Tube Amps / Music Electronics
For current discussions, please visit Music Electronics Forum. New: view Recent Searches.
New: visit Schematic Hell!
The sunn still shines online!

Listen to great tunes streaming live right now!

ampage archive

Vintage threads from the first ten years

Search for:  Mode:  
previous: Gord Hi,I have tried bui... -- 1098326632 View Thread

Re: TS9: Ge diodes = low output

11/16/2004 8:11 AM
Mark Hammer
Re: TS9: Ge diodes = low output
A stock TS-9/808 is not all that loud anyways. There IS some audible increase in output level, but none of the clones I have successfully built were drastically louder than the bypass unless you kept the level and drive controls up over halfway. That's not a weakness, just a characteristic.  
Unfortunately, if you switch to GE diodes, several things happen. First off, GE diodes can vary in clipping threshold from UNDER 200mv to over 280mv in my experience, while standard 1N914/4148's can vary from under 500mv to over 650mv. The odds are reasonably good that the signal level your clipping stage clamps the output at is less than 1/3 the original level. Given that the clipping stage is followed by a lowpass section that reduces the apparent volume, and a tone section that tends not to add much gain/level, it is not surprising that you would need to dime the volume control, even at max drive and max treble.  
The other thing that happens is that, with a reduction in clipping threshold, it simply isn't a TS-9 any more, but rather a poor attempt at a fuzzbox. I'm not saying this to diss anything. Rather, it's a bit like pouring ketchup on your salad and wondering why it doesn't taste much like a salad OR a hamburger. The TS-9 circuit is intended to be capable of producing modest, and EVEN clipping across the spectrum. Part of that formula lies in matching the gain and signal shaping to the diode properties (and vice versa). Change the diode properties and you've radically altered the balance within the design.  
The more puzzling thing is why you don't restore level when using several GE diodes to replace a single 1N914. In principle, this should work fine, although you'd want toverify the values. As noted, although GE diodes *theoretically* have a voltage drop of HALF what an SI type does, it is not uncommon for me to find 1N34's and 1N60's and unmarked GE types reading less than 200mv and SI types reading more than 650mv. In which case, it takes MORE than three GE diodes in series to result in the signal NOT being clamped at a lower level than a single pair of SI diodes would nail it.  
The other thing I wonder about is whether you'd actually hear anything different tonally from the GE experiment. The TS-9 is all about producing a more even tone by trimming back on low end at input to not over clip bass notes, and trimming back even more on the high end post-clip to compensate and yield a balanced bottom-to-top-note volume level. Whatever nuances might exist in the waveform and harmonic balance between GE and SI clipping may well be lost in process. That'snot to say there ARE no differences in how these might sound under other circumstances, but this isn't the circumstance to exploit those differences and make them audible.  
What I HAVE found useful is experimentation with number and balance of SI diodes. The BOSS SD-1 is ostensibly the same circuit as a TS-9 except that it uses a 2+1 back to back arrangement (i.e., three 1N914's). The added diode raises the clipping threshold for one half cycle. This increases the overall output level but also increases the dynamic responsiveness of the pedal - pick a little harder and you'll hear the volume difference - beyond what the stock TS-9 can deliver.  
I've made two clones that tinker with this, both using Francisco Pena's TS-9 layout available at One simply allows me to shunt the 3rd diode with a toggle and switch back and forth from TS-9 to SD-1 mode. The other uses a pair of diodes in TS-9 fashion but sticks a 10k pot in where the 3rd diode would go and "warps" the symmetry of the clipping in a continuously variable manner. (Kudos go to Jack Orman for the excellent work on clipping warp controls that he has posted over at I prefer the latter over the former.