Amek AT8 Hybrid Line Driver Questions

[krabbencutter]

I'm currently racking an Amek MZ15-RN Channel and the circuit uses Amek's AT8 Hybrid Line Driver for the Buss & Insert outputs. Unfortunately the service manual omits the component values for it and the PCB only contains the capacitors and some strange black pads. Does anyone know what purpose those black pads might have? As far as I know the AT8 was also used in the BC3 Console and maybe a couple more. Is anybody familiar with the circuit and could help me fill in the blank resistor values?
Cheers!

 

[Khron]

https://en.wikipedia.org/wiki/Resistor#Printed_carbon_resistors
You could always, y'know, try measuring those resistors, between the opamp pins they're between? Don't poke the carbon resistors themselves, you don't wanna risk damaging them...

https://sound-au.com/balance.htm

 

[krabbencutter]

https://en.wikipedia.org/wiki/Resistor#Printed_carbon_resistors
You could always, y'know, try measuring those resistors, between the opamp pins they're between? Don't poke the carbon resistors themselves, you don't wanna risk damaging them...

https://sound-au.com/balance.htm

TIL that the black pads are actually resistors. It's interesting that they chose to print them, since I'd assume you want very tight tolerances in a balanced line driver and the printed metodology can vary by up to 30% (according to the wikipedia article you linked).

 

[Newmarket]

TIL that the black pads are actually resistors. It's interesting that they chose to print them, since I'd assume you want very tight tolerances in a balanced line driver and the printed metodology can vary by up to 30% (according to the wikipedia article you linked).

I think those printed pads have been trimmed manually. See what appear to be cuts into them. Similar to trimming of printed conductive ink in fader manufacture. But why does the NE5532D have sixteen pins. Was / Is that an alternative package ?

 

[krabbencutter]

I think those printed pads have been trimmed manually. See what appear to be cuts into them. Similar to trimming of printed conductive ink in fader manufacture. But why does the NE5532D have sixteen pins. Was / Is that an alternative package ?

oh, that could indeed be the case! I'll make sure to measure them next week and maybe they'll indeed be close tolerance.
The N5532 is from Signetics with a non-standard pinout and I found a datasheet online: http://www.elektronikjk.com/elementy_czynne/IC/SE5532A.pdf It's also the first time I've encountered a dual op-amp in a 16pin package as well.

 

[krabbencutter]

I've measured the resistive pads and they are indeed very accurately trimmed. The only outlier was the 3k7 resistor, which measured somewhere around 3.669k. I'll confirm the values with my second AT8 PCB next week. I've also had trouble measuring the capacitors. The feedback caps were too small for my BM235 Multimeter to accurately measure while the compensation caps measured sometimes ~620nF and sometimes ~900nF. Will have to measure them properly when I'm back at the bench.

 

[krabbencutter]

I've measured and calculated the missing capacitor values with the help of a relaxation oscillator. I've also redrawn the circuit in KiCAD and included the external components, so that it can be used as a balanced output in any project. So if someone ever needs a Rupert Neve designed output stage, here you go

 

[krabbencutter]

Now that the NE5534 are EOL I've replaced them with another NE5532. Will need to test it in circuit, but I'm guessing it should be okay?

 

[DarkSky]

Now that the NE5534 are EOL I've replaced them with another NE5532. Will need to test it in circuit, but I'm guessing it should be okay?

Ouch - I wasn't previously aware that the NE5534s are EOL!

Seems like the end of an era ... I've been designing things around them since the mid '80s, so they feel like old friends - of a sort.

Do you know if there an 'official' alternative that will 'replace' them, or are we just going to have to run with existing options (such as the NE5532 etc)?

 

[krabbencutter]

Yeah, I just recently got a notification from mouser and unfortunately it's the whole series including SMD variants :/ As a replacement for PDIP TI recommends the OPA134, which unfortunately costs 2.5x than the NE5534:
OPA134

As a replacement for the SOIC TI lists the OPA1677 which looks like an interesting IC to keep an eye on: OPA1677
It's a fairly new part, has rail-to-rail output, is qualified up to +/-18V, has a 6MHz wider GBW, only needs 2mA of supply current and has only 10 pA input bias current! Slew rate & Vn are more in line with the NE5532 tho. However it is surprisingly cheap and even comes in dual & quad variants! This could actually be my new go-to audio opamp for SMD designs

 

[DarkSky]

Yeah, I just recently got a notification from mouser and unfortunately it's the whole series including SMD variants :/ As a replacement for PDIP TI recommends the OPA134, which unfortunately costs 2.5x than the NE5534:
OPA134

As a replacement for the SOIC TI lists the OPA1677 which looks like an interesting IC to keep an eye on: OPA1677
It's a fairly new part, has rail-to-rail output, is qualified up to +/-18V, has a 6MHz wider GBW, only needs 2mA of supply current and has only 10 pA input bias current! Slew rate & Vn are more in line with the NE5532 tho. However it is surprisingly cheap and even comes in dual & quad variants! This could actually be my new go-to audio opamp for SMD designs

Very interesting - thanks for sharing that!

I agree that the OPA1677 seems to have some promise. It will be interesting to hear how it sounds. A couple of the stand-out specs (other than those you mention) are that the quoted THD is 1/20th of that of the NE5534, and (FWIW) the operating temp range is more than double.

Along with the fact that it appears to have been developed specifically for audio applications it seems promising.

Meanwhile, price aside, the OPA134 looks even more impressive (well, as far as the specs that might matter go). Ignoring the fact that it will only operate at up to 85º C (as opposed to 125º C for the OPA1677). The FET architecture and "Burr-Brown™ Audio, Premium Sound" might even be worth paying a bit extra for!

 

[limo]

I have two 16ch frame AMEK BC3.
One has been perfectly maintained by me and is in good working order. The other one is sleeping right now.
I don't have numbers written in my circuit diagram either, so this topic is very helpful.
The two BC3s are from different years, so their specifications differ in some places. I have 3 types of AT8.

The first one is the 1991 Signetics NE5532, which has the same parts layout as the Krabbencutter one, but the small IC model number is different. I looked up about this IC but couldn't find anything. I was able to find out from your photo that it is an operational amplifier compatible with NE5534. thank you.

The second one is the 1995 Philips NE5532, with a simpler circuit. I don't have the schematic for this version.

The third one, from 1994, is the Signetics NE5532, which has a simple configuration like the Philips, but it is on a large board and has an analog switch on the back.
The connection is a 16 pin socket, but it is compatible with those without a large board.

Signetics=Philips' NE553x has different sound quality depending on the year and lot, so the sound quality of each one is different.
My impression is that the 1992 Sig has a close to vintage sound and is suitable for microphone input.
The 1995 Phil has a more modern sound and is suitable for LINE INPUT and mixed music sources.
Now, Signetics cards are running on my BC3.
This one has a sound that you will never get tired of and is fun to use.

 

[krabbencutter]

Very interesting, thank you for sharing those details! I'm completely geeking out right now and the longer I look at those PCBs the more questions I have. It's also interesting that the resistive pads are blue. I'm wondering whether this was just added pigments, to make it more distinguishable, or if it's some protective coating. And if it's the latter, how did they trim down the pads to the resistance values they needed? Gosh, I wish I could talk to the designers about these decisions
The mystery IC is weird Maybe it's an obscure NE5534-ish opamp variant from signetics? Or maybe they manufactured their own ICs for a while? There's some custom ASICs on the Mozart that control the switching logic, so they definitely had some sort of fabrication capacity.

 

[Newmarket]

Very interesting - thanks for sharing that!

I agree that the OPA1677 seems to have some promise. It will be interesting to hear how it sounds. A couple of the stand-out specs (other than those you mention) are that the quoted THD is 1/20th of that of the NE5534, and (FWIW) the operating temp range is more than double.

Along with the fact that it appears to have been developed specifically for audio applications it seems promising.

Meanwhile, price aside, the OPA134 looks even more impressive (well, as far as the specs that might matter go). Ignoring the fact that it will only operate at up to 85º C (as opposed to 125º C for the OPA1677). The FET architecture and "Burr-Brown™ Audio, Premium Sound" might even be worth paying a bit extra for!

OPA134 is not a great substitute for NE5534 if you are using 5534 for its low voltage noise EIN.. 8nV/rt. Hz Vs 3.5

 

[DarkSky]

Coincidentally, I had some of our Tonelux rack units on the bench today, and noticed that the units I was looking at were making use of a series of OPA2134 op amps, which (as the designation suggests) are a 2-channel variant of the OPA134.

The Tonelux circuits in question are EQs and the op amps appear to be providing buffering and makeup gain. I've always thought these units have a very light, clean and open sound, so count that as a solid vote in favour of OPAx134 chips - in appropriate implementaitons!

 

[Johnny1234]

Laser etched precision conductive plastic resistors. P & G faders used that technique also.

 

[krabbencutter]

I started working on a PCB for this. Since this is the first time I'm doing any PCB layout I'd be happy if someone might have some feedback on it Power traces are 20mil and all other traces are 10mil. The bottom layer will be a ground plane.

 

[Khron]

Since this is the first time I'm doing any PCB layout I'd be happy if someone might have some feedback on it

Are all the components on the board already? If not, place them all there first. If yes, you might wanna hold off on the trace routing until the component placement is (arguably) as good as you can get it (shortest feasible connections etc).

Trace width wise, it's usually best to go with "as thick as you can afford". With 0603 and bigger parts, i go with at least 0.6mm (or whatever that translates into, in mils) for signals, and even wider for power (where necessary; wouldn't do it in a phantom-powered mic).

 

[Tubetec]

Be interesting to hear from Matt about the AT8 variants ,
Ive two 24ch bays of BC3 ,
without the VCA's ,
Mine has the kind on the right in the first picture above ,
BC315 and 318 modules as far as memory serves ,

 

[krabbencutter]

Are all the components on the board already? If not, place them all there first. If yes, you might wanna hold off on the trace routing until the component placement is (arguably) as good as you can get it (shortest feasible connections etc).

Trace width wise, it's usually best to go with "as thick as you can afford". With 0603 and bigger parts, i go with at least 0.6mm (or whatever that translates into, in mils) for signals, and even wider for power (where necessary; wouldn't do it in a phantom-powered mic).

Yep, components are all on the board and the resistors & ceramic capacitors are all 1206. Is there a technical reason for your wide traces or is it personal preference? The pitch of the OPA1679 is 1,27mm (50mil) so if I go for 0,635mm (25mil) traces I would have very little separation between signal traces and pads.