The Repair Bench: Build Your Own Clone 5 Knob Compressor
- May 26, 2011
- in Guitar Kit Builder - Repair Bench
In this episode of The Repair Bench we are troubleshooting a clone of the famous Ross Gray Compressor called the 5 Knob Compressor and sold by Build Your Own Clone (BYOC).
We received this pedal from a customer who had purchased it, already built, on eBay. Apparently it had never worked for the original buyer/builder, but it had been sold through eBay to a new owner, who negotiated a reduced price when he learned that it didn't work. It was then given to us in the hope that we could get it working. Learning how to troubleshoot electronic circuits is one of the handiest skills to have if you are building or modifying amplifiers or effect pedals. At some point you'll power-up a circuit and find that it doesn't work, and then what? For this reason we write "The Repair Bench" section of Guitar Kit Builder about our own troubleshooting of amplifiers and other devices, to pass along to the reader the thought process, tips and techniques of troubleshooting electronic equipment.
BYOC 5 KNOB COMPRESSOR
The BYOC 5 Knob Compressor is essentially a Ross Gray Compressor clone with added Attack, Blend and Tone controls. The Tone control uses a pre-existing high-pass filter in the Ross circuit and makes it variable to better match the pedal to single coils or humbuckers. The Blend control varies the mix of compressed and dry signal from 100% dry to 100% compressed. The printed circuit board for this clone can use either a BA6110 or CA3080 voltage controlled operational amplifier chip. The pedal uses a triple-pole double-throw (3PDT) footswitch for true bypassing with an LED indicator.
First we should say that nothing is potentially as aggravating as troubleshooting a device that has never worked before. It exponentially increases the number of things that could be wrong. Our initial tryout of the pedal showed that it did produce an output when engaged, but it didn't sound like anything particularly good was happening. The controls seemed to have some effect, but the primary compression quality wasn't strong enough to believe that the pedal was operating properly. In evaluating a compressor we're particularly listening for two qualities. The first is that a hard-struck note is noticeably reduced in initial volume, compared with the compressor off. This is because the compressor should rapidly reduce the volume of a loud sound.
Secondly we're listening for greater sustain of a held note. This is due to the compressor raising the volume of a soft sound. Neither of these qualities was significantly present in the initial tryout. Next, we opened the case to look at the build quality, and this is where the likely problem became apparent. This kit appears to have been built by someone with little experience or skill in electronics construction. This is not particularly surprising, given that the market for building these kits is musicians who may or may not have any experience with electronics. Reviewing the pedal showed it to have many questionable looking solder joints, sloppy wiring dress, and some heat damaged components, at least externally. Some examples of these are shown in the following photos.
As the upper photo shows, the wire leads extend well above the circuit board, and are long enough that they can easily end up touching each other, either now or in the future. The lower photo shows the external heat damage to a couple of capacitors, and the socketed transistors, which are a reliability concern.
TROUBLESHOOTING THE 5 KNOB COMPRESSOR
Since the kit had never worked we considered the possibility that the printed circuit board had an error or that some other "fix" was needed that perhaps had never been done by the builder. So we visited the BYOC website to take a look at the kit instructions and schematic. In reviewing this material it was apparent right away that the board in our kit was different from the instructions. We assumed that the kit had been revised at some point, and a quick post in the helpful BYOC Forum proved this to be correct. We were pointed to the instructions for the original kit, and also did some searching to see if any fixes were needed for our board, but none were. We tried adjusting the internal trim-pot according to the instructions, but while the control had some effect, it was simply between bad and worse.
We used the instructions to do a quick check of the board - checking that diodes, transistors and the BA6110 chip were inserted correctly, which they were. We did notice that the transistors seemed a bit loose in their sockets, but all were inserted correctly. We then checked that the resistors and capacitors were the right values in the right places, and that the potentiometers were connected correctly. These all checked out as good, though we noted that a couple of capacitors had external melt-marks from a soldering iron touching the outside. We couldn't tell at this point whether the capacitor was damaged or not, but the melting appeared to be external only. At this point we surmised that the kit builder, while unskilled, had been paying attention to the instructions and been careful in placing each component. Our assumption was that the likely cause of problems was the build quality, most likely one or more cold solder joints or sloppy, loose wires making spurious connections.
We got out our magnifying lens and started going over each connection on the board, re-soldering any joint that didn't look good, and trimming back lead lengths to make sure that none were touching. The solder pads on this board are relatively small, and some kind of magnification and good lighting is very helpful to spot problems. For the joints that didn't look good, we just gave them a short touch with the soldering iron, to re-melt the existing solder, and added a little bit if needed. Any solder "blobs" were removed and the board traces checked carefully to make sure of no cross-trace connections. Wires were trimmed back to a length that could not cause a shorting problem. We also soldered the transistors into their sockets. This was to remove the uncertainty of a bad connection during troubleshooting, and a possible reliability issue in the future.
With everything looking good we again fired up the pedal and probed the signal path with an oscilloscope. Now we could see that we were definitely getting compression because the scope trace showed that the signal was being limited in amplitude, compared to when the pedal was off. Plugging the pedal into an amplifier confirmed that the pedal was now operating correctly, and producing good usable effects. We adjusted the internal trim-pot and it now had a definite range where it sounded correct, and fit the description of the instructions. We now had a working 5-Knob Compressor. As a final step we replaced the power input jack, which had a cracked nut that prevented it from being tightened. This was a simple de-solder and re-solder type of fix.
So what was the specific problem with the pedal? We don't know, other than it was a bad connection within the construction of the pedal. It's clearly better to use careful construction techniques as you go, rather than have to try to clean things up later. A more experienced builder would have known to do this, but new builders are often too excited to be careful. Our customer was very happy with the repaired pedal, and said that it now sounded the way he had hoped it would when he bought it.