Getting started: Building a First pass Audio pre-amp

Experiment # 1: Disconnect the speakers from an old radio
1. From an old radio disconnect the speakers by cutting the wires
2. Reconnect the speakers by joining wires: Still work? Good!
3. Connect another set of speakers in place of the ones you disconnected. Still work? Good!
4. Now time to build your audio pre-amplifier. This takes us to experiment # 2


Experiment # 2: Build a pre-amp
1. We're going to start off with something simple that will give us confidence to proceed. Go to this schematic and build it on the breadboard.
2. Connect the function generator sine wave input to this and see that you get a clean sine wave at the output: Done
3. Now connect the radio output to the input of this circuit and the output of this circuit to the speaker : Done

Some observations
1. Even though the output of the radio puts out a differential output signal, you can use it in a single ended fashion by grounding one of the inputs. A single output does just fine
2.  You can even invert the inputs to the amp we built and it still sounds the same
3. If you want to use it in single ended fashion, sure you must ground one end. If you leave one end floating somehow, the channel seems to be different. Why?  
4. Distortion observed as well. Saturation observed. Did not wanna do more as that would blow out the speaker.

List of components

Component types from 1-10 are absolutely essential
1. Lots of resistors running the gamut from 10 Ohms to 1MOhms
   a. 10Ohms, 22 Ohms, 33 Ohms, 47 Ohms, 68 Ohms: 5 each
   b. 100 Ohms, 220 Ohms, 330 Ohms, 470 Ohms, 680 Ohms: 20 each
   c. 1K, 2.2K, 3.3K, 4.7K, 6.8K: 40 each
   d.  10K, 22K, 33K, 47k, 68k: 20 Each
   e.   100K, 220K, 330K, 470K, 680K, 1M: 10 each

2. Lots of capacitors running the gamut from 100uF to 10nF
   1. 10p, 100p: 10 each
    2. 1n, 10n   : 20 each
    3. 100n, 1u, 2.2u, 4.7u, 10u, 22u, 47u, 100u: 30 each
 
3. Normal Diodes
    a. 1N4148, 1N4004,4007: normal diodes #20 total
    b. 1N5060: clamping diodes, #5
 
4. Zener Diodes: 6V, 12V, 1W: 10 total

5. A few opamps
              a. LM741: Classic. 10 total
              b. LM 386 Audio Amplifier which has internal feedback.
              c. TL074M : classic. TL084, OP176, AD797 are suitable substitutes: 20
                  total
              d. OP184: 5
              e. LM1458: NTE7784 is a suitable subsitute: 5 total
              f. LF353: 2-3 total

6. Some BJTs
              a. Normal: MPS8099 (NPN), 2N5401 PNP, 2N5551 NPN, 2SB649 PNP output buffer, 2SD669 NPN output buffer, MJ15003, MJ15004, 2SA1302, 2SC3281, 2N3904/06 (Classic BJTs), tip31c/32c, 2SC 3281
              b. Matched Pair: MPS8599, 2N5551, 2N3806

7. Some Fets: 2SJ162, 2SK1052

8. Couple speakers
9. Couple microphones
10. Breadboard
11. Gauge wires for breadboarding and wire cutter. Different colors. Use SO22 or SO24. SO 20 and lower are too wide in diameter to fit inside breadboard slots.

Components 11-15 are nice to have but optional for now 
11. Couple ADCs and compatible DACs
12. Clock generators & Timers: NE555 and MN3103
13. Programmable logic
14. Couple transformers
15. Voltage regulator ICs: 7805, 7812, LM337/317
16. Chokes/Inductors

*******************************************************************************
1. Some analog filter chips. like UAF42

Ideas for a nice circuit

1. Graphic equalizer
2. Play a guitar string. Can the circuit recognize it?
3. Some kinda circuit that helps with figuring out the guitar?
4. An instrument to automatically tune the guitar. Should light up when the string is tuned!!!! Awesome

Troubleshooting test equipment: Function Generator and Oscilloscope

Plenty has been written about this oscilloscope, the venerable and grand Tektronix 2465 4 channel 300MHz Oscilloscope. I do not intend to clutter the space. This page is just a log of the issues I faced and solving them

First, The manual is available free of charge at http://bama.edebris.com/manuals/tek/2465/


Issue # 1: Waveform display is pulsing. It is as if the CRT dot moves from left to right while tracing out the sine wave input. It is like an ECG screen. I want a steady waveform display

The applied frequency is probably too low in the range of Hz. This should not happen for higher frequencies in the range of KHz


Issue # 2: Oscilloscope channel has DC offset. The input waveform is at zero DC, but the scope shows it as having some offset. The scope itself has some internal offset. When you apply a ground coupling, you see that it is ground coupled at 200mV. Why is this? Can this be changed?


Issue # 3: Can't see other channels

That's what the mode buttons are for. You need to press the knobs for the respective channels to see the waveforms for that


Issue # 4: Signal not triggering on channel X
If only channel X is on, then Channel X must be the trigger source. Make sure Channel X is set to the source on the trigger menu.


                              Function Generator
You might have a similar issues with the function generator from time to time. I am using a GW instek GFG-8020H function generator

1. The triangle waves look like sawtooths
2. The square looks like pulse.

No matter how much I adjust the duty cycle knob, I cannot get the square to look like a square or a triangle to look like a triangle. Also the square seems to have some overshoot followed by first order settling indicating some sort of stability feedback loop issue. Finally the sine wave looks distorted. The 2nd harmonic is 10% of the fundamental.

I used high frequency cables and probes, but it stays the same. So does not look like a probe or loading issue. Finally, I see that the frequency seems to flicker a bit about +/-2%. Did you encounter anything similar from your function generator? Any clues as to what this could be?


1. The answer is that the function generator is being used at the edge of it's range. If the range setting of the function generator is at 100K, but you have turned the frequency knob so low that you are actually putting out a 5KHz wave, then obviously you are operating the FG at the edge of it's range. When putting out 5KHz, use a 10K range (assuming max freq at 10K range is 20K)

2. Another issue is you may be using an abnormally high offset, saturating the circuit beyond it's point of linearity

3. Finally, take a look at the duty cycle knob. Turn it all the way inside. Don't mess with that knob while looking at Duty cycle. It will screw up the wave

Buying the basic Lab eqipment

                                 Choosing a starting area
I'm going to start with Audio because I can simply breadboard the circuits and check them out instead of having to solder them. Audio frequencies are under 20KHz, easily the range of operation of any breadboard. Also, the oscilloscope hardly needs to be be high performance as do the function generators and probes. Being a conservative fellow, I always like to start slowly, inexpensively with simply projects and add complexity to them


                                       The equipment
We need the following equipment to get started
1. Oscilloscope and probes: Up to 1MHz BW
2. Function Generator and associated BNC connectors: Up to 1MHz
3. DC power source and associated power clips: Decent PSRR up to 20 KHz

This is hardly the equipment you can pick up at the local walmart. Let's start off with Ebay. Best place for exotic items like this


                                         The Sellers
1. Ebay : online. Surprisingly this is not the cheapest way to get what you want, if what you want is a generic item. Ebay is great for items that are hard to find. I found a DC Dual power supply to be a lot cheaper when sourced from a shop selling test equipment than from ebay. Also, the shipping on ebay products is prohibitely high as well. Up to 3 times more than from a commercial seller. Either the commercial seller has some deal with the shipper due to the volume or else shipping costs are a big part of revenue for the individual sellers on ebay. Not sure.

2. Alibaba: online. Some very cheap Chinese equipment. But not sure about quality.

3. http://www.ladyada.net/library/procure/hobbyist.html

4.Some local electronic test equipment sellers

5.  https://www.valuetronics.com/ : looks like a great site if legit
           a. tek 2213 for $229. others for up to $90: https://www.valuetronics.com/detail/Used-tektronix-2213.cfm#.UaD2bs-fEuo
           b. kenwood Pw-18 power supply for $65: https://www.valuetronics.com/detail/Used-kenwood-pw18-2.cfm#.UaD1qc-fEuo. This stuff is going for $250-500 on ebay, Yikes!
           c.   Function Generators are a little more expensive than in ebay. Here is a tek one for $125. https://www.valuetronics.com/detail/Used-tektronix-fg501.cfm#.UaD3Ms-fEuo:


                                       The Economics
1. First hand agilent sells a function generator for $2000, Yikes! That's too much for something under 1 MHz. Funny that function generators are harder to come by than oscilloscopes on ebay. You would think these are easier to build than scopes.
2. A shitty brand new hand held oscilloscope goes for over $300.
3. DC power supplies from China seem to come in around $50 on ebay. What do they cost in china itself?
4. Damn, passive probes are expensive. Brand new 100 MHz is over $65. At ebay, it is $14. 100Hz is really the cost effective BW for a probe. Till that point, the prices do not increase a whole lot with BW. But after that the price just shoots up.  Considering that a 400 MHz probe costs $40 on ebay, perhaps a 300 Mhz oscilloscope used fair price is probably around $250?

                      

                                     Final Buying Price
The First 5 items are from ebay
1. Tek 2465 4 Channel 300MHz Oscilloscope: $270 + $42 (ship) = $312
2. Function Generator GW-Instek: $72 + $15(ship) = $87
3. 100 MHz OScilloscope probes: $14/pair
4. BNC cables: $10/pair
5. Banana to alligator clips: $3/pair

Did not get any great deals from ebay for sure, but did not get ripped either.The next item is from valuetronics

6. Kenwood PW18-2 Dual DC power supply: $65

Ebay bidding tips

If you are going to be setting up an electronics laboratory on a reasonable budget, you will be looking to ebay a lot to find used test equipment. Some tips are in order. These are not the best, but basic

1. First start looking outside of ebay to see what these items cost outside of ebay, via established businesses. Typically I would be willing to spend 75% that amount on ebay.

2. On ebay itself, Start of with looking at items that have already been sold which are similar to the items you are looking to buy. That will give you a ballpark idea of how much you may need to spend. This amount is usually 65% of what you could expect to buy for from an established businesses

3. Start bidding on items that are going to expire soon, i.e. within 2-3 hours. No point bidding on items that have two or more days to go. The bidding process seems to get very hectic near the very end, within the last few hours. Sniping might even happen at the last minute.

4. Take a look at similar items (to the ones you are bidding on right now), but which will expire a little later (than the item you are currently bidding on). You should be willing to pay about 10% more on the current item. For some specialty items, this may be hard.

5. Just make sure you do not place a maximum bid that is higher than 1. or 2.

Ebay should probably add a new algorithm to prevent sniping. Every time someone makes a new high bid, there should be an automatic time extension of 5-10 minutes to give the other bidders time to respond.  Basically, there must be at least 5-10 minutes for the new guy to respond with  higher price. This is better for everyone except the sniper.