1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

Building Preamps

Discussion in 'Preamps / Channel Strips' started by Spase, Feb 3, 2009.

  1. Spase

    Spase Active Member

    Hey everyone!
    I have recently found myself with a lot of time on my hands, and have been looking at building some preamps. My hope is that I can maybe make some money selling them if I can find a good price/performance point. My father is helping with design, but while he is a PHD EE, he hasn't done much for audio. So I was hoping some of you might help with some of the audio specific pitfalls that we might encounter.

    The basic things that a preamp should have: Phantom power, phase reverse, -20db pad, some metering, gain control.

    We were thinking of going with digitally controled gain, which we could then use for a unique (to my knowledge) feature: Self setting gain. Basically there would be a "set" button. You press it, check the mic, press it again, and the preamp would set the output to the peak of the check. There would still be a manual gain control, so this wouldn't be the only way to set the gain.

    We have decided to start the design around a Burr-Brown chip. I know that we want metal film resistors and generally using electrolyitic caps (Panasonic FM, Nichion HE are 2 lines that seem to be popular) and avoiding tantalum caps is good. I also know that transformers can be nice, but push the price up fast.

    Anyone know of a good rack mount box supplier? Most of the other components seem easy by comparison.

    Any info would be appreciated.

  2. Link555

    Link555 Well-Known Member

    Yes- rackmount boxes
    They are fairly cheap.

    When using digital pots and the Microcontroller, PCB layout is critical.

    Get all digital control well away from (physically isolated) from the analog path. Opto-couplers for control signals may be your best friend here, but keep in mind the I2C or serial controlled pots may have communication issues with an opto-couplers in the path.

    The common (or circuit returns) for digital and analog circuit need to be keep separate. If you must tie them together, do so only at the star point connection.

    Feel free to PM me if have any specific questions.

    Happy building.
  3. Link555

    Link555 Well-Known Member

    oh and about caps, take a read of this before you settle for Electrolytic caps.....

    (Dead Link Removed)



    Food for thought...
  4. dvdhawk

    dvdhawk Well-Known Member


    Here's a supplier of rack enclosures I've dealt with:


    I think they still do custom metalwork too, if you want a pre-punched panel etched with logos, control names, and numbers etc.

    Those Middle-Atlantic enclosures at RP at nice too.

    Some enclosures, like Raxxess and the Mid-Atlantic EC series, are nice for some projects. But they are 2-piece, which might be hard to work with. Front & bottom are 1 piece / back, top, and sides are the other piece. This might not accommodate your plan for mounting your circuit boards, pots, and connectors. The problem is, that makes the back panel a takeaway, and if that's where you planned to put the connectors you have to leave excessive wire lenghts inside or put a multi-pin connector on the board.
  5. Boswell

    Boswell Moderator Distinguished Member

    How you do the digital gain selection depends on which B-B part you are considering. The INAxx series are instrumentation amplifier configuration with internal gain resistors and so do not generally have high-impedance points available for gain resistor selection. The exception is the venerable INA103, where you can use purely external resistors for gain and hence pick off the required taps using a dual multiplexer to feed into the inverting inputs. In this way, the signal current that sets the gain does not have to flow through the on-resistance of the multiplexer, which varies with applied voltage causing distortion.

    You could consider the new PGA2500, which does the job for you, and all you have to do is supply an opto-isolated serial bitstream from a local PIC or AVR microprocessor to set the required gain.

    You need to take real care and give great attention to detail to get a good-sounding result. For example, PCB track impedance (resistance, capacitance and inductance) can make a surprising difference. Star points for ground helps keep signal currents out of power current paths.

    I've done professional designs with most of the B-B and AD parts, and there can be pitfalls at each corner.
  6. Spase

    Spase Active Member

    Wow! Thanks for the great responses.
  7. Link555

    Link555 Well-Known Member

    Send some clips and photos when your get the pre built. I would be interested in the results.

    Have fun!
  8. Spase

    Spase Active Member

    Well, we have the first prototype boards in, and I have populated the power supply board. Here are pics of the boards:

    I hope to have the pre built and tested in the next week so we cn get all the bugs figured out.
  9. Boswell

    Boswell Moderator Distinguished Member

    Looks good. Let us know how it sounds. Which amplifier did you go with in the end?

    PS img specifiers don't like "=" characters in them.
  10. Spase

    Spase Active Member

    we did this one around the BB INA 217. I'm going to use a socket for the chip so we can easily sub in different chips and see how it changes the sound. I'm going to start on the preamp board today and hopefully I'll let you know next week how it sounds.
  11. Spase

    Spase Active Member

    Finally got through most of the testing. There have been just a ton of issues, though when it finally got through a speaker, the sound seemed pretty nice. We checked the frquency response, and it looked pretty flat from 3hertz-30khertz. We'll probably raise both of those a bit, I'm thinking more like 15-45k. I checked it against my Soundcraft Compact4 and it was a bit quieter and sounded at least as good (hard to say exactly where it might fall from just talking through it for a few minutes). When we went to check the phantom power we had some issues and I need to make some more changes. I'll check it against my Great River tomorrow if I don't have any issues getting it back up. I'm hoping to have it ready in time to bring it out for my weekend gigs and check it out in a live situation. That should give me a better idea of where it's sitting sonically.
  12. Kev

    Kev Well-Known Member

    There is so much out there now since the old Tech Talk
    the 217 fits into an old SSM2016 and now 2019 ... try the THAT unit
    so you are welcome to any of my old work
    and hard to find anything new that hasn't been done to death

    but hey
    go for it

    that's the old GroupDIY.com

    look for the baby animal kits
    the go between boards are useful

    Seventh Circle have a nice 1272 derivative and some other topologies

    Purple Audio, John Hardy and Fred Forssell for DOAse is so much out there
  13. Spase

    Spase Active Member

    Alright! I have put it through a few gigs, and I have to say it sounds pretty good. I have recorded it next to my Great River MP2-NV and it's pretty hard to hear the difference. I can hear some subtle differences, but the others helping me with it think it sounds pretty much identical. This is with some issues that we know need some fixing - like I need to change some op amps to get rid of a little crossover distortion.

    I did recordings of guitar - I ran 2 separate MD421s to the two preamps on the same speaker, and vocals with my Beyerdynamic Opus 89 split to the 2 preamps. I'll post some clips when I get a chance.
  14. Boswell

    Boswell Moderator Distinguished Member

    You've got crossover distortion but it still sounds identical to a Great River MP2-NV?

    What op amps have you got in there at the moment and what are you thinking of changing them to?
  15. Spase

    Spase Active Member

    we put in OPA 137 op amps on the outputs, and that is where the distortion is coming from. It is not a lot of distortion, but we want there to be none if possible (actually we checked the GR and found a little there as well - though it was just a slight flat spot on the sine wave we put through). We are changing those to OPA 2604 which look to have been designed to eliminate this issue, though it seems we will have to be careful of oscillation.

    As to the sound, I am recording these at a live show, so the recording situation is far from perfect. I think there are some differences, though they are subtle. The others who were listening are definately not trained ears.

    I'll get the clips up so you can hear for yourself, but so far I am actually pretty impressed with how well it seems to be coming out.
  16. Spase

    Spase Active Member

    OK, here's some links, if they work:

    Well, they seem to work.
    The 1's are all guitar. They were recorded with an MD421 through the Great River through my EMU 1616M line in, and another MD421 on the same amp through another line in on my 1616M. A and B are just the guitar, and C and D are mixed in with the rest or the tracks (wich were taken through the board subgroups as I only brought the base unit for 6 inputs - it's just a rough mix). The 2 set is similar, but instead of 2 separate lines on guitar, I split the lead vocals from a Shure KSM9 through the 2 pres.

    So, the gear is all decent except the Great River(which is pretty darned good). The music was recorded at a live show. The 1's are from a group called 6 Wheel Drive and the 2's are from a group called Trick Daisy.
  17. Spase

    Spase Active Member

    New pics! Prototype #2 is here!


    Definately a few changes on this one.
  18. Spase

    Spase Active Member


    It's getting close!
  19. Link555

    Link555 Well-Known Member

    Cool pics, on the pcb where does the micro live?
  20. Spase

    Spase Active Member

    On the pic, the input/output is on the right. There is a connector on the left that goes to the front panel board(the small one). The version pictured has the processor about in the middle(on the bottom). We have a problem with that version; the processor turning on the 7 segment to display the gain is causing noise in the audio. We have (hopefully) fixed that by separating out the power and ground for the processor and the LEDs and moving the processor to the bottom left corner by the connector to the front panel. We have a few other things going on to try to ensure a cleaner signal as well.

Share This Page