They are more precise than normal 5% or 1% resistors, in those resistors the sound that will be generated depends on those resistors, i think they are suppose to represent octaves, like their value is doubles each time 5k to 12k to 20k 40k and so on.

I didnt have 5k ones, so i did solder 10k in parallel, which is a good solution 10k is a pretty usual E value, the others are not, so not that easy to get.

To be more precise, I think those resistors are there to convert some digital sample values form the EPROMs into analog values, which then are amplified and a bit filtered so you can hear a nice sound.

So are they supposed to be exactly 5k, 10k, 20k, 40k, 80k & 160k?

Mine are not and they are the Metal Film Resistors - Through Hole 1/10watt xxKohms .1% 25ppm in the offical Nava Mouser BOM.

Having the octaves exact on the crash and ride doesn't make a lot of sense when you have a potentiometer thats adjusted by hand for tuning

I guess thats why in the build guide it says "RA9B 10kohm must be a resistor you measure at the beginning of assembly." Cause the 10k ones are rare?

But I found these http://au.mouser.com/ProductDetail/Vish ... OoEJyq4%3d

Yes, 5... The best you can get They deliver sounds from 27C, but this is a point of characteristics and comparison also, I suppose.
In theory should be exactly as mentioned.
I had no problem to get 10k by Vishay.
Because values should stay the near as possible. Thats why.

cheers

if you refer to the potis to adjust the hihat?
They have different purpose, one is standart for volumn, the other two are simply for timing purposes int he circuit, they convert the trigger pulse form the sequencer to hihat sounds output timings.

On the other eprom circuits you do have tuning potis, you could also tweak the hihat circuit to do so, with resistor R493 or R492 should probably tune something as well, or put another capacitance on C160 that will alter the clock speed with which the epprom is read out, makes for some interesting sounds in my opinion, but will be a bit tricky to implement that as a mod.

The RA9A-F resistors form a 6-bit DAC. The hi-hats share one audio sample and hence one DAC; the ride and crash use 2 DAC's each: one DAC to play back the audio sample and one DAC in their VCA's.

On the ride and crash the clock speed of the digital audio playback is variable (this is what the tune pots vary). Because changing the clock speed changes both the pitch and the duration of the sound, the ride and crash need variable duration for their VCA envelopes. The second DAC's on crash and ride are used to create a VCA envelope with a duration dependent on the clock speed.

The resistors values in the DAC's have to be as accurate as possible (especially for the low values) because they affect linearity of the output. Google 'weighted resistor DAC' if you want to know more. Also see viewtopic.php?f=24&t=802&start=30#p7283

Nice one.

Cheers guys