best place and value for volume pot

[pcdata76]

Hello

I built updated version of JLH class-a amplifier(10W@6ohm design, MJ15003's for output, 470uF feedback cap and 1 uF input cap) and Rod Elliot's preamplifier (the only difference is i used bc309c transistors instead of bc559) for doz.

Both amps works perfect. But since this is the first power amplifier which i have built, i can't decide the proper value & place (even type log or linear) for volume pot.

The best solution looks like the building the same preamplifier with smaller (or unity)gain and putting the pot between preamps, but if putting the pot before jlh power amp will not make any considerable problem then it will be more convenient (means easy ) What will be the best solution?

I'll buy ALPS RK27 series pots after deciding the value.

[jaycee]

22K if you can get it. "Standard" is 47-50K. It should be a log pot unless you use a "law faking" resistor (Rod has an article on that I think). Put it between the preamp and the power amp.

[pcdata76]

Thank you for response.

I can't get quality brand 22k pot so i should choose 50k.

Is there any disadvantage of using 10k pot? (except very little attenuation on max. setting of the pot)

[Xoc1]

I find I get good results with a linear pot and a law faking resistor. This will give a better channel balance than most log Pots
Rob Elliot has a good article on this on his site.
ESP - A Better Volume Control
He recomends a 100K pot and a 15K law faking resistor. This is probably a good place to start. One advantage of using a law faking resistor is you can change the value to give you more control at your preferred listening level. For instance reducing the value of the fixed resistor would give you more control at lower levels.
I use a higher value fixed resistor on amplifiers used for PA, this gives me more control at high level.
The volume control should be close to the power amplifier input for best performance. Try to avoid driving long cables after the volume control.

[pcdata76]

I make several combinations with 100k linear pot + log faking resistor. With 15k fixed resistor, there will be no sensible change (Usable volume levels for daily listening are in the very very small rotation on the pot so it's very hard to adjust volume in lower end. To increase sensitivity on the lower end (similar sensitivity like commercial amps), i have to put a fixed resistor somewhere between 1-2k\ but i know that this is not a good thing and makes middle settings of the pot almost ineffective( no volume change by turning the pot on middle zone)

I think something is wrong, I used BC309C's instead of BC560/BC212/2SA970 written on JLH web site for input transistors. This can create the problem?

[Peter Pan]

This is kind of off topic, but this whole methodology of simulating a logarithmic response from a pot using a "faking" resistor is very interesting to me, for another application. As a guitarist, I've both built and worked with a number of store bought volume pedals, and in many cases I've not been happy with the physical response of pedal position vs. volume. In a volume pedal, the full pedal movement won't turn the pot more than about 1/3 at best, and it's up to the supporting electronics to compensate with extra gain. While that isn't an issue, the response of the pot often is. Now I'm thinking this law faking resistor might be just what is needed. I might have to use a trip pot for that faking resistor and have another to tweak the gain of the subsequent pre-amp circuit to get the desired effect, but I've a suspicion that a linear pot with the right value of faking resistor might just do better than even a good quality log pot in this application. If anyone is interested, I'll let you know what I find in a separate thread.