Audio Design *Input from PhiPi needed*
average joe
Posts: 795
I know this isn't directly related to a parallax product, BUT... I don't know who else to ask. I have designed a preamp for my guitar *part of the LARGER project that DOES involve parallax products. I am working on the final revision and was wondering if what I can do to improve s/n ect. If you want more information on the whole project, let me know. Any thoughts, comments, suggestions are appreciated!
728 x 716 - 12K
Comments
The diagram is kinda incomplete, so a bit hard to read and make some judgements. I'm guessing it's a multiple output buffer w/tone control? The two bottom 74's are set up for unity gain - did you want that? And what are those "terminal strips" about?
Need more info, man.
Fellow guitar-electronics tinkerer,
DJ
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Instead of:
"Those who can, do.· Those who can't, teach." (Shaw)
I prefer:
"Those who know, do.· Those who understand, teach." (Aristotle)
·
I know this is very incomplete, so I will try to explain. This is a schematic for an output buffer, built into my guitar. It will send signals from both pickups, and from the pickup switch to the brain over 50' of 6 - 24awg. The brain *a bs2 and 4066s select bridge/neck/both for 2 effects processors. The terminal strips are as follows...
3pin- Pickup in, Neck-Gnd-Bridge
4pin- Pickup Sw, pwr-gnd-m-n switches
6pin- connection to 1' extention cord from guitar to brain..
I am re-designing the brain, if I get schematics done soon I will post those to this thread. I also have a "bypass" unit with a 4066 and 1/4 inch out that runs off a 9v.
Current changes to the schematic will be ...
#1 Supply going from +5 to +12 *or maybe* +15v
#2 Bias resistors changing from 1k to 47k
#3 Trim pots changing from 2.2k to 1m
Hope this helps.
Why is there a 1uF and a 100uF in parallel on the input of the top '74?
Why are there two filters on the input to the top '74?
And to answer your original question on how to improve S/N...in general, use metal film resistors and the lowest value of resistors.· Large value caps are great for wide audio bandwidth, but this is a guitar with not much harmonic content above 10KHz, so keep that in mind.· Limiting the bandwith can do a lot for reducing noise.· The type of cap can matter; I use polypropelyne for signal coupling·in the amps I've built but they can be physically large with high values of capacitance.· Wiring can play a part also.· With any kind of gain in the circuit, input/output cross coupling can occur if the in/out wires are too close.· Ground is always important - single point "star ground" is the best.
Later,
DJ
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Instead of:
"Those who can, do.· Those who can't, teach." (Shaw)
I prefer:
"Those who know, do.· Those who understand, teach." (Aristotle)
Post Edited (davejames) : 7/27/2009 5:11:55 AM GMT
Joe
(smiles)
DJ
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Instead of:
"Those who can, do.· Those who can't, teach." (Shaw)
I prefer:
"Those who know, do.· Those who understand, teach." (Aristotle)
·
Thanks again Dj
Joe
A question and a few comments (probably nit-picky stuff):
Q - why the "remote switching" of the pickups?
C - the 10uF output coupling caps could be reduced to 2uF (probably 1uF) if size is an issue. The 10uF doesn't buy much as guitar doesn't go very far below 110Hz.
C - the 1uF/100uF combo on the top '74 could be replaced with a .1uF ceramic/10uF tantalum for reasonable filtering
C - I see the value of the bias resistors has changed to 470K, good
I'll be interested to see the rest of the project!
Later,
DJ
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Instead of:
"Those who can, do.· Those who can't, teach." (Shaw)
I prefer:
"Those who know, do.· Those who understand, teach." (Aristotle)
·
Thanks again DJ
Joe
Post Edited (average joe) : 7/29/2009 2:26:10 AM GMT
I forgot you mentioned 50 feet of cable that the supply would be traversing. The '74s won't take much current to run, and the voltage divider for biasing the other '74s will only draw 10 - 15 mA (depending on the supply voltage)...so my thought of 10uF should be ok, but you might note a need to evaluate the value when the guitar-side is powered up.
You're welcome Joe!
DJ
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Instead of:
"Those who can, do.· Those who can't, teach." (Shaw)
I prefer:
"Those who know, do.· Those who understand, teach." (Aristotle)
·
In looking at the brain schematic...I see two '74s (IC4A/B) being driven by the same input, their respective outputs connect to two analog switches (IC1A/1B), and the output of the switches are summed back to a single buffer (IC5A). I can see that each analog switch is under BS2 control, but why the double buffering/summing? Also, the two output buffers (IC5A/5B) are routed to the OUT connector (I get that), but they are also connected to another set of analog switches in which the outputs are hard-wired together. What's that about?
I'll be very interested to see the final product!
Off for a family outing,
DJ
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Instead of:
"Those who can, do.· Those who can't, teach." (Shaw)
I prefer:
"Those who know, do.· Those who understand, teach." (Aristotle)
·
The "next" step is going to be interesting. I want to free up a wire so I can use the volume knob for RCtime. My thought is only using one wire for the pickup selector. For an on-[noparse][[/noparse]on]-on sw use gnd-[noparse][[/noparse]6v]-12v and sort the mess out later. I'm still trying to get the basics of selecting a pickup outside of the guitar. It doesn't help that I've had no formal education in electronics. It is very interesting , minus the steep learning curve, and I'm starting to get it *I think*
Yes - I'd like to see the schematic of the P/S. Unless the regulators you used as replacements had a different pin out, I don't see why the fuses would blow. One thing with negative voltage regulators like the venerable 79xx series is that the tab is NOT ground as it is with the positive voltage versions (78xx).
The center pic is kinda fuzzy - what is it?
Later,
DJ
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Instead of:
"Those who can, do.· Those who can't, teach." (Shaw)
I prefer:
"Those who know, do.· Those who understand, teach." (Aristotle)
·
Anyway, after I got the power supply working I experimited a bit. I have no signal output. I'm thinking I might need to swap the summing resistors with the bypass caps. Any thoughts? Things I should look out for? *It's times like these I wish I had a scope*
*EDIT* Fixed schematic!
Post Edited (average joe) : 8/14/2009 5:31:05 PM GMT
I can't hazard a guess as to why you have no output with the p/s attached versus the 9V battery other than the obvious cynical question of "is it hooked up correctly?" (smiles)
It would appear that the guitar signal is routed through the AND gates? (IC1D-12 connects to IC2A-1,2 and LED1) I don't understand this unless you *want* distortion.
Later,
DJ
BTW - what are you using to draw the schematic?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Instead of:
"Those who can, do.· Those who can't, teach." (Shaw)
I prefer:
"Those who know, do.· Those who understand, teach." (Aristotle)
·
I played with different ideas of how to improve this simple setup for a while. The one thing I wanted to be able to do was use the bridge pickup for my distortion, while using the neck *or both* for the clean channel. This meant selecting pickups outside the guitar. I have several ideas on how to use the volume knob, pickup selector as a user interface.
I'm using eagle 5.3.0 to draw the schematics.
Sooo, are you running multiple amps or multiple channels on a mixer?
DJ
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Instead of:
"Those who can, do.· Those who can't, teach." (Shaw)
I prefer:
"Those who know, do.· Those who understand, teach." (Aristotle)
·
How 'long' was the cable run? Using twisted/pair & balanced is a solid way to go...but a bit overkill?
Dave
I have decided to revisit this project because I feel that I will be able to improve it knowing what I know now. As you can see, I decided to use a mpf102 FET as my first stage, followed by '74s for the rest of the circuit. My design referenced the "600 ohm line driver" and "preamp to handle anything" posted here http://www.diystompboxes.com/smfforum/index.php?topic=82609.0;prev_next=prev I am currently building the in guitar part for testing. My previously posted schematic had some errors, so I am posting the updated guitar buffer.
Now here's my first question. On the updated schematic, I am not sure if I need R28 and R30 or what the ideal value will be. Also, I am not sure if it makes a difference if I place R31, R37 before or after C19, C20. Also note that Q1, Q2 are MPF102, not J201.
Also, I will be using a Propeller to control said project with. I'm implementing features discussed in [url] http://forums.parallax.com/showthread.php?136315-Prop-Sigma-ADC [/url] I need to get the input stages finished before I do any more work on the Envelope Generator. Anyway, what are your thoughts?
More to come, I will keep you updated as I progress.
Lemme think about this a bit.
Hey! Congratulations!
After a bit of thinking, I revised my schematic yet again. This time I am using a AD5206 6 - 10k digipot as a volume control right before the switching. These are R28, R30, R57, R58. This works quite well despite the fact there is no gain. My invert circuit didn't work as expected, so it's being removed from the schematic. I might revisit the idea later. I need to finish wiring up logic control and do some further testing. So far I am very happy with the sound. My guitar sounds very clean and warm. The low end response sounds a bit flabby in my opinion, so I will be changing input caps on the in guitar board to metal-film and the rest of the audio caps to polyester. I do hear a bit of "chatter" when adjusting settings on the digipot. I believe this is coming from the bs2 I am using strictly for testing. I plan on controlling the final product with a propeller chip. User Interface is still under design. Still a ton of work left to do, but some progress at least.
The channel that actually works sounds really good. The neck pickup sounds warm and full and the bridge sounds nice and plucky. Both pickups without the invert stage provides a funky out of phase tone. I REALLY like this sound. With heavy distortion it has a very nice bite, clean it sounds very thin. Noise is minimal. Still some when bs2 is communicating with the pc. I tried recording some samples, but I will have to do that later when I can actually play the guitar. I can't wait to get both channels working and test with both of my processors.
The board looks a bit sloppy, I guess I should have spent more time getting the wiring to look clean.
I've been looking at the schematic and will have some questions/comments later today.
Glad to hear it's working to your tonal-satisfaction!
BTW - thanks for "like-ing" my recent efforts.
I finally got a chance to tinker on my project again and discovered why the second channel had issues. I used 1k pull down resistors on 2 of the 4066's instead of 10k. I replaced the resistors and everything worked great. I am revising the final summer stage because it introduces switching noise. The solution is to add ac coupling caps after the 4066 and change the tl074 reference to ground instead of the 2.5v ref. I believe I will also need a resistor from the junction of the 4066 and cap to 2.5v ref. Not sure of the value here yet, in previous designs I used 1M, but I'm trying to keep resistor values low to reduce thermal noise. I THINK 100k should work, but I will have to play around with this a bit.
I still need to wire up some transistors to a 374. I am thinking about using SPI to control the switcher since I am already using serial for the AD5206. I wish I had some 595's laying around right about now.
The user interface will take quite some time to develop. I will be using a 3.2" touchscreen and possibly 8 rotary encoders. Hopefully the screen will be here in the next few days.
Audio clips still need to recorded which requires me hooking up both processors. That's all I have for now. I hope you're having success with the Director.
I'm still collecting my questions from the last communication...
I'll be very interested to see/hear the "final" (is it ever?) project.
How's the switching noise when you cut in/out the 4066s? That's the one thing I'm contorting over now. In some way, some how, whenever the BS2 asserts a "mute" condition, I get this 100 - 200mV transient in the output...which of course translates to a real nice "TICK!" with every switch press.
Anyway, I'll be back to debug later this evening. I've already yanked the audio buffer board and mux board, so that doesn't leave much else to trouble shoot. We'll see...
Later man.
I have a fairly bad switching transient right now, don't have a scope so I can't really SAY BUT. The way the schematic is drawn is insufficient. This results in large transients when switching. Clean it is a slight tick, with ample distortion it is a loud pop. The problem is, the summing input varies from about 0v with no switch on, 2.5v with one and 5v, with both. The output from the 4066 needs to be AC coupled and a return path for the switching current needs to be bypassed.
I think this should work, I will test it when I get a chance. R1,R2, R3 should bias the junction. I would like to lower the values to limit thermal noise. I'm ordering metal film resistors for important points such as this. All my 4066 designs suffered switching noise. A design similar to the the one above WITHOUT the bias resistor gave best results. Hopefully the bias resistor will fix this. I'm trying to talk my wife into letting me buy a scope with some of our tax return. The idea is being met with some resistance so we'll see. LOL. I'd be interested to see how you designed your switching. Dual supply seems to work the best but it adds some major challenges as well. Voltage translation is the big issue with dual 7.5v rails. Keeping the control signal biased around a -4v and +4v swing is a bit tough without using a lot of current during. There are better analog chips, and other designs. I expect to see some transient when switching, but it would be ideal to keep it to under a few mV.I haven't used relay based switchers, but I would imagine they suffer transients as well. When I think about contact bounce and EMI generated, I question what they really do to a signal with significant gain.
And figure 2... *edit* sorry I forgot to change the input resistor to a cap.