Phono 3S build

[colin.hepburn]

Hi Richard
Thank you for the update on the the variable caps and resistor error
I have also pm Matthew about the errors as you advised
Colin

[Matthew Snell]

Hi All

How's your memory?

I'd really appreciate input on the R21/22 issue. It might have been a mistake, but we've built AND tested using 2K2 and we've got a lovely phonostage. Does anyone feel like testing it and getting back to this post?

Sorry if I'm covering ground that's been well tilled previously, but I'm new to this business!

Cheers

Matthew

[Richard]

Hi Matthew,

R21/22 value affects the higher 2 corner frequencies of 75uS 2122Hz and 3.18uS 50KHz.

It works with the cap and cap trimmer to set those 2 corner points and the result is somewhat adjustable with the trimmer.

The 75uS corner is found by adding the valve output impedance to the following series resistors and working that value against the cap and trimmer setting.

Output impedance of that second ecc83 without a cathode bypass cap is 153700. The following series resistors in your drawing are 180K and 2k2 so this totals 335700.

The 75uS corner is by far the most audible of the two and will dominate a square wave test and check of riaa curve. If you set up the amp with those resistors and tweak the cap trimmer you'll get the 75uS corner coming in nicely on target with 223pF.

The 3.18uS corner is found by working R21/22 alone against the cap and trimmer value.

Using 2k2 with that 223pF puts the corner at 324573Hz. This is clearly way off the target of 50KHz. (But it is still a small correction in the right direction compared to not having the 3.18uS correction at all. It is also probably way above what anyone is hearing. The difference between incorporating this 4th corner at 50KHz or 324573Hz may well be academic and the amp may well sound the same.)

However, let's now follow the likely instructions and set the trimmer midway, so 180pF +25pF = 205pF. With R21/22 value 2k2 those corners are now at 2313Hz (nearly 10% off) and 353072Hz (way off).

Compare that with R21/22 value 14k3 and we get a best result using 215pF for the 75uS corner at 2129Hz and the 3.18uS corner then comes in at 51792Hz. So both very close indeed.

In practice there are stray capacitances and the very highs are lost in this circuit. I found by adjusting the trimmer as shown in my circs for ecc83 and 5751 that a slightly lower capacitance value is needed to keep the high frequencies on-curve.

(The reason for the difference in trimmer setting btw between the 2 valve types is that 5751 has a lower output impedance than ecc83.)

Does any of this matter? Well yes, if you're going to incorporate the 3.18uS corner then it's as easy to use the correct resistor and set the corner at the correct 50KHz as the wrong one which will only ever set it around 300KHz!

The trimmer plays a big part in getting the 75uS on-curve and it is this that which I believe will be heard most. That's why I'd advocate you product test then supply the kit with fixed caps as a small turn of the trimmer by a builder without test gear will throw off the 75uS corner quite a lot.

[John Caswell]

Hi all,
Just to jump in here and correct Matthew and hopefully put this to bed.
The combo I just built for him had a 14K3 resistor for R21/22 so I do not know how this error has crept in.
As far as the trimmer is concerned, I take Richard's point entirely re the voltage at switch on, however my phono stage is powered by a modified Graetz bridge a la Andrea Cuiffoli consisting an EZ80 valve and a pair or UF4007, so has a slow warm up. Hope that helps clear up some confusion.
I also use the same type of bridge in my version of the 300B PP amp but use 6D22s and UF5408s

John

[Matthew Snell]

Thanks for all that (Richard, I think I'm going to have to go back to school!)

I don't know how 2K2 came to happen, but it was there in my inherited Circuit Diagrams, Parts List and Assembly Instructions. I was convinced.

Still an error though, and thanks for spotting it in time. It's official now: R21/22 is 14K3!

Cheers

Matthew