RIAA preamp

[dominicbeesley]

Hello all,

I've not been doing much but this week I've decided to attack some of the projects I've had lying around for months. One of these is an RIAA preamp. I designed the circuits myself from scratch with some help from Morgan Jones.

The circuit actually uses Russian 6п16ъ-в (not 12au7's as shown on the diagram) sub-miniature pencil triodes. The input is a cascode running from a regulated voltage source. The "hum bucker" feeds any noise from the power supply back into the "screen" of the cascade and is tuneable. It works surprisingly well. The noise of the regulator valve can be made inaudible by setting it at around mid position. The reason for having this part run from a regulated supply is in part to reduce the h-k voltage on the top valve to around 50V but also to fix the operating position which is quite critical in determining the output impedance of the stage which is used as part of the RIAA filter.

Power supply for HT is pretty simple, using an ancient choke I found in a box of junk. The LT supplies are separate regulated 400mA current sources using LM317T regulators.

The second stage is a simple gain stage and the final stage is just a buffer. At some point I aim to add another pair of triodes to have a baxandall tone bass/treble circuit.

I tested it with a sig gen and also a swept tone/lisajous both giving good results <0.5db error until about 15kHz where there is a slight deliberate lift. Also playing back a pink-noise track on a test record gives a nice straight line in audacity's spectrum plot.

Construction wise it's all thrown together into a crappy project box until I get round to either finding a suitable chassis at the tip (not easy these days with the electronics dump being guarded) or get my metal working mate to make something.

The circuits I've built as little modules on pcb material, this is a nice way of working with these little valves as circuit mods are a doddle and there's a nice ground plane to work with.

Listening it sounds great to me though my hearing does stop above about 8kHz! The only trouble is a slight hum. First up I'll try adding an extra 1k/100uF cap to the power supply. After that I'm not sure what to do.

I'd really like to have the signal lifted from the chassis earth but I'm not 100% sure how to do that and keep it safe. At present there is a star earth point.

My thoughts are either to somehow lift the input ground and feed some of it's signal to the "screen" of the cascode or maybe to lift the whole signal ground from chassis with a resistor/low value cap but then I'm not really sure how to make that safe and it will still all just end up being grounded by the output load....any ideas?

Cheers

Dom

[dominicbeesley]

Here's a few pics of it running...sorry about the quality

[audiomik]

Have you tried rotating the EI Mains Transformer (top left in your picture) to assess stray induced fields?

Also a screen between your Power Supply and the low-level active circuitry may well help

Mik

[Radio Wrangler]

One thing I spotted, after your Laplace-defined VCVS, the 47k resistance in the cartridge model... Is this the right sort of value? I associate magnetic cartridges with wanting to work into about a 47k load, which isn't the same thing as being a 47k load. I'd expect their source impedance to be smaller. The load isn't supposed to be a match, but a resonance damper... I think.

David

[G8HQP Dave]

Yes, an MM cartridge needs to see about 47k - not be about 47k. The resistive part of the MM impedance would be much smaller than 47k.

Given regulated supplies and DC heaters, hum could come from:
1. unscreened signal wiring
2. poor grounding - a star ground will not solve all hum problems, it could create some instead. It is vital to keep PSU charging pulses well away from the signal ground yet most people route them through the star point!

[fetteler]

Hi Dom,
Thanks for posting such an interesting circuit. I've often pondered an RIAA preamp using those little pencill valves - well done!
I agree with the others re. the cartridge model, messure the dc resistance of your cartridges couls to get a better idea of a value for the series resistance.
I also agree with the comments that your hum is likely to be caused by layout and screening considerations. A little tweaking and you'll soon be there

Cheers,
Steve.

[dominicbeesley]

Thanks lads,

I thought I'd replied to this earlier but must have forgotten to click on the post button. Thanks for the interest.

The 47k series resistance is a red-herring - that was me not resetting values when I'd been "playing" in LTSpice. I've measured the _real_ response (with a pink noise recording) and its more or less bang on so I'm not worried about that part of the model.

The HT supply for the output and 2nd stage valves isn't regulated just the input stages - I suspect a bit of smoothing there won't go amiss.

I have noticed that the input valves are quite sensitive to hum - a finger near them is audible so first up will be to screen them - not sure the best way but for a starter I'll just roll some metal into a loosely fitting shield and solder it to the ground plane.

I'll try a bit more with screening and HT smoothing but I'd still like to know a "kosher" way of lifting the grounds - if there is one.

On the plus side I can actually - even with my ruined hearing - hear the difference between this and my previous op-amp based amp...plus it just makes me happier that I made and designed it myself from my play pile!

Cheers

D

[Radio Wrangler]

If you're using equipment you're building yourself, don't lift grounds. Use balanced signals and do the job properly. It solves a lot of problems and doesn't compromise safety aspects. Liifted grounds can be done safely, but the liability aspect is harder to handle.

David

[dominicbeesley]

No thanks, I thought of that but it would mean re-rewiring the turntable and it not being as easy to switch between my various amplifiers.

The question still stands. There must be a way of doing it on the first stage at least without a major redesign. How about something like the attached. I'm fairly sure the common mode noise on the ground must be fairly low impedance.

D

[fetteler]

Dom,
Are you sure that the hum is coming vis the input? Have you tried shorting (or putting a resistor across)t he input to ground and seeing what happens?

Sorry if I'm suggesting egg sucking here!!

Cheers,
Steve.

[dave cox]

Regarding ground lift, I usually wire a seriously rated bridge rectifier across the lift resistor - something I hope won't open before the fuse or wiring !

dc

[dominicbeesley]

Thanks lads, I'm sure some of it is coming from the input. shorting the input reduces it significantly, the rest is coming from the PSU which is easier to attack.

I like the diodes/rectifier idea - I suppose here I could put that at the input as that is capacitor protected....

D

[daviddeakin]

Originally Posted by dominicbeesley

I like the diodes/rectifier idea - I suppose here I could put that at the input as that is capacitor protected....

This is indeed a common and acceptable method of cutting down ground-loop hum without violating mains safety. Here's a nice diagram!
http://sound.westhost.com/earth-f4.g...m/earth-f4.gif

[dominicbeesley]

Hi david, I'd be interested in seeing that but the middle of the link looks to have been replaced by "...". Please could you try posting it again.

I've gone a bit quiet on this as I just bought an SDR radio and have been busy making lots of (mostly hopeless) aerials....

D

[fetteler]

A little editing of the nonsense generated after pasting the link into my browser yielded this

Cheers,
Steve.

[dominicbeesley]

Aha, thanks. I think I'll give something like that a quick go first but I'm not sure I like the idea of the whole thing floating...

D

[fetteler]

Yes, I agree. See if it solves your immediate hum problem and allows you to listen to the true sound of the amp and then you can have a leisurely think (to music ) about a long term solution.

Cheers,
Steve