Towards a better sounding discrete preamp / buffer for Hypex NC1200 / 500 modules

[Art Vandelay]

Having recently acquired hypex mono amplifiers based on NC500 and NC1200 modules I've now started work on a discrete class A buffer / preamp that's capable of easily driving the unbuffered Hypex inputs, either unbalanced or balanced.  My motivation is to avoid using the typical all op-amp based design as outlined in the Hypex application notes. Also, by locating the preamp externally and with its own dedicated power supply there's less potential for EMI related issues that can degrade performance.    

 

Output requirements are 15 vrms @ 3mA with low THD (<0.001%), and with sufficient voltage gain such that a 2vrms source can drive the module to full output. Via unbuffered inputs the Hypex modules are a low gain stage (11.6dB), hence require an additional 10-15dB of gain to be compatible with typical 2 vrms sources. 

 

My fist prototype will actually use a single op-amp (opa2132/4) followed by a buffer / gain stage configured with approx 8dB of voltage gain. In keeping with the "kiss" principle the output stage is relative simple and I've decided to go single ended rather than complementary push-pull. In simulations the open loop distortion is inherently low (0.01%) and almost entirely second order, and closed loop is pretty much non-existent with a low distortion op-amp. 

 

Initially, I'll drive the NC1200's in unbalanced mode, but down the track the preamp will have 2x stereo boards for balanced capability as well, and maybe even a quad gang attenuator.    

 

 

 

  

[acg]

Looks like a good project Art. 

 

Personally, I'd at least try to do it with tubes...mainly because of the contradiction of topology and technology and just to see.  And because I like tubes.  And big things.  And, well, there really are no compelling technical reasons to be honest.  The voltage gain and current drive sides of your requirements are fine, but I doubt anything better than 0.05% THD could be achieved with a single stage valve design.  Imagine a thoriated tungsten filament 801a DHT,  a hybrid mu-follower and an autoformer for volume control:  gain is 8x, current 30mA or so; Zout is about 1.6R at -3dB and much lower with more attenuation; lots of unnecessary complexity and size (but it would glow!); stupidly expensive to do well.  Only fly in the ointment is that a decent autoformer is generally only good for about 8Vrms @ 20Hz input and may or may not be better in this case at input rather than output.  Come to think about it, incorporate an input selector and you have a complete preamp/amp with tech at each end of the spectrum.

 

Sorry, I'm externalising my angst...I really need to find time to build that pre-amp for myself...carry on. 

 

I'm am interested to see what you come up with having taken partial notice of the various attempts of others elsewhere.

 

Anthony

[Art Vandelay]

5 minutes ago, acg said:

 

 

Sorry, I'm externalising my angst...I really need to find time to build that pre-amp for myself...carry on. 

 

I'm am interested to see what you come up with having taken partial notice of the various attempts of others elsewhere.

 

Anthony

lol.

 

Thanks Anthony. You could drive speakers with what you're proposing, but that definitely sounds like a project for a covid-19 induced sabbatical, and no doubt of interest to a large audience.

  

I'll post periodically as I progress to the point of build and testing etc. I hope the finished product will be worth the effort but if it isn't it's at least given me something to do in the weeks and months ahead.  

    

 

 

 

 

[mwhouston]

I assembled on Vera-brd this discrete ESP transistors preamp for another. He boxed it and made the PS. He claimed it was really excellent from only off the shelf components. Wouldn’t mind building another.

 

https://sound-au.com/project37.htm

 

http://mongreldogaudio.blogspot.com/2010/10/esp-p37a-distcrete-component-class.html

 

Mongreldog is an old blog of mine. 

[Decky]

12 hours ago, mwhouston said:

I assembled on Vera-brd this discrete ESP transistors preamp for another. He boxed it and made the PS. He claimed it was really excellent from only off the shelf components. Wouldn’t mind building another.

+1 for ESP DoZ (Death of Zen) preamp. You will be surprised. I sold my beefed up, full blown, no corners cut, build on SNA two years ago but I should have another assembled PCB in couple of months time. This is in case esp is out of the PCBs or if you feel like being to lazy to build one yourself. I will be happy to give you the main PCB for a test run if you like. 

 

This was my build 

 

 

Edited April 7, 2020 by Decky

[aussievintage]

13 hours ago, mwhouston said:

I assembled on Vera-brd this discrete ESP transistors preamp for another. He boxed it and made the PS. He claimed it was really excellent from only off the shelf components. Wouldn’t mind building another.

 

https://sound-au.com/project37.htm

 

 

I was just looking at this.  There were a couple of things about the circuit I was not happy about, but there revised one here  https://sound-au.com/project37a.htm  is better.  Interesting simple preamp.  Thanks for posting the link.

[robmid]

16 hours ago, acg said:

Looks like a good project Art. 

 

Personally, I'd at least try to do it with tubes...mainly because of the contradiction of topology and technology and just to see.  And because I like tubes.  And big things.  And, well, there really are no compelling technical reasons to be honest.  The voltage gain and current drive sides of your requirements are fine, but I doubt anything better than 0.05% THD could be achieved with a single stage valve design.  Imagine a thoriated tungsten filament 801a DHT,  a hybrid mu-follower and an autoformer for volume control:  gain is 8x, current 30mA or so; Zout is about 1.6R at -3dB and much lower with more attenuation; lots of unnecessary complexity and size (but it would glow!); stupidly expensive to do well.  Only fly in the ointment is that a decent autoformer is generally only good for about 8Vrms @ 20Hz input and may or may not be better in this case at input rather than output.  Come to think about it, incorporate an input selector and you have a complete preamp/amp with tech at each end of the spectrum.

 

Sorry, I'm externalising my angst...I really need to find time to build that pre-amp for myself...carry on. 

 

I'm am interested to see what you come up with having taken partial notice of the various attempts of others elsewhere.

 

Anthony

Hi Anthony.

 

For the last 2 years, I have been refining Bi-Amped NC122MP modules with buffering and LDR attenuator, to drive 505 Mini ESL panels.

 

It could be a successful solution for less technically inclined DIYers who have access to Hypex amps and uses an integrated passive 3 input LDR passive preamp with active analog X-kitz crossover to provide a pristine signal to the Hypexes. The kits require very little soldering as most connections use screw blocks. Total cost was around $500 for parts.

 

I have had pleasing success with LDR attenuation over 2 years, due to the silence that accompanies when music is not playing, and presenting with music, the best possible audio experience. (Details on my LDR build and others) https://robmid42.wixsite.com/diyaudiokits/stereo-coffee-preamplifier

 

Cheers,

Rob

 

 

X-kitz crossover boards mounted vertically and LDR boards mounted flat

 

[acg]

3 hours ago, robmid said:

Hi Anthony.

 

For the last 2 years, I have been refining Bi-Amped NC122MP modules with buffering and LDR attenuator, to drive 505 Mini ESL panels.

 

It could be a successful solution for less technically inclined DIYers who have access to Hypex amps and uses an integrated passive 3 input LDR passive preamp with active analog X-kitz crossover to provide a pristine signal to the Hypexes. The kits require very little soldering as most connections use screw blocks. Total cost was around $500 for parts.

 

I have had pleasing success with LDR attenuation over 2 years, due to the silence that accompanies when music is not playing, and presenting with music, the best possible audio experience. (Details on my LDR build and others) https://robmid42.wixsite.com/diyaudiokits/stereo-coffee-preamplifier

 

Cheers,

Rob

 

 

X-kitz crossover boards mounted vertically and LDR boards mounted flat

 

 

Hey Rob,

 

I'm not using Hypex anything in my system, but do need a preamp with gain and real current drive to manage my six channel SET amps with passive line level crossover filters for all six channels at the amp input.  It is a tough load, circa 5K, with inductors, caps et cetera making it NOT an easy resistive load.  So no passive preamp for me, only active, and something that can turn 1v from my sources into 4v+ so I can clip the amps, and given some nice "quiet" recordings I am going to need maybe up to 8x gain to make things really loud.  I have a handful of SS active preamps here that vary from acceptible to not acceptible in use, but nothing really exceptional, and philosophically I would like complete hollow state from source to speaker:  2 stage valve LCR phonostage >> DHT preamp with autoformer volume control >> DHT/IDHT SET amplifiers.  The DHT preamp that I intimated to earlier is really the only shot I think I have for valves to work in my particular circumstances, and the parts are pretty much all here to get started...just have to design up a chassis and get to building.

 

But this is Art's thread, about a driver stage for a Hypex amp module.  I'll stay tuned in here to see what he comes up with...

 

Anthony

 

Hey @Art Vandelay, are these amps for fullrange or just the bass?  Or not sure?

[Art Vandelay]

6 minutes ago, acg said:

 

 

Hey @Art Vandelay, are these amps for fullrange or just the bass?  Or not sure?

 

TBD. At the moment I'm using Hypex only for the woofers, with a class AB driving the mids tweeters (B&W800 d2).

 

But an objective is to see (hear) how they sound when driven direct (without the usual op-amp based balanced driver). This could also apply to the now available Purifi modules. 

 

A side effect of balanced drive is that the even order harmonics are neatly cancelled, leaving the odd order products which can result in a dry,  clinical sounding treble, so my DIY preamp will allow me to drive the modules "unbalanced" to full output, and ultimately no op-amps in the signal path.