Intro and PSU opinions

[ricothetroll]

Just to say, I've tested the THAT1512 circuit with a 10k rev log pot, and the gain linearity is REALLY poor, like you have half the gain (30dB) on the last 10% of the pot. I'd suggest you use 1k rev log instead, and as you have a PAD at the input you won't need the 0dB setting (who needs it anyway ?). THAT1510 shouldn't be very different I guess.
My advice is you do some tests with cheap (Alpha) log pots wired backwards before ordering custom rev log OMEG pots
Best regards.
Eric

 

[joaquins]

The OPA is there as DC servo, instead of using a cap which may or may not be a good, and using a servo you avoid this problem. It don't protect as much for the outside world but who cares, strange that something as bad may occur to damage an output stage of a mic pre and in the case it happens you know how to solder so, stay with the OPA I'd say.

Well, forget about META, there is somewhere a three with this type of preamp, with already worked out boards, a lot in fact, but I don't remember and the only option is search, if you can't find it just run a test with what you have here and try it out, other better documented projects and almost as cheap are amek 2500 and green pre, both worthwhile I guess, always about to build one but never did, maybe worth the difference in a studio situation, for live performance or not so high end studio ICs are really good, and in a really high end studio some better preamps are a must, transformer input for sure, not because the transformer it self but because the guys who have transformers are really good (and because of the transformer it self) but those two discrete input may be still useful in some cases against tx coupled.

JS

 

[Giggity]

ricothetroll.......  I was considering a 12 position switch for gain control but haven't a clue how to implement a trim for the gain.

joaquins......  I did some more searching in general as well as in the META links.  I didn't really come up with anything useful.  Thanks for the explanation about the OPA606.  I have looked at preamps that use transformers and ran away when I saw the cost of the transformers by themselves.  This is for just a small home studio for only myself so, I'm not wanting nor expecting great things.  Just simply a hobby I like to play with.

So...back to work on laying out the PCBs.  When I get all this done a verified that it actually works, I'll post all the relevant files for everyone to use.

 

[ricardo]

You mention that some of the later mods are not required./

Ooops!  I thought your mods were to the Preamp circuit .. not the PSU.  :-[ Your mods are good.

I've tested the THAT1512 circuit with a 10k rev log pot, and the gain linearity is REALLY poor, like you have half the gain (30dB) on the last 10% of the pot. I'd suggest you use 1k rev log instead, and as you have a PAD at the input you won't need the 0dB setting (who needs it anyway ?). THAT1510 shouldn't be very different I guess.
My advice is you do some tests with cheap (Alpha) log pots wired backwards before ordering custom rev log OMEG pots

I think you connected your pots wrongly.  With the pot at the centre position, the end to use is the one with less resistance to the wiper.

With a rev log pot, the Gain will increase with clockwise rotation.  A 'normal' log pot connected properly will decrease Gain clockwise.

You shouldn't use a pot to control more than 40dB Gain range so the pot should be less than 100x the smallest feedback resistor.

Use switched gain if you want more gain range on a single knob.

 

[JohnRoberts]

That is a well known issue with that topology, which is so good at everything else it is worth living with that gain characteristic.

If you want precision "and" wide range a mutli-position switch is not a bad idea.

While back at my day job where we were using millions of pots a year, we had the luxury of designing fully custom potentiometers especially for this application. As I recall the improved gain pots used a 3 step screening process, instead of the standard 1 step screen used for typical pots. By overlaying 3 different resistive ink screens it is possible to get good adjustment sensitivity down in the low ohms end, while simultaneously providing a few 10k of resistance at the high resistance end.

When buying hundreds of thousands at a time the extra processing cost something like $0.10 or $0.15 extra, but is well worth the cost hit for good user ergonomics.

While I do not have any part numbers or models to suggest, you might be able to buy a gain pot as a repair part from a major manufacturer that has a better than standard off the shelf pot taper.   

JR

 

[Giggity]

If you want precision "and" wide range a mutli-position switch is not a bad idea.

I just don't know how to incorporate a trim pot to work with a multi position switch.

While I do not have any part numbers or models to suggest, you might be able to buy a gain pot as a repair part from a major manufacturer that has a better than standard off the shelf pot taper.   

JR

Not a bad idea.  Researching what would work will take some time.

Ooops!  I thought your mods were to the Preamp circuit .. not the PSU.  :-[ Your mods are good.

Actually, I have tweeked both, on the advice of other members here.  Minor ones to the PSU but, more on the micpre.

The only other thing that may be worth adding to the micpre would be a PEAK indicator LED.

FWIW....  My goal here is to get rid of my Soundcraft E8 console that has served as nothing more than a micpre and an outboard monitor volume control.  Takes up a lot of desk space to only be doing those two things.  I can handle the monitor volume control, no problem.  The micpre I'm looking to build here surely will be better than those found in the Soundcraft.  I posted a schematic of the Soundcraft ones I believe on the previous page.  I really don't need the mixing capabilities because I can do that in the software for the M-Audio 2496 card.

 

[ricardo]

I just don't know how to incorporate a trim pot to work with a multi position switch.

The Seventh Circle T15 design is actually quite competent if a bit too complex for my liking.  Their trim pot is in the right place and properly buffered.

I second JR's suggestion of getting a replacement pot from a reputable mixer maker.  Anyone know of these?  8)

 

[Giggity]

I think I may have figured out a way to use a rotary switch for the gain and not need a trim pot.  But, I have no point of reference.  So, just how course/fine is a 2.5dB step?

 

[joaquins]

That's pretty good, I think with 6 or 10 dB/step is fine if you have a trim also, I figured out a way of doing both in a single stage, instead of having two separate stages, but not for a mic preamp, which may be a problem since your feedback resistors shouldn't change much for good noise/THD performance, so that's why preamps mostly have two stages for doing so. John Hardy uses two gang pot with different ranges of gain to get better range and tracking, just a switch that selects one or the other, the hard part is to get a pot with one 10K half and one 500Ω half, IIRC.

JS

 

[Giggity]

Sounds good.  Inspired by what ricothetroll was saying in the first post on this page, I took a look at the THAT design note #138 some more.  It shows the top 30dB (+30dB~+60dB) of gain is covered from 327 ohm to 10 ohm.  With 12 steps in that range, I believe I can get 2.5dB increments (4 steps for every 10dB).  And with the 20dB pad before the signal ever gets to the gain stage, I should also be able to cover a 30dB range from +10dB to +40dB.  That will give a total range from +10dB to +60dB with an overlap between +30dB to +40dB.  And of course with resistors, I can set the LAW up how I want and change it if I feel the need.  Obviously I agree that the 0dB to +10dB range is pretty useless.

Sound like a good idea?

 

[joaquins]

Sounds good.  Inspired by what ricothetroll was saying in the first post on this page, I took a look at the THAT design note #138 some more.  It shows the top 30dB (+30dB~+60dB) of gain is covered from 327 ohm to 10 ohm.  With 12 steps in that range, I believe I can get 2.5dB increments (4 steps for every 10dB).  And with the 20dB pad before the signal ever gets to the gain stage, I should also be able to cover a 30dB range from +10dB to +40dB.  That will give a total range from +10dB to +60dB with an overlap between +30dB to +40dB.  And of course with resistors, I can set the LAW up how I want and change it if I feel the need.  Obviously I agree that the 0dB to +10dB range is pretty useless.

Sound like a good idea?

In that case you will have 2.7dB on each step rather than 2.5dB, 30dB range, 12 steps, one for +30dB, one for +60dB and 10 steps in the middle are 30/11~2.7, but still pretty good.

0dB to +10dB isn't that useless, I think +50dB to +60dB is more useless than 0dB to +10dB, a kick drum kicked hard in a kick drum dedicated microphone produce a very high level, and even 10dB gain may be too much for it. Over +50dB will only be used in low sensitivity mics with low SPL sources, let's say a ribbon on a shy singer, and for ribbons and mostly in this condition a dedicated mic pre have great benefits, in any case, if you don't have a ribbon mic you most likely won't need it, or you can use +20dB to +60dB in your switch so 40dB range in 12 positions is 3.6dB/step which is quite usable, I want to hit around -20dBFS on average peaks so I get 20dB of headroom for loud moments, I could live with -17dBFS or -21dBFS for my average peaks, but that's me, some people prefer less than that and don't bother with eventual clips when recording, it's not bad but you need a really good converter to work this way properly, in many converters when you are -3dBFS already has much more distortion than at -10dBFS.

JS

 

[Giggity]

No acoustic drums and no ribbon mics here.  The only thing that gets a mic are guitar amps and vocals.  Drums are electronic.  Keyboard electronic and bass is DI.  Only dynamic and condenser mics.  Given my environment, about 90~95dB SPL (fairly quiet) is about all I can do without making other people mad.  That and I don't want to listen to it any louder.

I looked for specs on the micpres built into the Soundcraft E8 to use for reference but, couldn't find any. All I know is that if the gain knob turns from 7:00 o'clock t o 5:00 o'clock, I use the area from 10:00 ~ 2:00 the most.  Pretty much right in the middle.

I'll figure out some control for gain but, it would be helpful if I knew how to add a trim circuit.  Just copy the parts around the OPA134 from the Seventh Circle pre and insert that between the DC servo and the phase circuit?

 

[joaquins]

There or after, it's the same, just add an amp stage with your desired trim gain, but it isn't really necessary, what I like to do when I design something is to make the PCB thinking in things I might need so no need to rebuild if I actually need it, but not build in the first try... what I'm saying, design the trim part and add it to your circuit but don't build it till you need it.

JS

 

[Giggity]

Thanks, joaquins.  I can just jumper it for the time being.  I also think, at first, I will just go for a 5k rev log pot.  That should cover +9.5dB to +60dB, according to the THAT design note 138.

Time to get back to PCB layouts.

 

[joaquins]

You probably would be fine with maybe 45dB max gain...

JS

 

[Giggity]

I just wanted to thank everyone again, particularly joaquins for all the help.  But before I jump in to laying out and etching a PCB for this, I want to do one a bit simpler and less consequential.  I have been playing with different mods to a BMP the last few days and thought it may be better/easier to layout a Cornish mod BMP.  Slightly different output, tone and clipping topography than a standard BMP.

Once I get this done successfully and feel more comfortable doing layouts and etching, I will start the etching on this project.  Meanwhile I will still work on the layout for both the PSU and the micpre and post them for review.  Probably in a different thread.

Thanks again everyone.

 

[ricothetroll]

Another option : use a pot with a switch for the first low gain setting...