8 channel Neve-ish mic pre

[CM84]

I've been messing around with this for a while, and now that I've finally got something working I thought I'd share it with you all! I've picked up loads of good info on here over the years, so nice to finally post this! I had an account here but I can't seem to find my details (it may have been linked to an email account I've binned - too much spam!!). Anyways...

I built a 1073/1290 kit a while back, which mostly worked, so I thought I'd expand it to be something I'd actually use. I'm rarely recording just one or two channels at a time, so a single channel of mic pre isn't going to get used - 99% of the time I'm recording on location, I'm not interested in carrying any more weight than I need!

So I set myself the challenge of making an 8 channel, 2u version - something I could squeeze into a rack and not worry about. It had to have an internal power supply (the weak link, time to do some serious learning!), some sort of metering, and I thought it might be nice to have a HPF - 8 channels of neve on a kit with some decent filters? Drum heaven! I was fully aware that there'd be inevitable compromises from the tight-ish packaging, but I've plenty of modern clean mic pres if that's what was required. So I laid out some circuits and a couple of revisions later I had a working mic pre. It seems very stable and solid, and nice and quiet when using a bench supply - quieter than the RME 12mic I compared it to - a good start! The filter is just the inductor based circuit from the 1073. And I popped a pad circuit on the front of it all.



Next up was a power supply. I knew full well that a linear supply was a bad idea, what with it being inches from the mic pres, but I was impatient to get it moving forward. So it's a +24V, +48V and +15V supply using LM350/317/7815 regulators. And it works nicely of course - even I can layout a linear power supply! And it's not as noisy as I feared.. It's not great, and it's not good enough for what I want, but it's usable in the short term. It's one of those things that I had to hear for myself. You read so much about distance being the key to everything when it comes to linear supply, but now I actually know!



Finally I set about a basic level indicator board - optimistically called my "vu pcb". It's a super basic comparator circuit with 4 LEDs per channel. I'd love to make a proper peak indicator circuit for this, but that's a job for another day. Of course there's been plenty of fudging bits to fit in the case too, but compared to my first revision of this (which was more jumper wire than circuit board!) this is perfection!






So next step is a switching power supply.. I'd love some recommendations on good books if anyone knows any? I know that just putting the linear supply in an external box would solve most of my issues, but I don't want an external power supply.. It's actually something that puts me off buying certain pieces of gear - it's one more thing to pack, one more thing to get tangled... That said, I do want to finish this and move on with my life(!) so any and all thoughts are welcome!

Another thing I don't understand is why the PAD circuit is so crap. It's just a 3 resistor circuit (https://www.jensen-transformers.com/wp-content/uploads/2014/08/as016.pdf), but I get quite a lot of buzz when I engage it...

Lots to finish off!

 

[dmp]

Wow, that is an impressive goal. Making a great preamp is much harder once you pack channels in tightly.
You have pad and phase in the push button switches and the HPF frequency in the lower knob?
There's nothing wrong with the pad circuit - if you are getting buzz you have a implementation error. What's your actual wiring of the pad + input transformer? A buzz implies there is a bad interaction with ground.

Also, I don't think you need to go to a switching power supply. The issue you'd be most likely to run into is magnetic coupling with the toroid. Channels close to the psu would have more hum than far away. enclosing the toroid with mu-metal might be necessary to make this work.

 

[CM84]

Here's what I've got before the gain switch... It all looks fine to me, but then again I've been working on this on-and-off since the first covid lockdowns, so I've probably forgotten more about what I've done that I remember! Maybe you can spot the obvious..


The trick is finding affordable mu metal! It is definitely proximity that creates (most) of the noise.. Also, right now I've got a 50VA toroid in place for the power, which clears the top and bottom of the case by a couple of millimetres top and bottom.. I don't really have space. Depending on the final power draw I might be able to squeeze this down to a 40VA though, in which case I could try something.

No phase switch yet - in the end I had the space I had and went for 48V and PAD front switches. HPF on the rotary, with a bypass being the first step. If I could find some smaller LED push buttons I could redesign it one day, if I get the bugger working!!

 

[dmp]

I don't see anything wrong with that in principle. I'm guessing the problem is occurring through the ground. As a test, disconnect the +24v LED and see if you still get the buzz. You're going to want to be careful with your ground routing. The audio, switching, pin 1, and chassis. Keep them separate to the star point. The 9045 also has a pin 11 connection to the mu metal case that needs to go to chassis. Not sure offhand if pin 6 should go to chassis or audio ground.

I got adhesive backed mu metal from Noo elec and it worked great in a similar situation. A sheet is only $15 now.
You can see here how I wrapped the toroid in this v76 project. You want it to be continuous so I put the mu metal on a round can that bolted over the transformer and then you'd want to put a layer under.


But you'll want to make sure everything is working before doing that
Neve channels take a lot of current. 8 channels might be 1200 mA @ 24v? You could have a problem with the heat from the linear regulator at that amount of current - you might want to run it to the back wall of the enclosure instead of right on the bottom (b/c heat rises). Before making final decisions test the psu out with suitable 2 watt resistor (ie 20 ohm)
I see the challenges about the linear supply, but I don't get "a linear supply was a bad idea"... a switched supply has high frequency switching right in the box with your preamps - that seems like a BAD idea.
Best idea would be external, but I see your reasons for wanting it internal. I built a 8 channel Jensen 990 preamp in 2U but had to do an external PSU. Would have loved to put it inside but didn't have the room

It looks like you have some extra height allowable for the toroid, so potentially you could stack two smaller ones and wire the outputs together to get the VA summed?

 

[CM84]

Thanks,

Lots to think about there.. Thanks for the mu metal link, I'll check that out.

I just thought I'd always be fighting an unwinnable battle with the the linear supply! Everything and everyone I read told me so! I think you're the first to suggest it might be possible.. There are reasons most equipment is switch mode these days though, and listening to modern mic amps you can hear they're absolutely silent. I hadn't thought at all about two smaller toroids though... Presumably they'd throw off less noise in the first place too, as well as being easier to shield. This isn't something I've ever really looked at, how does this work in practice?

Regarding heat, one mic amp draws around 70mA @ 24V on my bench supply. As I showed on the schematic, I've got the 48V and PAD LEDs also powered from that (bad idea?) - with them both on I'm around 100mA, so I reckon around 800mA or so all in. Plus the 15V LEDs and the 48V, could be 30 Watts in there.. The regulators are all mounted on 3mm steel, which I hoped would help dissipate a fair chunk - optimistic?

 

[dmp]

This isn't something I've ever really looked at, how does this work in practice?

I was thinking you could just wire the secondaries in parallel. I haven't done it - but I don't see why it wouldn't be fine. You want to make sure they are exactly the same (turns and phase) so they don't fight each other

The regulators are all mounted on 3mm steel, which I hoped would help dissipate a fair chunk - optimistic?

Attached to the case is a good way to conduct heat to the case, but the heat still needs somewhere to go. It will conduct out radially through the steel somewhat, but the free convection into the air will be rising right into your regulator / power supply. Contrast that to a mounting upwards, then the free convection rises up and away from your components. Which is why I suggested attaching it to the back so the heat can dissipate outside the case.
The amount of heat the regulator needs to dissipate is the current draw (i.e. 800 mA) times the voltage drop (Vin-24v)

The reason you want to make sure things are working well is the regulator has a high temperature shutoff. I'd suggest testing it out with a little safety factor (1000 mA) and see what temp the regulator is running at. If it's too hot, redesign.
Not sure how you have 48v also powered by the 24v regulator?

 

[shabtek]

I have no real experience using switch mode for this kind of thing, but it would likely save some heat and definitely weight and power used.
dmp has good advice
remote power supply is generally good for high gain, especially when packed like a keg of fish
impressive project so far

 

[rp3703]

I am interested in a 24V switching PSU as well, although I only need it to power two 1290's. I am probably the least knowledgable person on this forum but I just swapped out my linear PSU in my DIY 8 Channel API-312 unit for an external switching PSU and it was dead quiet. Prior to that, I was getting low level EMI in the channels closest to the toroidal. I was lucky enough to come across an external switching PSU being sold by DIYRE that met all my requirements and all I had to do was put together a filter circuit for the PSU to plug into. luckily, I was able to retrofit my old linear PSU PCB. I would love to find something similar in 24V & 48V but have not had much luck as of yet. From what I've read switching PSU's are fine as long as they are built right and have enough filtering to eliminate the high frequencies. Almost all the 500 series racks use them which is why I decided to.

 

[amplexus]

I used a bunch of off the shelf switching modules (mostly MeanWell units) for an 8 ch Neve VR racking project a while back. Built a CLC filter pcb for each rail (I know the preference around here is a Capacitance multiplier) and it was dead quiet. Light weight and low heat output, not to mention cheap AF to get the current I needed on a 5 rail supply. Honestly I'm on Team SMPS these days for new designs.