Preamp Project: Samuel Groner's Shared Gain Preamp

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

rodabod

Well-known member
Joined
May 12, 2005
Messages
2,896
Location
London
I've been thinking about making a thread about this for ages. This is because I am so slow at getting this finished.

Anyway, as a brief introduction, this project stems from when I asked about what to do with a Klein and Hummel preamp board which had a Beyer input transformer. As it turned out, it looked best to forget the preamp board and just use the transformer in a better-performing circuit. After a few lab members offered suggestions, Samuel offered to adapt a schematic of his which would be suited for the application.

The design is a "shared-gain" preamp using a single linear pot to control gain for both opamps with transformer-balanced inputs, and unbalanced outputs. The design can obviously be easily adapted for particular needs.

sharedgainpreampr12ny.gif


So first things first, I order in most of the components to make a 1u rack of four preamps. It is at this point that you remember just how many bits you need to make a simple preamp!

sgp1pic11ap.jpg


That's some of the main components - most notably; beadboard, 4X beyer trannies, 4X OPA604s, 4X NE5534s, bipolar electrolytics, SSL Tech PSU board, toroidal tranny..... You get the idea. The silver box is my original prototype which has served me very well.

Other than that, there is my hardware - eg. rack case, knobs, stand-offs, zip ties, blah, blah.

Now, when doing DIY, try not to get ahead of yourself! I had fun messing about with Jaydee's and Frank from NRG Recording's .fpd files and printed out a lovely paper front panel..... But wait - there isn't any circuitry to go behind it :razz:

sgp1pic29ic.jpg


Now on to the fun part (this is where all you fake-ass pre-fab PCB users should go to the brewery :razz: ) - breadboarding!

Here is the first channel in place with IC sockets ready for plugging in trannies and ICs :

sgp1pic31bb.jpg


This worked straight away, but it had very little headroom and sounded distorted unless it was at very low levels... I wondered if my NE5534P opamps would be to blame (I thought the suffix P meant higher temperature though...) so I swapped them for OPA604s and right enough, it worked perfectly! I wonder what is up with my (free) NE5534s? I am running on bvatteries at +/- 9V for the meantime.

Here's a close-up:

sgp1pic42be.jpg



I will update this thread as I go along. The plan just now is to populate the rest of the board so I have have four working channels...
 
I like this :thumb:
Did you make any planning or did you just go "free form" on the breadboard?
I also see you don't use the "lanes" from the breadboard. Is it all point to point underneat?
 
[quote author="radiance"]I like this :thumb:
Did you make any planning or did you just go "free form" on the breadboard?
I also see you don't use the "lanes" from the breadboard. Is it all point to point underneat?[/quote]

This is a cool design, because there is scope for doing anything you want with it - you can add or take away bits and easily customise it to your needs since it is so simple.

I did plan my breadboard - this is actually quite important as fault-finding can be difficult. It's a little bit like hand-drawing a schematic - you draw the main blocks (ie. transformer and opamps) and then fill in the remaining space with the smaller components.

I do use the lanes on the breadboard. I just cut tracks and use jumper wires on the board to get the flow correct. There is no point-to-point underneath.

Roddy
 
Aah, no I see. The lanes are also silk screened (or something like that) on top...
I like planning breadboard. If you think long enough you always find a way to make it more compact (although I see that's not really your goal :grin: )...
When I plan a breadboard circuit I use this huge paper with a milli meter scale printed on it.....
 
[quote author="radiance"]
I like planning breadboard. If you think long enough you always find a way to make it more compact (although I see that's not really your goal :grin: )...
[/quote]

What are you trying to say? :evil:

Only kidding! :razz:

I deliberately give one or two holes space between each component so that I can try any mods or add components later. Also, I bought a fixed size of board and simply divided it into four for each channel.

[quote author="Kit"]Looking good, Roddy. :thumb:

Strange about them 5534´s though. :?[/quote]

Thanks mate!

The 5534s is a bit strange... I will need to look at the datasheet for the NE5534P.


By the way, I have finished the whole board tonight, so I will post photos tomorrow.

Roddy
 
hey roddy,

looks like real neat and tidy work!
strange thing, the first thing i´ve done with the beyerdynamics from kubi was trying to plug them into the ic sockets from a ssl comp! but i can see that you didnt ground the can, maybe you should do that.
by the way, how about your beyerdynamic mics...did you come around recording something with them?
-max
 
[quote author="ioaudio"]
looks like real neat and tidy work!
strange thing, the first thing i´ve done with the beyerdynamics from kubi was trying to plug them into the ic sockets from a ssl comp! but i can see that you didnt ground the can, maybe you should do that.
by the way, how about your beyerdynamic mics...did you come around recording something with them?
[/quote]

Cheers, Max. I think these transformers suit ribbon mics well as ribbons have quite low outputs. I am going to add four JLM go-betweens to add a 20dB pad for condensers and phantom and phase reverse too. You are correct about me not grounding the can - I need to do this. Also, there is a pin for "static shield" - should I ground this too? I guess I should.

I recorded for the first time last week with one of the Beyer CK70* condensers. I had one on cello which I replaced with a Beyer M260 as it sounded smoother on the high notes and was more natural. But, I had one on overhead for a Jazz drum kit along with a Beyer M88 on kick which was great - the CK70* has a nice high end, and seemed to sound a little compressed on the drums (I didn't use the -10dB pad) whichI guess was maybe the transformer saturating. Have you used yours?

[quote author="Samuel Groner"]
Did you add a compensation cap (22 pF) to the NE5534? Otherwise it will oscillate once in a while...
[/quote]

Hi Samuel, yes, I have 22pF caps in the feedback path of both opamps.... Strange. I will need to try the NE5534 again to see what happens. Old Signetics ones worked fine in my prototype model.

Here is where I got to last night - the board is almost complete:

dscn14916zs.jpg
 
[quote author="Samuel Groner"]
Yes, I have 22pF caps in the feedback path of both opamps...
Not in the feedback path, between pin 5 and 8. :wink: [/quote]

Between pins 5 and 8?

But are they not left as "not connected" usually?

I thought we just wanted to reduce gain at high frequencies by adding a small cap in the feedback path.... What does connecting a cap between pins 5 and 8 do?

Thanks,

Roddy
 
Roddy,

It looks BEAUTIFUL! Well done my friend! The panel looks great as well, are you planning on using a JLM Go Between in front of the pre, or will you wire up the switches yourself off board.

Well done again, it looks great and your perf board layouts looks a HECK of a lot better than mine... I now have a project board to try and play with first before being gungho about using perf board.

Oh and your camera takes great shots :)

Cheers

Matt
 
What does connecting a cap between pins 5 and 8 do?
Without the compensation cap the NE5534 is not stable at gains below 3 (about 10 dB). By increasing the capacity you reduce the open-loop gain of the amplifier which makes it stable at some point for a given closed-loop gain. The main drawback is probably that the slew-rate comes down as well.

A detailed analysis of the topic is somewhat lengthy, but for the moment: just add a 22 pF (likely a 15 pF does the trick) between pin 5 and 8, and we're happy!

It is surprising that the newer brands are less stable than the Signetics parts, but that's live...

Samuel
 
[quote author="matta"]
...are you planning on using a JLM Go Between in front of the pre, or will you wire up the switches yourself off board.
[/quote]

Cheers Matt,

I'll be using the standard Go-Betweens on the front panel. I was wanting to use push-buttons mounted behind the panel (just the buttons sticking through) but this is not an option since they won't fit the board.

I think you've had some great photo shots too - these ones are just rushed whilst I was soldering. Also, they have heavily jpeg compressed which gets nasty on the high frequency detail (eg. wood grain and sharp edges...) As always, all you need is your flash switched off, a steady hand and the macro for the close-ups.

[quote author="Samuel"]A detailed analysis of the topic is somewhat lengthy, but for the moment: just add a 22 pF (likely a 15 pF does the trick) between pin 5 and 8, and we're happy! [/quote]

I didn't know that the NE5534 had compensation pins for both pins 5 and 8 - I was just referring to my 741 pinout! I understand using the compensation pin for cancelling offset, so I can see how this could be a solution...

Thanks again,

Roddy
 
[quote author="rodabod"]This is a cool design, because there is scope for doing anything you want with it - you can add or take away bits and easily customise it to your needs since it is so simple.[/quote]

excellent
opamps when used correctly and in sensible topologies can work just fine.
see
F100, Neve-IC-pre, Amek, JLM-Di and the 99v mother board with IC's ... see the Rane stuff
and you can also find ideas on HiPass filters and simple EQ's etc

lots of good out there
 
[quote author="Samuel Groner"]Just found this thread while updating the MicPre META--did you finish this design already, Roddy? I'm curious how you like it!
[/quote]

Hi Samuel,

It is still under construction! I've just got back from Berlin on holiday, and I will still be very busy when I get back, but don't worry, I will get finished eventually!

I've tried one channel so far with the Beyer 1:7 and really like it - it is very quiet, and sounds more musical than my other electronically-balanced mic pres.

I will update soon!

Roddy
 
Ok,

My parents are away on holiday this week, so I have turned the living room into a workshop. :twisted:

So, I've fired-up one channel of the preamp in the rack case using Keef's 18V PSU as shown below:

sgpopenrackis0.jpg



Now, in the top two thirds of gain it seems to work very well. Even at max. gain with a 200 Ohm mic, it remains very quiet. :thumb:


However........ In the lower third of gain it seems to suffer a "buzzy" 50Hz hum. I can't seem to figure out why. The 50Hz hum is not being "masked" by noise at the higher gain settings. It is definitely louder at lower gains and stays around the same level.

I checked for grounding problems by lifting the ground on the XLR out but this did not help.

Any ideas?

Thanks,

Roddy
 
Ok, now I have been fault-finding for a while, and have sorted a few problems regarding the grounding (my laptop's PSU has to be unplugged as the ground is filthy).

Anyway, my problems now seem different.... :?

Now I am getting clean and quiet gain for the first half, and then fairly suddenly I start getting hideous noise and interference sounds, and the audio gain seems to go down.

This is with OPA604 opamps. With NE5534s I get no audio at all :shock: - even if I try adding compensation caps.

At first I wondered if maybe the opamps were unstable, but the problem only occurs at higher gains.

Here is an MP3 recording of me speaking into a mic through the preamp whilst turning the gain pot up:


http://myweb.tiscali.co.uk/rodabod/SGPNoise2.mp3


Does anyone have any suggestions?

Thanks,

Roddy
 
That's oscillation (yeah, I heard that before!). It happens at higher gains because somehow a signal from the output couples to the input. And you'll get a similar effect as you know it from your PA gigs if you crank up the monitors loud enough (just that it happens at much higher frequency).

It's all about routing and grounding. Keep the input from the output, carefully check and re-route the ground pathes. Putting things into a grounded and shielded case usually helps a good deal as well. Increasing C4 a bit might help too. Do you have a proper PSU? If there's a lousy ground, such stuff is almost intractable.

No idea what's wrong with the NE5534s.

Samuel
 

Latest posts

Back
Top