Ashbory Mk I modification
I found that, relative to my other basses, the output volume of the Ashbory is on the low side. I decided to increase the gain.The information I needed for this was obtained from Largesound.net, where there's an article on replacing the LED, accompanied by a circuit diagram and component layout.
The Ashbory's electronics are pretty simple. There'a a tone-shaping section which is a Baxandall-type network feeding the volume control, and after that there's a gain block around the TL071 op-amp.
In the circuit diagram on Largesound, R5/R3/C1 and R6/C2 form the gain determining feedback resistance (Rf) and,for want of a better term, ballast resistance (Rb). As there are capacitors in the circuit, there is a frequency dependency to the gain block. This wouldn't matter if the gain was determined by the ratio of feedback to ballast resistance (which it does when the op-amp is used in inverting mode), but in non-inverting mode, as this is, the gain is 1 + (Rf/Rb), so you can't just divide the values of R6 and C2 by ten, for example.
I took five sample frequencies - E, A, D, G, and G' (or at least the integers just by them) and found that the gain at each frequency was
| E | 4.02 |
| A | 3.98 |
| D | 3.85 |
| G | 3.65 |
| G' | 3.11 |
So there's some treble roll-off. I knocked up a spreadsheet and played with some values for R6 and C2 (a little less complicated than playing with the feedback path). Values of 15k and .22 uF looked promising, more than doubling the gain and with just a little less treble roll-off. To check out the extent of the treble roll-off, I tried some higher frequencies - 1k, 2k and 4k.
| 1k | 2.28 |
| 2k | 2.15 |
| 4k | 2.07 |
The values I'd come up with still looked good. Here's the new gain and the ratio of new gain to old gain, and how much the gains vary taking the gain on E as the base value:
| Freq | New gain | Old gain | Ratio | Ratio variance |
|---|---|---|---|---|
| E | 8.95 | 4.02 | 2.22 | - |
| A | 9.03 | 3.98 | 2.27 | +2.09% |
| D | 8.87 | 3.85 | 2.30 | +3.59% |
| G | 8.49 | 3.65 | 2.33 | +4.56% |
| G' | 7.22 | 3.11 | 2.32 | +4.40% |
| 1k | 4.97 | 2.28 | 2.18 | -2.20% |
| 2k | 4.57 | 2.15 | 2.13 | -4.32% |
| 4k | 4.36 | 2.07 | 2.10 | -5.61% |
Ideally, all the ratios would be identical.
I could get higher gain by decreasing R6 to 12k but this gives a bit more variation on the gain front:
| Freq | New gain | Old gain | Ratio | Ratio variance |
|---|---|---|---|---|
| E | 9.76 | 4.02 | 2.42 | - |
| A | 9.99 | 3.98 | 2.51 | 3.60% |
| D | 9.94 | 3.85 | 2.58 | 6.60% |
| G | 9.65 | 3.65 | 2.64 | 9.04% |
| G' | 8.41 | 3.11 | 2.70 | 11.57% |
| 1k | 5.90 | 2.28 | 2.58 | 6.63% |
| 2k | 5.44 | 2.15 | 2.53 | 4.43% |
| 4k | 5.18 | 2.07 | 2.50 | 3.04% |
So, what gain would I need?
I only really needed a rough idea so I switched the trusty DVM to mV and connected the first two basses that came to hand, a passive Peavey Grind and an active Antoniotsai. Volume set to max, ditto the tones on the Peavey, the active tones and pickup blend set to centre, and the output from a normal pluck was of the order of 100mV (lots of plucks done) for both. Did the same for the Ashbory and the output was of the order of 50mV, so I need to approximately double the gain, which is what I'll get with 15k and .22uF.
The operation didn't take long. The replacement 1/4W resistor was quite a bit bigger than the original. I might replace it with a metal film one at some later date.
I recorded the Ashbory prior to modification, the Peavey on the same gain settings on the input device (a Tascam US428), and then the Ashbory post-mod. The volume is definitely up and the tone sounds the same to me (though I have cloth ears).
Recordings done to Audacity through a Tascam US428, then converted to 192k MP3s. All on the same gain settings which were about -6dB on the pre-mod Ashbory. I didn't set the gain to suit the higher volume instrument so it's been driven to clipping.