Questions for API 312-esque circuits - gurus??

[tk@halmi]

[quote author="fallout"]Has anyone built his PCBs yet? I was wondering how they sound.. Might wanna build a couple..

-Jay[/quote]

I have populated a board that I bought from Joe with a JT-110K-HPC, a JFET-992 and an OPA2604 for the servo. All I can say that it works and in this configuration sounds well. The board itself is excellent quality and it is very flexible.
This flexibility does demand some planning on your part to decide on what configuration you want to build.
The relays that fit on the board are ten pin miniature ones, smaller than the usual DIP relays. They can be had at Digikey: part# Z740-ND.

Tamas

 

[Flatpicker]

[quote author="Bauman"]...I can post the untested version in a couple of days if you like.[/quote]That'd be great, Fabio.

I'm working on a version like that too. The thing that seems to take the most time is making all the special component footprints.

 

[Flatpicker]

[quote author="TomWaterman"]I wonder if Tim from 7th circel knows about this board...maybe he could shed some light???[/quote]I'm sure he could since he probably designed it. He used to post over at TT, but I wonder if he even knows about our new home. Drop him an email with our web address and ask him to stop by.

 

[Kev]

JLM audio
http://www.jlmaudio.com/JLM99V.htm

as far as I know Joe is working on a new web page to help explain the options for this one. It shouldn't be too much longer.
singles, duals, servo, gain block or IC opamp, transformer output and NON-transformer output ... and more

it's got it all

 

[TomWaterman]

[quote author="Flatpicker"][quote author="TomWaterman"]I wonder if Tim from 7th circle knows about this board...maybe he could shed some light???[/quote]I'm sure he could since he probably designed it. He used to post over at TT, but I wonder if he even knows about our new home. Drop him an email with our web address and ask him to stop by.[/quote]

Doh!!!! of course he knows...... He obviously designed it but you know what I meant....right?

I'll shoot him an email to tell him about this place. I have been wondering about the JLM boards....they have everything there and they are pro quality boards. Might work out better than home etching some.

Still I figure its not an over-complicated circuit to lay out and I would like the practice with Protel. I was thinking about using his power supply board though. I might make a clean pair of twin servos with Hardy990s with the JLM boards.

Tom

 

[JLM Audio]

Hi All

You can actually create an output trim at the input if that makes sense. By changing the transformer load resistor to a pot with a resistor in series. Eg if the load resistor on the secondary of the input transformer is 10k use a 5k pot in series with a 5k resistor which will give you 6dB of trim and then the gain around the opamp can be switched in 6dB gain steps.

Also Don't think of the JLM99v as colourful as it is very pure & smooth sounding especially in dual JLM99v mode. The JLM99v when running on its +/-34v rails will stay cleaner but with the silky Class A top end to higher levels than any opamp around. It only gets a 2nd Harmonic when putting out high output levels which are above where all other opamps have hit the power rails and are square wave clipping (odd harmonics) . It has a definite smooth Class A sound. Also don't just try the JLM99v. I would recommend trying the Fet 992 opamps designed by Fred Forssell and John hardy 990 based on Deane Jensen's original design. Also if you are after the more transient sounding A/B sound don't forget the old NE5534A IC as dual or single with or without DC servo setup with a good input transformer as it will put you into the premium end of the Mic pre spectrum.

The Mic pre colouration comes more from the input transformer used. If you want pure almost no iron sound use Lundahl 1538 as 1:2.5 ratio and dual 99v and no output transformer or our JLM mini input 1:4 ratio which tightens the bottom end and has a clean but smoother top end. The higher the input level the more coloured it gets which is great on drums. The fattest most coloured sound but in a good way comes from using the OEP A262A3E as the input set to 1:6.45 ratio with one JLM99v and a API output transformer running as 1:2 or 1:3. This would approach the Neve 1073 fatness but with faster top end. Plus the sounds of all the other input transformers that fit the PCB.
You cannot use DC servo easily with the 99v due to its 1v offset on the output. I am not a big fan of DC servo and would prefer a good capacitor to block the DC instead like most old designs used. The DC servo rolls off at 12dB per octave and needs to be set to a very low frequency to not interfere with the audio low end. But in reality a higher frequency has to be used to trim the DC quick enough during pot or switching of Mic pre gain. It is on the PCB so DIY people can play with all configurations and make up there own mine.
Hope this is actually helpful and not just more confusing.

Also Thanks to all that have bought PCB's as we have almost sold out of the first batch of 60 PCB's. Please any questions feel free to put them here or email me on [email protected]

Thanks

 

[Flatpicker]

[quote author="JLM Audio"]...Hope this is actually helpful and not just more confusing...[/quote]Very helpful. Thanks!

 

[Bear]

Joe, what does the Dual config offer that the single opamp doesn't, especially in terms of tonal quality? And would the servo option work better with other opamps like the 990?

Bear

 

[TomWaterman]

Thanks a lot Joe!

Great info there - cheers :sam: :guinness:

Does the trim control at the input TX offer any benefits in performance over the one in the feedback loop (7th Circle design)?

I got hold of Tim from 7th Circle and he said he has already found us and would drop by at some point.

He also said that the compensation zobel network for the output TX was not needed for a 2503 but was there incase anyone decided to use something a little different. The 600 Ohm resistor is apparently there just to simulate a load.

Cheers Tom

 

[JLM Audio]

Hi All
The Dual opamp configuration lets the opamp work with a better range of negative feedback and keeps the gain bandwidth product (GBW) & Power bandwidth for the opamp well above the audio range at full gain. Keeping the Mic pre sounding much better and more constant across its gain range. Eg. to get 60dB of gain with a 1:2 input transformer proving 6dB of the gain a single opamp has to provide 54dB of the gain which is a voltage gain of 500 (hard work). Where as the Dual opamp divides the gain evenly between the opamps at 27dB each which is only a voltage gain for each opamp of 22.4 (cruising).

Gain layout single opamp Mic pre with 1:2 input transformer 2 x 500 = 1000 = 60dB
Gain layout dual opamp Mic pre with 1:2 input transformer 2 x 22.4 x 22.4 = 1000 = 60dB

The Dual servo will work with any opamp that has a minimal offset JH990 MM99 Fred992 and the end of this year the JLM99s. But the JLM99v is based on old opamp type of circuitry which is very minimal and uses a transistor base for the + input and a transistor emitter for the - input of the opamp which causes a 1v offset which works perfectly with an output electro cap and sounds great this way. Similar in ways to a Neve BA283 preamp but running on dual power rails. Similar types of circuits can be found in parts of Old Helios & Trident consoles. It is the large 1v offset that makes it hard for the servo to remove it without giving the servo large current control. Like I have said before I am not a big fan of DC servo in Mic Pre amps and the thought of giving it high current control to remove the 1v offset is something I would never do.

The input type of gain trim has no real benefits over the feedback type. it is just a different way to do it.

Joe

 

[Plexibreath]

Would the input trim on the resistor across the input transformer, be a better pad if set up to pad down 20db, than the typical pad that's inserted before the input transformer? I'm thinking it would be a cleaner way to pad down the signal going into the opamp, but is it necessary to pad down before the transformer? In this case I'm using a 1:10 into a 2520 opamp in a 312 circuit.
Kelley

 

[JLM Audio]

The Pad would usually be better before the input transformer so the input maximum level increases when the pad is in. Example would be using the Jensen JT-115k transformer which saturates at -2.5dBu @ 20Hz. It would remain at that point with a 20dB PAD after the transformer but putting the 20dB PAD before the transformer would increase the input saturation to +18dBu @ 20Hz allowing the Mic Pre to hand any level up to line level at its input.

Joe

 

[Plexibreath]

Thanks for setting me straight Joe, I had thought the pad was to keep from overdriving the opamp, I wasn't sure it was to keep from oversaturating the input transformer.
Kelley

 

[Bryson]

With a dual op amp circuit, would it be easy to have an ultra high gain mode, say 80 to 90 db, that could be switched in when needed (Ribbons)? Why haven't I seen this yet?

 

[JLM Audio]

Yes the JLM99*MB PCB with dual 99v on it you can just make R15 & R22 equal smaller values. An example with a Lundahl 1538 wired as1:2.5 ratio you can get 82dB of gain and with it wired as 1:5 ratio you can get 88dB of gain by making them both 300ohm instead of 1.8k. You could easily switch between these values with a switch. Thing to remember is that the noise floor once 80dB of gain has been used is not far under the input signal so hiss will be becoming audible. We have already sold some dual 99v kits setup so they can do 70dB of gain for ribbon microphones which is enough in most cases.

Joe

 

[TomWaterman]

Thanks very much for your knowledge Joe!!

I have also been thinking about putting a small meter circuit (either peak reading or VU - probably peak) on my mic pres so I will be able to see my output level (although I could monitor my DAW inputs I would like to have some indication on the box).

Would it be possible to hang it off an output transformer secondary. With the 2503s being wired to 1:2 and the spare 1:1 winding feeding the meter??
I guess that would not take the extra gain of the op tx into account though?

I have drawn up a little circuit straight from the application notes of the LM3916 IC - a 10 step meter with either peak or VU detection, but am unsure of the best place to put it in the API312 circuit.

Any ideas?

Thanks again

Regards Tom

 

[JLM Audio]

You can take the meter drive from where the output transformer connects to the opamp output cap if using a LM3916 and not a mechanical VU meter. As from memory the LM3916 only need about 1v to full scale. So by using a 10k 10 turn trim pot wired as a volume pot to the input of the LM3916 you can align it to match what ever output level you want. So having the output transformer giving 6 or 9db of extra gain will not be a problem. If using Mechanical VU meters use a TL072 opamp as a buffer between the audio opamp output and the VU meter otherwise the VU meter will usually raise the output THD up by roughly a factor of 10. Also if you want peak hold check out the LB1412 IC from Sanyo. It has 12 LED's displaying up to +12dBVU (+16dBM) when 0dBVU = +4dBM. The LED's past 0vu have peak hold and release and chips can be linked together so peak hold release is synced between channels. It makes a great output meter.

www.datasheetcatalog.com/datasheet/L/LB1412.shtml

Joe

 

[TomWaterman]

Thanks a lot Joe - I'll check out that chip and see what I can come up with!

Cheers :sam: :guinness:

The circuit that I have drawn out from the LM3916 application notes uses two halves of a LF353 IC to create full wave rectification and and apparently peak detection that complies with the German DIN PM spec.

This then feeds the LM3916 which can drive 10 L.E.Ds and has a trim pot for reference calibration and L.E.D brightness. I will post it up here as soon as I can.

Do you think this circuit could also sit at the opamp output like the LM3916 - potential problems?

Thanks for assisting a novice

Tom

 

[JLM Audio]

Sounds like it would work fine in the same way. Would be interesting to see the circuit you mention using the LM3916 running 20 LED's.

Joe

 

[TomWaterman]

Hi Joe!

I'm drawing it out in Protel now!

In the application notes the LM3916 can be cascaded with LM3915s to drive up to 100 L.E.Ds. It can be set for bargraph or dot-mode pretty easily.

I'm thinking about trying out a 40 step stereo meter for my monitor console (project at uni) that will be bargraph and switchable peak/VU.
The application notes have it all in there, it even suggests combining a small AC signal at the input to smooth out the meters visual response. I'm sure it would look pretty classy but not too sure about the performance.

I'll be back

Tom