Jensen Twin Servo 990 - Design Ideas

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

mrvision

Well-known member
Joined
Jul 18, 2009
Messages
183
Location
USA
Hi guys,

I have been working in the past days in prototyping a Jensen Twin Servo Preamp based in the original Jensen Design:

http://www.jensen-transformers.com/as/as083.pdf

So far everything works and the sound is awesome. Warm, full, 3D and punchy.

So here is the problem: I would like to add a switch for 1 or 2 stage gain. Putting a simple true bypass switch between the output from the 1st Stage to the output transformer will produce the expected loud "pop" due to DC offset changes, capacitances, etc. The design must be free of capacitors in the audio path so a coupling capacitor between stages is not an option (even though I tried a capacitor to block the DC and popping was minimun... but there was color on the sound and sound quality was not the same) There are some conventional and popular fixes like bleed resistors and capacitors... the ones people do in guitar amps, but I think were are talking here about something a little different because we are talking about Discrete Opamps, DC voltages present sometimes in the audio path and a servo DC Offset servo compensation circuit.

The goal is a clean switching between 1 stage and 2 stage preamp.

Any of you people with experience in Op Amps circuits have a sugestion?

Picture:

388483101_o.jpg
 
Any chance you might share the etch files?  I've been looking for a basic twin servo pcb for a long time.
 
Yes, there are servos in the circuit. I am using LM11CN on both stages.

I regards to the etching files, I will probably make some PCB's available for the Prodigy community once the Preamp is fully tweaked. I have been also looking for a PCB to build such a legendary Preamp. I found some clones out there but the prices are ridiculous... to much for a DIY guy  ;D
 
Nice project. Definitely interested in pcb's if you're making a small batch.

Looking at the schematic I see there are two servo's right?
Are these to get rid of any DC offset which might still be there even though there are transformers to block the DC?
Especially the first servo seems a bit overkill. The second one might be there to get rid of the DC offset from the first 990 opmamp.

I'm not questioning this BTW. Just trying to understand the circuit.
 
radiance said:
Especially the first servo seems a bit overkill. The second one might be there to get rid of the DC offset from the first 990 opmamp.

The second servo is there to rid of the DC offset of the second stage alone. The servos are not "cross-linked" to the stages is any way.

The first stage will absolutely need a servo (or a cap) to remove any DC offset as well. The offset would otherwise severely impair the headroom of the second stage (asymmetric clipping). Think of a servo as an active DC coupling cap. It just replaces one BIG cap with an opamp and a smaller cap (470nF in the twin-servo design). Which both are in the signal path by the way. The selection of the two is somewhat critical. Otherwise you would decimate the benefits a great sounding discrete gain stage may have.
 
Kingston said:
The first stage will absolutely need a servo (or a cap) to remove any DC offset as well.

But a transformer does exactly that (and more), getting rid of DC offset. Or am I missing something?
 
radiance said:
But a transformer does exactly that (and more), getting rid of DC offset. Or am I missing something?

Yes.

A gapped transformer can handle some amounts of DC, but none of transformers used in with twin-servos in general can handle DC. They will clip (and play weird in other ways) just as easily as an opamp.

The first discrete opamp will produce a significant DC offset.  It needs to be removed, because it would impair the second stage (might not even work, oscillation, or a nasty clip, depending on the amount of DC). Standard choices are a cap, or a servo. The second discrete stage will also produce DC offset. The output transformer can't handle any, and it needs to be removed.
 
mrvision said:
The goal is a clean switching between 1 stage and 2 stage preamp.

Any of you people with experience in Op Amps circuits have a sugestion?

Jfet's ? They can switch with a very fast ramp, a very fats fade in ...
 
Use switch to replace first stage gain pot with resistors set to unity gain? I don't fully understand the circuit but this came to my mind :)
 
Well said Kingston... you have explained the servo circuit really well...  I like the basic explanation: "Think of a servo as an active DC coupling cap" That is an easy way to understand the concept.  ;D

I also thought about putting some FET's configured as switches in the Audio Path but then I also thought about distortion and then we have to provide an additional flip-flop circuit to turn the electronic bypass on and off complicating more the design.

I had not thought about the idea of using the switch to bypass the Potentiometer and connect it to a net to set unity gain on the first stage instead of switching the whole stage. It sounds like a good idea, I may give it a try. Thanks tmuikku.

Thank you guys for all your contribution to this post. I am looking forward to make this project available to everyone in a near future.

I have posted a Picture of the Prototype board on the first post ...still going under a lot of revisions, just to give you an idea about the design.

The PCB is intended to be munted in a 1U Rack case, it uses lightpipe buttons, it has +48V, Impedance selector, Pad, Phase, Gain Stage Switch, Gain Control, and I am thinking about adding also a DI/Instrument section using an extra jensen JT-DB-E.

VU? ...maybe I am looking for a good VU circuit with peak hold falling capabilities. Anyone has a good circuit in their collection?

 
Ideally speaking there is no DC present, but due to changes in temperature,changes in the gain control, noise, and other factors, the DC fluctuates temporarily and the servo starts acting to compensate the DC detected at the DOA output. In my particular circuit under test, once the servo adjustment has settled, the DC present on the output of the DOA's stays in the first stage between 0.0 - 0.1mV and in the second stage between 0.0-4.2mV. That is very good considering the fact that an offset voltge is considered "bad" by some people from 5mV-8mV and up, at least for audio applications. My circuit under test is not inside a grounded case yet BTW, there is no oscilations thanks to the ground plane on the PCB, but I am sure electrical noise is all over it trying to find a way on the circuit. The Jensen OLI-3 inductors are still pending, they are in the outputs to help control unwanted noise also.


Also, my circuit is not under ideal conditions right now, one of the DOA's was really noisy and unstable so I used a FET version of a 990 that I had on my bench, and I have read that FET DOA's have more problems with DC offset. i am not sure if that is why the DC offset is a little higher on the 2nd stage. A matched pair of DOA's is better for more consistency. ptownkid's 990's sound more warm, full and 3D in this particular design, APP99F will do a good job in API circuits.

The circuit as it is in the "basic" Jensen Twin Servo Schematic has a higher DC offset, I does not produce an effect that can be perceived by human ear and it can probably only seen with test bench equipment. Still, I have added some DC compensation circuits to the DOA inputs to have better control over offset voltages and to bring DC offset close to ideal.
 
APP99F has a low input offset voltage, often around 2-3mV . Its front-end is done using two mached 2SK170 Toshiba ultra low noise fets.
I've designed the APP99F to work in API312/512 pres, since the JE-990 is not for high ratio input transformers and several people don't love it in these pres.
Infact, customers said me APP99F is better than JE-990 in API312/512 pres.
If JE-990 was always better than APP99F I would stopped its production.
However, APP99F is very similar to JE-990 on its topology.

APP99F does not need the OLI on its output because it has more stability than JE-990.

I council you to try APP99F on front-end stage and the JE-990 in the second one.

kind regards
Pier Paolo

 
 
Sorry Pier, I didnt mean to say that your APP99F was a bad DOA, As a matter of fact I believe is a really good opamp, but in this particular design, I think that DOA's with more similar design and characteristics than the legendary JE990 seem to work better.

I am actually keeping the APP99F's for a future API project I have on the shelf. In this case I grabbed one of the APP99F just to temporarily replace one of Ptownkid's 990's
 
Not sure if this is completely stupid question, but has anyone ever considered a DOA as the servo opamp as well?

Some might say it's wasteful and stupid, but I would be very interested in testing something like that. I was easily able to hear a difference between OPA604 vs. LME49710 as servo in another design, for example...
 
isn't a stupid question, the only problem is the higher input offset voltage and/or currents of the DOA's than the IC's, but using a trimmer we can use the DOA's for servos. In my opinion a servo based on DOA sound generally better than an IC servo. It's not a rule, because this depend of servo topology, DOA and the IC used, the pre topology ....etc....
But however, we can have a good IC servo using an IC opamp.
Jensen twin servo doesn't use an "invisible" topology for the servos, even if it is for the 50% "invisible".
 
 
regarding the servos in the pre I coulcil to test the IC opamp's output and to see if these outs are saturated. If one or both the outs are saturated than two or one resistors should be modified to have the best reduction of DC on the DOA's out concerning the servo saturated.
- if in the first stage the servo's opamp is saturated than R9 must be reduced until the servo's opamp is not saturated with good margins.
- if in the second stage the servo's opamp is saturated than R15 must be reduced until the servo's opamp is not saturated with good margins.
However setting r9=6.8k and r15=100k is a good way to use several DOA's in the market, comsidering that this pre is designed around 990 so it migh not run with other DOA's.
 

Latest posts

Back
Top