIGNAL PATH / BLOCK DIAGRAM FOR CHANNEL AMPLIFIERS 1066, 1073, 1078


BLOCK DIAGRAM

This diagram is confusing, but it makes more sense with the reference above and the other items below, where I've covered the EQ sections individually.

NEVE CHANNEL AMP - BLOCK DIAGRAM


NEVE MIC PREAMP SCHEMATIC

Three different preamps are used in the channel amplifier, which are used as gain stages in various places.  The multiple boards are made up of two or three of these preamps combined.  Some are used as gain stages for the active EQ, Mic and Line inputs and, finally, an output amplifier.

As it happens, this was used as the clean preamp on the Way Huge Pork Loin.  

NEVE PREAMP SCHEMATIC

MODDED MIC PREAMP SCHEMATIC

MODDED PREAMP SCHEMATIC

Just gain and level at the moment.  The frequency response is very flat for a pedal, so for my taste, it doesn't need a tone control.  Even though it has an element of fuzz, it doesn't have excessive high frequencies.

LTspice

It doesn't look exactly like this on the scope; the wave isn't as squared off as this.  It was more triangular with flat tops.  Hopefully, after I work out the gain sweep properly, I'll add some actual traces at a later date.







OUTPUT AMPLIFIER

NEVE OUTPUT AMPLIFIER SCHEMATIC


LTspice





Frequency response varies quite a bit, depending on where I take the trace from (note letters on trace vs schematic).  I have done a bit of a dodgy workaround where the transformer should be, so it might be a result of this.




MISC


Now I know you can't see any details here, and there's still a little more to add - but this is why I will not be attempting to recreate a Neve channel amp on vero - it would be gigantic.




THE NEVE 1073 EQ SECTION

NEVE BA205 BOARD (LOW & HIGH SHELF EQ)

NEVE BA205


THE ASSOCIATED PREAMP

NEVE PREAMP SCHEM

WHICH TRANSLATES INTO THIS

It's quite similar to an active James EQ, with the low shelf frequency being selected by a rotary switch (switching between different values capacitors), with a pot to control cut and boost.  The high shelf just has a cut and boost pot.

Hopefully, this is easier to understand, and it gives me a starting point if I ever decide to use some of it in a pedal or adapt it in some way.  The next section in the signal chain is the mid-range control (Presence), which looks quite challenging.   

NEVE BA205 BOARD (LOW & HIGH SHELF EQ)




NEVE BA211 BOARD (SWEPT LO-MIDS TO HIGHS)

I'm pretty sure this has an error - instead of C & E (which are preamp in/outs), it should read K & M.  K & M are the in/outs of the other preamp on the same board.  C & E have already been used for the low-high EQ section. 

There also seems to be some differences between the schematic, some including a 10k option, others not.  There's also an alternate schem for the board above, with an extra resistor. 


THE ASSOCIATED PREAMP



WHICH TRANSLATES INTO THIS (HOPEFULLY)

I say hopefully, as the frequency response didn't exactly match the stated values for 2, 3, & 4.  Looked good for the rest.




NEVE BA182 BOARD (HIGH PASS FILTERS)

This is the simplest of all the EQ sections, but it did have me stumped for a while as to why there are two identical sections with the same lettering - it's pretty obvious now; it is because they are connected...    As mentioned above, this is a passive filter, so no preamp is required. 



WHICH TRANSLATES INTO THIS





19 comments:

  1. Following this with big interest...

    ReplyDelete
    Replies
    1. I’ll update over the weekend. There’s another section that I’m working on that I haven’t added to the post yet. There’s a few different preamps and amplifiers in the channel strip.

      Delete
  2. Love this idea! I'm also in to adapting pro audio & synth builds for guitar, and it can definitely be a challenge to change the supply voltage, especially on finely tuned circuits like Class A amplifiers. I wonder if you need to check the sim on +24V to compare the collector & emitter currents to the +9V version. A big supply voltage change like that has probably affected the transistor biasing, but can be adjusted with some resistor value changes. That also might be causing your sudden clipping.

    I find it difficult to read the old Neve schematics that use letters & subcircuit blocks, but fortunately a lot of clone versions have schems that are easier to read. Check these out:
    - AML ez1073pre (page 39): http://www.audiomaintenance.com/downloads/ez1073pre_colourbook.pdf
    - DON Classics NV73 (also has transistor reference voltages) - https://thedonclassics.com/nv73/schematic/nv73-diy-schematic.pdf
    - Analog Classics ez1290 (1073 w/o EQ or Line input) - https://cdn.analogclassics.com/wp-content/uploads/2021/07/ez1290_build_guide.pdf

    You should be able to see the rotary gain switch wiring w/resistors a lot more easily on the AML schem, which I couldn't figure out for the life of me on the original Neve block version. It looks like the 1073 uses 2x BA284 stages, the first of which feeds into the second at higher gain switch settings. It might be worth trying that for a smoother-sounding overdrive sound. This GroupDIY thread talks a bit about the gain staging: https://groupdiy.com/threads/sonic-penalties-for-2-gain-stage-vs-3-stages-in-neve-type-circuit-1073-1272.74456/

    Keep us posted on your experiments, and if you're looking for another preamp circuit to try, check out the Helios Type 69 Preamp: https://groupdiy.com/threads/helios.58393/

    ReplyDelete
    Replies
    1. Thanks for those links (and your interest). I’m modelling the entire thing in LTspice at the moment, including EQ sections, to get a better understanding of how it works.

      To keep things practical for a pedal - I’m still looking at just one overdriven preamp, and ideally at 9v but who knows at this stage. I may add some EQ along the way - the presence control looks to be the most difficult to implement with an odd choke. The rest is pretty straight forward.

      Delete
  3. That last LTSpice screenshot is absolutely ridiculous! Will it actually run a .tran command? Do those dashed lines around parts of the circuit serve a function in LTSpice or is it just for organization? Simulation is where I start as well, it's a great way to play detective on what's actually happening in a circuit and what sections are critical.

    The 1073 EQ has been cloned and modified quite a bit, and one of the first steps is usually to remove the chokes & inductors. There is a great DIY build called the "Not Yet Another 1073 EQ" that I've played around with in LTSpice and on a breadboard. It's much more doable than the original Neve circuit: https://groupdiy.com/threads/nyan-1073-eq-x-formerly-cheaps73-poor-mans-1073.70096/
    There are even some Spice sim files on the Github (although the original versions I DL'd had errors, so maybe check the documentation): https://github.com/ravettel/Nyan-1073-EQ

    One thing that has made working with pro audio circuits easier is using the original PSU voltages, at least until I've got a good feel for the circuit. I use either an LT1054 bipolar charge pump (for +-15V circuits) or a 1044/7660S-based 25V charge pump (for +24V circuits) for breadboarding. Tagboard has a vero layout for the latter: http://tagboardeffects.blogspot.com/2016/11/charge-pump-extravaganza.html

    ReplyDelete
    Replies
    1. Hello - yes it is fine running a tran. Dashed lines are just to help me stay organised. I'm laying it out the same way as the schematic to some degree, and I've also pulled a sections out to work on individually. I'm tackling one EQ section at a time. Just did the high / low last night. Will do the HPF next as it looks easy, and will look at the presence later.

      I've been looking at some of the work out there that others have done as a reference. Thanks for sharing links - feel free to send me more.

      I have a bench power supply, so I can easily run at the original voltages, and swap voltages around while working on the breadboard. Makes life a bit easier.

      Delete
  4. Hi this is interesting, but one issue I see is if you plug in guitar directly the low impedance loads down the guitar pickup. The end effect is the low pass filter (caused by the huge inductance of the guitar pickup) -3dB point gets shifted way down. I measured my Strat inductance at about 1.2 Henries, so it's huge. Another issue is BJT's are noisy for a high impedance source. I'm designing some guitar circuits myself and the first stage is much quieter using an FET due to it's inherent high input impedance. Look at noise simulation in LTspice.

    ReplyDelete
    Replies
    1. Hello - with the Neve I was looking for the sound of plugging directly into an old console and driving it. So I was ok with the impedance limitations in that regard.

      Delete
  5. Replies
    1. Hi Pete - no layout for the channel strip as it would be gigantic and full of hard to find parts. I’ll probably look at the preamps again over the coming holidays. The Revox preamp is really good if that’s of interest (with vero layouts)

      Delete
  6. I did some preliminary PCB tracing on the JHS Crayon, which is supposedly based on the Neve 1073 circuit. Looks like 3 BA284 stages in series, with an opamp for the tilt EQ & Hi-Cut controls: https://forum.pedalpcb.com/threads/jhs-crayon.12256/

    ReplyDelete
  7. I used your diagrams to assist in re-drawing the entire circuit as one full sized schematic. Thank you they were very helpful. I noticed however that for your drawing of ba205 that RA1 12k and RA2 39k are on the wrong side of R3 6.8k, In the block diagram the resistors are between pins f and k not pins f and v: Can you please confirm if I am correct. thank you

    ReplyDelete
    Replies
    1. Confirmed: you are correct. Good spotting and thanks for letting me know.

      Delete
    2. Thanks Andy! yeah I just got into electronics last year and as a first project I basically read the entire manual over and over until I was able to put all the pieces together. I actually drew a full size diagram on my dining room table on butcher paper. Your diagrams helped with the shelving and high pass filters. I'm diving into the math and art of electronics in conjunction with all this. My first home build consists of wire wrap dip sockets with pluggable headers which the components are soldered to. My EE friend is kind of guiding me through it. Hoping to start testing soon. Next step will be winding my own output transformers. I'm having so much fun with all this.

      Delete
    3. That’s quite the project - wow….

      Tranformers can be tricky things. If they aren’t properly shielded they can cause noise.

      Are you also making you own inductors?

      Delete
    4. I bought the inductors from Don Audio, a Carnhill for the HPF and two of their air gapped inductors for the presence circuit.

      for the output transformer I dig tons of digging through gear space and pieced together enough documentation to duplicate the Marinair from a few guys that reversed engineered it. I have a source for the bobbins and laminations too. Would be glad to send you everything I have. They're kind of weird with their information and weren't too hot on sharing it with me but luckily the internet holds info forever so I was able to find everything I need. Let me know on those transformer documents. They're pretty cool actually. and it
      's supposed to be way better than the carnhill in terms of low end distortion.

      Delete
    5. I see you updated the diagram. fantastic!

      Delete
  8. I am by no means an expert on transformers and inductors - good luck. Let me know how it all works out, you can reach me on guitarverop2p (at) gmail.com or @balmoralelectric on instagram

    ReplyDelete