Given the amount of time I've already invested in this project and the growing realisation that I'm going to have to invest a whole lot more if I want to get to the end of it, I started to turn my attention to making my experimentation a little bit easier.
See here a sort of test jig I created to allow me to tidy up the connections. It still looks a little awkward here as I plugged in my Mark 1 perfboard circuit, but it does what it was intended to do.
If you peer at the picture you'll see that the jig is simply a block of wood with L-shaped brackets attached. I've drilled holes in the brackets to attach various sockets and dials. Furthest left is a power socket followed by a 6 throw switch, potentiometer and two audio sockets. These have then been wired up to various terminal blocks which in turn can be connected to whatever needs to use them.
It's a fairly crude piece of kit, but it's working surprisingly well, allowing me to concentrate on the circuit with a free hand to scratch my chin.
Having done everything I could think of to debug my Mark 1 circuit, I started to wonder whether any of the other versions of the circuit on the internet were exhibiting the same sort of buzz. Maybe, just maybe, this could give me the insight I needed to crack this problem.
I returned to Scott Helmke's mint box buffer which had been a huge inspiration for me starting this project in the first place. On closer inspection I spotted a glaring difference in his design and mine; I had biased my 2n5457 JFET to take the audio output from the Drain (as I'd thought was the conventional pattern); Scott's design takes the audio output from the Source. How had I never spotted this before?
Could this be the missing magic I needed?
Long story short... no.
See above a short video of a circuit I pieced together based upon Scott's. About the only difference I can recall making was to reduce the pull-down resistor to 680K. This resulted in a quieter signal, but also a less noisy one. Can you hear the confounded buzz? Yep... me too!
On reflection, I feel that Scott's circuit is more light-weight and elegant than my Mark 1, but at my hands, it suffers the exact same circuit noise problem. I experimented with different bypass capacitors and biasing resistors, but I couldn't get it to sound any better than what you can hear above. DAMN!
As I pondered the problem I started to pluck out a tune on the banjo. Above is the short idea I came up with. I must be growing a bit soppy in my old age; This is far too happy a jig! ;-)
Although there are no words written down at the moment. this tune tells a tale of young love. Beautiful! For those who like my more sombre ballads... you have nothing to fear... this all ends in tears. Ha ha.
If I ever summon the courage, you'll get the whole story in a comic sometime.
I love the Internet. It is a haven for truth and lies in equal measure. The challenge, as always, is to try and figure out which is which. Earlier this week I was researching my Piezo Buffer when I stumbled across a curious site curated by Ivor Catt.
It's probably the ugliest site on the web. As I scrolled through seemingly bottomless pages of poorly formatted text I couldn't help but feel that I'd stumbled across something special. Had I unwittingly logged on to some ancient monolithic computer? Was I experiencing the internet version of time-travel? Everything felt dusty and complicated. The more I read, the more confused I got. There seemed to be hints of conspiracy intermingled with arcane electronics theory in a rambling prose that spun in ever decreasing circles. Brilliant!
Please go read it for yourself and when you understand it... come back here and explain it to me. I haven't got a bloody clue what it's all about!
And there you have it. I leave you as confused as I feel at the moment. Would it be bad form to give up now?
Oh dear... what have I gone and done...