|
Thought I`d throw this[1] out there again. I ran across it on an old M.O. disk where I archived it a long time ago.
If you can use it, great. Have fun. [1] Here`s a bit of help for those that may encounter a dead Small Clone. I recently resuscitated two and from that have some basic pointers on what to look for.
BTW, this isn`t an "exhaustive circuit theory" post, it`s just a "shot gun" approach at isolating trouble...you`ll need a schematic of course for things that I don`t mention and for basic circuit layout.
1. Check battery and battery "clip". It`s a "duh", but you`d be surprised. Next verify that the typical cheesoid EH external power adapter is not "open". This is a switching jack that removes the battery from the circuit when a jack is placed into it. If the "shorting action" isn`t working, the + from the battery will not make it to the circuit. 2.Verify that all semi-conductors are powered. This kills two birds; it not only tells you that all devices are biased, but checks the action of the {stereo input jack}. (applies the battery (-) to the circuit.
*RC4558 pin 8=9volts (battery voltage VCC) *LM358 pin 8=9volts *CD4047 pin(s) 4,5,6,14 and SAD1024 pin(s) 4,5,7,9,12,13 all should equal battery voltage minus an si emitter-base drop. MN3007 has pin 1 at this voltage. *There is a voltage divider (betw. VCC & gnd.) using two 56K res., there should be 1/2
VCC at their junction to offset input signals into devices using a single rail supply. (The entire small clone) *The signal input [pins 2&15 SAD1024, or pin 3 of MN3007] also has a separate dc offset that is adjustable via a trimpot. When adjusted, the last one I repaired was at exactly 4.5Vdc at the delay signal input pin. *The switching FET`s gate has a 4148 diode in series with the LED, the footswitch and a voltage divider. When the pedal is off, the diode`s CATHODE is at almost ground potential and when the unit is switched on, the cathode measures about 5 vdc. (cutoff/ sat. of FET) *There are several transistors living in the circuit. Use your schem. to figure biasing.
Let`s now look at the basic circuit function.
As long as the 4558 has bias, is "good", and the signal path is uninterrupted, you will have an output(at the "output" !jack!). The input comes in on pin 3 and exits on pin 1. It then is routed to a "mixer" section (remember this mixer for later), and then to the other half of the device.
Pin 6 is the input and the signal exits on pin 7 and goes to the output jack. There are two oscillators on the clone.
One is the CD4047 multivibrator and it is set up to run in a "astable mode". It`s base frequency is determined by a resistor (per device spec, but a diode is used in this case) at pin 2, and a capacitor (100pF) at pin 1. This frequency setup by the diode and the capacitor is approx. (120Khz). This waveform has a very short duty cycle (I didn`t measure it). These 120Khz waveforms appear at pins 2 and 1, and are 180deg. out of phase.(inverted) The CD4047 output is the Q and (-)Q square-wave (pins 10,11) that is required to "clock" the BBD. The frequency of pins 10 and 11 is half of the osc. freq. or 60Khz each. Without anything else, the small clone could simply operate as a delay pedal. The "twist" and the reason the unit "choruses", is that the final 60Khz BBD clock [frequencies] are shifting up and down based upon the low-frequency-oscillator. This circuit is built upon the LM358. It`s frequency spans between ~1-10HZ and the output on pin 1 is a saw-tooth wave. The output frequency is based upon the front panel control (1Meg pot.). This saw-tooth is sent through two resistors (4.7K and 2.7K) to control the "depth" of the device. Placing the "depth" switch in the "shallow" setting simply decreases the waveform`s amplitude by dropping a portion across the 2.7K resistor to ground. The signal "exits" at the junction of these resistors, and for reasons beyond me, has it`s "points" rounded-off by a 10uF electrolytic. (If this cap. is removed from the circuit, audible "clicks" at this low frequency rate will show up on the output.) Finally, it is connected to the diode at (pin 2) (cathode) and the 100pF cap. (pin 1) and directly to (pin 2)of the
CD4047. So now, we have a saw-tooth (with rounded peaks) modulating the 120Khz frequency setup of the CD4047. So, if I had to guess, the Q and(-)Q of the CD4047 is centered at 60Khz but swings within a window between 1- 10 hz. That`s what my scope tells me anyway.
Ok, back to the 4558. Right after pin 1, you`ll see that the guitar(?) signal splits and is sent not only to the output jack, but also goes through a 1uF cap. This signal rides on a dc level setup by the bias trimpot. It basically goes straight to the BBD input via an emitter follower. The level of the dc here is simply set to minimize distortion of the BBD`s delayed signal waveform.
If everything is chugging along so far, we should now see a (audio) guitar signal on the BBD output. Also present at the delay output is the clock waveform. These signals are sent through a filter. The "audio" portion, is sent through another emitter-follower where it arrives at the
"drain" of the switching FET. The "filter" mentioned is important; it is necessary to roll-off the
BBD`s clock frequencies. You also may have thought, if the BBD clocks were at 60Khz, who can "hear" those, so why do they need rolled off? Well, I don`t know, but I can tell you that without the filter they CAN be heard.(I tried to build a BBD delay unit years ago and inappropriate calculations for this filter is why I shelved it.) I can also tell you that there is a
"buzz" around the net about a "secret cap." value. Could it live here? I would suspect it does and it may be the one that connects between the two 39Ks and the emitter/10K junction of the buffer. The last clone that I serviced had this cap. value at .0037uF, yet the schem. shows a .0029. Others that I`ve seen also had differing values, so I suppose that the "best" cap. was chosen during final test of each unit. I guess this mojo-cap., "buzz" issue deals with the [amplitude of the delayed signal] and it`s ability to be perceived with the "raw" signal. If the filter attenuates incorrectly, or too much, some of the "audio" signal could be dumped to ground as well. The latter circumstance would yield weak chorusing effects. Of course, this is all just a SWAG on my part. Getting close now, good! I`m getting tired of typing. We last left with the delayed signal arriving at the FET`s drain. When the FET is biased "on" the signal passes through it and to a
27K resistor. This is the "mixer" that I referred to earlier. This is the point where the delayed or "chorused" signal meets the raw signal coming from pin 1 of the 4558. They both go to input (4558) pin 6 and are sent to the output jack from the 4558 output pin 7.
Of the last two clones I repaired, the first had a faulty SAD1024. I got one
for 17.00. The problem was basically easy to track down. Everything (circuit-wise) seemed to be in-check, but there simply was not an output from the BBD. [You can also easily set a clone up to use the MN3007 simply with a wire change.(the board is pre-drilled for the eight pin chip.) MCM electronics has the MN3007 for about $7.00.] The second repair was a bit more challenging! EVERYTHING seemed to be functioning, yet the chorus effect was barely noticeable, if appearing at all. Well, after alot of cursing and "take fives" on it, I found the
100pF CD4047 timing cap. to be flakey! The oscillator was still ticking away, but at the
*wrong* frequency! What a pain in the arse! I placed a silver-mica here, and....presto!
Because this "base frequency" is so important, I believe a silver-mica is the way to go in this position. Another thing that I did was to change ALL of the electrolytic caps. The parts count is only about eight or so, and although some may view this as a drastic measure, I can say that more than half of the old ones were out of spec. Over thirty years, they DO dry out and drift.
Well that is it. I hope my ramblings can help someone out there! Please feel free to add to this and make any corrections.
Have a good one, Bill Kahle ?
|
more posts...