Neve V3 Reduce Power Consumption

Hello everybody,

I've got a Neve V3 to play around with. Now I've been reading a lot on the experiences of other people with that console and there always seem to be the people who advocate upgrading the chips to a newer generation, ideally a more efficient design.
Now I've been looking at the OPA16xx designs and some of them really seem to cut down on the current they draw, compared to the 5532s that make up a lot of the used opamps.

Now my primary concern is cutting down the power that the whole desk draws. Right now, each year it consumes between 30000 and 50000 kWh. That's quite a chunk.
A lot of it obviously comes from the crazy heat the console produces. I suspect that most of it comes from the number of opamps. My hope is that by an increase of efficiency, I can not only reduce the used power, but also bring down noise, prolong life of electrolytics and also save some power on the AC.
I am aware that the sound of the console can change by a customization like this, but the power consumption is a bigger concern to me, as long as everything works as it should. Obviously, it would be ideal if the sonic changes are minimal and I would definitely try some different Opamps.

Has anybody of you possibly done a similar change to a Neve V Series? Would be interesting to hear about some experiences. Also even if not, does anybody of you have some useful additional information?

Best,
Phil

Almost every single op amp made operates in class B. This means when there is no signal, the idle current is very small and the power consumption should be quite modest. As soon as you send signal through it, class B stages up the current draw depending on the the signal level and the load. So, unless Rupert did something sexy to persuade the 5532s to operate in class A , I don't think changing op amps will make that much difference.

Cheers

Ian

> each year it consumes between 30000 and 50000 kWh.

Really? 5,000 watts 24/365?

I heat my garage with a 5,000 Watt heater. I can't imagine running that thing for long in an insulated space.

50,000KHh/year is $800 a month at my electric rate.

In Orbison's V3 I see maybe 2,000 opamps. At a generous 0.3W each that is 600 Watts not 5,000 Watts.

I'm wondering if a decimal point slipped.

o.k. the V3 runs at a constant temperature of over 108 degrees under the hood. It kills caps. and causes issues no different then a VR in that regard. It came out after Neve was sold and was under AMS/Neve There are things you can do to make it more efficient.

For starters I would recommend  checking out atomic instruments. They make a power supply that will reduce the electric bill by a lot. From there, there are some things you can do to reduce the heat. There was a chap who had a thing called the ice mod. it dropped the internal temperatures by 20 degrees.  He still may be doing those.  There is also the mad labs upgrades, that John Musgrave would do. the new 88R is said to be a Neve VR with Musgrave  upgrades. I may have the mad lab notes floating about.

PRR said:

> each year it consumes between 30000 and 50000 kWh.

Really? 5,000 watts 24/365?

I heat my garage with a 5,000 Watt heater. I can't imagine running that thing for long in an insulated space.

50,000KHh/year is $800 a month at my electric rate.
...

Click to expand...

I'm sorry, I did actually make a mistake. The kWh number is for the whole electrical bill. Although, the combination of console and AC makes up most of it. Something between 600€ and 1000€ seems about right.

ruffrecords said:

Almost every single op amp made operates in class B. This means when there is no signal, the idle current is very small and the power consumption should be quite modest. As soon as you send signal through it, class B stages up the current draw depending on the the signal level and the load. So, unless Rupert did something sexy to persuade the 5532s to operate in class A , I don't think changing op amps will make that much difference.

Cheers

Ian

Click to expand...

I also don't believe that there was something done to make the ICs run at higher idle currents. Still, according to the datasheets, the Iq of OPA16xx can be less than half of the 4 mA that the 5532 draws. The ICs are the only thing that I can imagine to create the excessive amounts of heat the console radiates.

pucho812 said:

o.k. the V3 runs at a constant temperature of over 108 degrees under the hood. It kills caps. and causes issues no different then a VR in that regard. It came out after Neve was sold and was under AMS/Neve There are things you can do to make it more efficient.

For starters I would recommend  checking out atomic instruments. They make a power supply that will reduce the electric bill by a lot. From there, there are some things you can do to reduce the heat. There was a chap who had a thing called the ice mod. it dropped the internal temperatures by 20 degrees.  He still may be doing those.  There is also the mad labs upgrades, that John Musgrave would do. the new 88R is said to be a Neve VR with Musgrave  upgrades. I may have the mad lab notes floating about.

Click to expand...

I did found a Bobby Summerfield who seemed to accomplish a cooling system for the console. Sadly, I can't find too much visual information about how he went about it. I will certainly follow that lead, thanks! Also the tip for the Atomic Instruments PSU is awesome!

from what I know of the Ice mod. It works around  a vacuum sucking out the hot air. I am not quote sure what mods to metal work, etc are required but I have seem some video with a shop vac  taking measurements before turning 0n the shop vac and after the shop vac.

Sucking out the hot air will not change the power consumption. (Actually raise it a bit.)

PRR said:

Sucking out the hot air will not change the power consumption. (Actually raise it a bit.)

Click to expand...

true but it does reduce the temp under the hood which will prolong the life of capacitors and  plastic switch cams which fail from heat.

Hello,
I have modified some channels, with the Opa 1642 and the neve works colder..1.8mA, has better slew rate than the 5532 and you can eliminate electrolytic capacitors from the signal path. There are some ICs that can not be changed in the equalizer because they oscillate

you can try the opa 1692, only 0.65mA  ;D

bi fets in the EQ make a nice sonic change and gets rid of the pinched feeling. IC 10 in the EQ will sing like a bird... cant change that puppy.... you can also add power switches/relays to turn on and off buckets. Atomic is a great choice (rip norm) and if you are real tweekie you can strap the transformers slightly low and up the voltage to 18 to take a little dissipation off the supplies. 240v also helps a tiny bit for psu heat. I am around you can pm me on fb messenger always.
John m

Cheagor it is a bit knee-jerk to blame the internal high temperature with the opamps.  Like your initial mistake with the kWh/annum that the desk consumes, it is easy to miss seeing what power dissipation items exist in the desk and individual modules etc, and how different contributors affect the temperatures in the regions where higher temperature cause part lifetime concerns.

Certainly well worth investigating further, especially along with internal cooling flow routing, but perhaps only as part of a higher priority concern such as very old electrolytics.

ruffrecords said:

Almost every single op amp made operates in class B. This means when there is no signal, the idle current is very small and the power consumption should be quite modest. As soon as you send signal through it, class B stages up the current draw depending on the the signal level and the load. So, unless Rupert did something sexy to persuade the 5532s to operate in class A , I don't think changing op amps will make that much difference.

Click to expand...

Actually, many Neve products add a 10k resistor from one rail to the output, so the output stage operates in class AB. the idle current for one full 5532 goes from 8 mA to 11mA.
This has been designed so that class A operation is guaranteed for signals lup to 6dB above nominal.
There's no increase in current draw from idle to this elevated level.

trobbins said:

Cheagor it is a bit knee-jerk to blame the internal high temperature with the opamps.

Click to expand...

It's not like there were so many usual suspects. Indeed, LED's and illuminated push-buttons in the mid section cannot be ignored.
Or are there relays in that beast?
I would think LED's are powered by CCS at about 10mA, so a whole string of LED's would draw about as much as one 5532.
Indeed, I would try the OPA1632, for preserving the sonic character of BJT's - for whatever it means- or the OPA1642 or 1652 for the benefits of JFET inputs.

I got a tour of a big London studio complex years back from a former work colleague,
Their Neve V ran hot as hell  and would need modules pulled and sent to the service dept on an almost weekly basis. The service guy didnt have anything good to say about it .
I did get a chance to look over a pulled module and just like Abbey said ,theres a lot going on under the hood , dozens of relays, leds and op amps per channel.

I doubt subbing op amps(where possible) would make a huge difference overall,
maybe skipping a channel module every 4 or 6 slots and replacing it with a vented panel might help or as previously suggested an active cooling system.

The actual cost of the power for the console might not be the only factor , if you end up having to run aircon to make your control room habitable  you could easily end up adding an extra zero to costs.

Cheagor said:

I'm sorry, I did actually make a mistake. The kWh number is for the whole electrical bill. Although, the combination of console and AC makes up most of it. Something between 600€ and 1000€ seems about right.

Click to expand...

Auch...

ruffrecords said:

. So, unless Rupert did something sexy to persuade the 5532s to operate in class A , I don't think changing op amps will make that much difference.

Cheers
Ian

Click to expand...

Hi Ian, this console was released in 1987, Rupert was long gone from the company by then, 12 or 14 years after any involvement with Neve products.

Substituting the 5532 for opa1692 the temperature drops about 49ºC to 36ºC (in the equalizer). Mine works at 26ºC with a little help from fans .. and it sounds great I am in love with this table !!

A cheaper alternative to atomic is to use these PSUs:

https://www.tequipment.net/TTi/EX2020R/DC-Power-Supplies/Lab-Power-Supplies/

These are used by raindirk

kaguenpituak said:

A cheaper alternative to atomic is to use these PSUs:

https://www.tequipment.net/TTi/EX2020R/DC-Power-Supplies/Lab-Power-Supplies/

These are used by raindirk

Click to expand...

I had to install these PSU's on a Raindirk Symphony, Cyril supplied them.  They worked very well & replaced 2 really heavy 5U PSU's, but were quite expensive.      The only thing that wasn't so good was that 3 of them supplied the whole desk + & - something like (17.5v & a logic supply).  The original PSU's supplied half the desk each, but had a neat arrangement with back to back diodes between the 2 halves, so that if one supply died the other automatically took over, supplying both halves of the console.  This was rather useful since the console came from a mobile truck.  They were much more reliable than the old supplies which blew pass transistors once in a while.