Anything obviously wrong with my mic pre amp ?

[nigelwright7557]

Has anyone got any comments on my mic pre amp cct ?

If I short out the AC input to it it puts out a sine waveform which looks modulated at quite a low freqency.

The power supply is possibly over decoupled with 100uF electrolytics at 220Volts. The output of this stage goes into the other half of the ECC83 which is a copy of this cct. By the the time the noise gets through both stages it is quite big 200mV.

The sound this signal causes is almost bearable on the output but is a nuisance if I turn the input volume down and the amp gain up.

The heater is driven from a DC source.

The pre amp is almost there so I have tried very hard, I just need a little help to sort out the last bug.

I enclose a cct diagram of the first stage.
Thansk in advance for your help and/or suggestions.

[nigelwright7557]

Here is the second stage with the PSU.

[astouffer]

Your power supply doesn't look quite right. Why did you go with half wave? I'm not sure if that 100nf cap before the diode is going to do much of anything. Oh and what software was used to draw that schematic

[nigelwright7557]

Originally posted by astouffer
Your power supply doesn't look quite right. Why did you go with half wave? I'm not sure if that 100nf cap before the diode is going to do much of anything. Oh and what software was used to draw that schematic

The 100nf kills the diode rectifying glitches.
They were huge without it.

Half wave just reduced the number of components, space was a bit tight.

I use PCBCAD30 which I got off ebay.

[kevinkr]

Do you have a scope? Can you tell us more about the low frequency noise you are hearing - like amplitude, frequency waveform?

I notice that you have an unbypassed cathode bias resistor on the input tube so you in effect are gaining up the johnson noise of that resistor by mu+1 - I wouldn't be surprised if it was substantial.

I don't think placing a 15K (?) pot ahead of the input is a good move from an SNR standpoint either with a microphone as a source, probably better to use a 100K - 250K pot after the first stage.

How much ripple on that half wave supply? Microphone pre-amps need quiet, well filtered dc, and you may not have that..

Also ac mains voltage tends to undulate at subsonic frequencies as a function of constantly varying loads on the grids - take a look at your supply with a simple resistive load drawing the same load current as you active circuitry and make sure this is not the case - if this is the case you need a much longer time constant in your supply circuitry. This is one of the reasons I originally started using regulated supplies.

Download ltspice switchercard 3 from Linear Tech (it's free and very powerful) and get the tube libraries from people here, will allow you to analyze your circuits and post schematics that us hoary old folk can see....

[nigelwright7557]

Originally posted by kevinkr
Do you have a scope? Can you tell us more about the low frequency noise you are hearing - like amplitude, frequency waveform?

I notice that you have an unbypassed cathode bias resistor on the input tube so you in effect are gaining up the johnson noise of that resistor by mu+1 - I wouldn't be surprised if it was substantial.

I don't think placing a 15K (?) pot ahead of the input is a good move from an SNR standpoint either with a microphone as a source, probably better to use a 100K - 250K pot after the first stage.

How much ripple on that half wave supply? Microphone pre-amps need quiet, well filtered dc, and you may not have that..

Also ac mains voltage tends to undulate at subsonic frequencies as a function of constantly varying loads on the grids - take a look at your supply with a simple resistive load drawing the same load current as you active circuitry and make sure this is not the case - if this is the case you need a much longer time constant in your supply circuitry. This is one of the reasons I originally started using regulated supplies.

Download ltspice switchercard 3 from Linear Tech (it's free and very powerful) and get the tube libraries from people here, will allow you to analyze your circuits and post schematics that us hoary old folk can see....

Many thanks for your input.

1/ Adding bypass capacitors makes minimal difference.
2/ I tried adding more smoothing both before and after the resistor in the smoothing chain and it amkes no difference.
3/ The input pot is 100K.

I am beginning to think the valve cct is fine and there must be hum getting in somewhere near the front end. The valve stages do have a gain of 10 each so this is a total gain of 100 through the mic path.

I have just used hook up wire on the inputs so I might replace it with screened wire and see what difference that makes.

[rdf]

Almost sounds like motorboating. Is there room to decouple the first and second stages with a resistor and second bypass cap as a test? BTW, Spice suggests an unloaded gain closer to 1000 than 100. It approaches the latter when driving a 10k load typical of computer sound cards but the output 12ax7 is being seriously abused by then.

[nigelwright7557]

Originally posted by rdf
Almost sounds like motorboating. Is there room to decouple the first and second stages with a resistor and second bypass cap as a test? BTW, Spice suggests an unloaded gain closer to 1000 than 100. It approaches the latter when driving a 10k load typical of computer sound cards but the output 12ax7 is being seriously abused by then.

The load is the input to an amp which is around 47K.

It doesnt need much hum to be amplified by a hundred to become significant. 5mV becomre 500mV !!!

I noticed the hum gets better or worse depending on where the heater wires are in respect to the signal wires.
The heater supply is 12 volts DC but is not regfulated just smoothed. I will try to get some screened wire tomorrow for the heater supply wires.

[kevinkr]

Originally posted by nigelwright7557




The load is the input to an amp which is around 47K.

It doesnt need much hum to be amplified by a hundred to become significant. 5mV becomre 500mV !!!

I noticed the hum gets better or worse depending on where the heater wires are in respect to the signal wires.
The heater supply is 12 volts DC but is not regfulated just smoothed. I will try to get some screened wire tomorrow for the heater supply wires.

Ripply dc is worse than ac for heater supplies IMO. This is because the sawtooth is rich with harmonics not present with AC heating or regulated dc.

For this sort of application you should also float the heaters on some percentage of the available HT using resistors and a bypass cap to ground. Typically something like 40 - 50V will suffice.

[nigelwright7557]

Originally posted by kevinkr


Ripply dc is worse than ac for heater supplies IMO. This is because the sawtooth is rich with harmonics not present with AC heating or regulated dc.

For this sort of application you should also float the heaters on some percentage of the available HT using resistors and a bypass cap to ground. Typically something like 40 - 50V will suffice.

If the heater is left floating the hum gets worse.
I currently have tied one side of the heater DC to ground on the HV.