Phono amp - need less output!

[Top Cap]

Hi, Can anyone tell me the best way, IE component values, to reduce the output of my phono amp to my PC soundcard? No problems running into the valve amp (50k input) but into the PC soundcard with volume setting at minimum, the signal is almost hitting the max point. Both channels are identical of course and consist of a cathode follower feeding the phono socket via a 2.2uF cap and a 1k resistor. The phono is also connected to ground by a 1M. I do not know the impedance of the sound card (Bet Kat knows it) and I can only guess at 47k?
So the plan would ideally be to have a 2 pole toggle switch so that in one position I get full output but in the other say 20db reduction?

Maybe just another resistor in series between the 1k/1M junction and phono would be better?

Perhaps even splitting the cathode resistor and tapping down on this?

Many thanks for any ideas.
Les

[paulsherwin]

The simplest solution would be to replace the 1M resistor with a pot and connect the output to the slider. You could also use a preset in this position if you don't need to adjust the output regularly.

1M seems a rather high value and any pot between 100k and 1M should be OK. It's also unclear to me what the 1k resistor is doing though it won't do any harm.

Paul

[Tim]

Hi Top cap.
Increasing the the value of R24 to 1 Meg will give half the present output in the circuit shown, so try 100k.
Mr Sherwins idea of a preset or pot is good. Tapping the cathode resistor might work too. I s'pose the output voltage would be fairly stable too, but the input impedance of the device you are feeding into may have some effect(as R22 is fairly high) unless isolated by a suitable capacitor.
As regards switching make R24 1 Meg and switch in say 220 k in parallel.
It's been a long time since I messed around with this sort of thing.
I have found the best way to sort out this sort of thing is to use the empirical method!! (And a scope!)

[Top Cap]

Hi Paul, well the amp is a chinese copy of the WAD (now defunct) phono amp except that I used a different valve for the follower. I did wonder about the 1k myself, the 1M I think was to bleed away any voltage on the phono socket from the capacitor at switch-on if un-connected. I was more concerned with the effect on freq response and is why I thought that a reducer in the cathode circuit would help keep things as they are. (see attachment that would give me about -10dbV reduction).

I had given thought before to using two separate output presets as I found the channel balance was a little out and I had to select ECC83's to get it right. Certainly there is space for two presets and I take it that I can feed one channel to get the desired output then parallel the inputs and adjust the other preset for minimum (null) output across the two phono signal pins. At least it would show up any phase errors .

It might be fun to swing the input freq through the audio range and see how well matched or un-matched they are .

If I go the preset route then I think I will dump the 1k resistors unless someone knows something to the contrary.

Best regards
Les

P.S. I wonder if the 1k resistor is to help impedance differential? Xc falls from 723 Ohms to 4.8 Ohms between 100 Hz and 15kHz but with the 1k resistor Xc+R varies from 1.234k Ohms to 1.02k Ohms over the same frequency range. Cannot otherwise see any reason for the 1k !.

[oldeurope]

Hi Les,
hm, I think this is killing the advantage of the kathode follower. (Low impedance output.)
Normally the output impedance is 1/S (plus 1K here) but after your attenuator it is somewere above 30K Ohms.
Question: How many Volts PP AC and how many volts DC are at the kathode?

Kind regards,
Darius

[Top Cap]

Hi Darius, a DMM shows around 80V on the cathode of the follower and there is 80V on the anode of the previous directly coupled stage. I have left the expected test voltages that WAD gave and it does show the cathode being a bit low. I expect a lot of this is due to my use of a different valve here namely the 6N1 as I had a few spare ones (new) in the component store.

The DC heater supply (just measured as 6.26V) sits on a high voltage of 44V to further reduce hum from that area, the result is a preamp as quiet as a mouse.

Also I added an extra 22k resistor in the main HT line, it is 250V before the resistor and 182V after. But for a phono amp I thought that this should be ok, it sounds very nice indeed, just a bit too much amplitude for the PC soundcard.

[oldeurope]

Originally Posted by Top Cap

Hi Darius, a DMM shows around 80V on the cathode of the follower and there is 80V on the anode of the previous directly coupled stage. I have left the expected test voltages that WAD gave and it does show the cathode being a bit low. I expect a lot of this is due to my use of a different valve here namely the 6N1 as I had a few spare ones (new) in the component store.

The DC heater supply (just measured as 6.26V) sits on a high voltage of 44V to further reduce hum from that area, the result is a preamp as quiet as a mouse.

Also I added an extra 22k resistor in the main HT line, it is 250V before the resistor and 182V after. But for a phono amp I thought that this should be ok, it sounds very nice indeed, just a bit too much amplitude for the PC soundcard.

Hi Les,
400mVpp is ok for a sound card, silli question did you use the mic in or the line in? (Line in is the blue i think.)

Darius

[Top Cap]

Hi Darius, Yep it was the line-in as can be seen by the volume panel attachment. That's the position I have to use and the Sony Sound Forge software struggles to contain it, it is just a fraction from clip level.

I just checked the cal of my scope and sure enough the AC peak to peak, playing a loud old 45 gives me around 400 to 500mV.

I have a similar problem with overdrive when my mate brings his computer round here and it has a higher quality soundcard in it with which he masters CD's then reduces the bit rate to make them compatible with what we buy!

He does not like the phono amp driving his soundcard so much but he has much higher head room, as he calls it, at the higher bit rate so he can tolerate it better.
He likes to use my amp and transcriptor as he thinks the quality is so good for getting the best out of vinyl. It just gives out too much
Les

[Top Cap]

Darius, I just checked the output from the phono socket after the DC blocker and I find it to be much greater, around 1.5V p-p. This may be due to my scope not liking 40V dc on its AC coupled mV ranges.
Les

[Merlin]

You could try applying feedback from the cathode of the follower to input?

Or split the cathode resistor of the cathode follower (say 120k top, 33k bottom) and take the output from there at an attenuated level. That should also increase the bias on the cathode follower a bit too.

This amp doesn't appear to have RIAA correction??

[Top Cap]

Hi Merlin and many thanks to everyone for the helpful comments. I tried the simplest thing and that was an external attenuator, placed on one channel for evaluation against the other. I convinced myself the attenuator was taking off the crisp top end of the signal. This top cut also seemed to be felt if I increased the 1k resistor to 47k so I tried the cathode approach. I replaced the 150k with a series 56k/100k combination and the results are most satisfying.
The soundcard input slider control now gives a correct level at mid setting rather than an excess when virtually just off the bottom end stop.
Sorry for misleading you Merlin, I did not enclose the full circuit but attach it anyway plus a photo of the unit sitting on its companion power supply.
I did notice a component numbering error on the circuit which explains why R22/R23 has suddenly become R23/R24.
Les

[oldeurope]

Hi Les, thanks for the schematic.
Question: Why do you use some 1% tolerance components there?
Darius

[Top Cap]

Hi Darius, The original WAD amp used 1% in places, presumably they are part of the RIAA characteristic circuitry. In fact one resistor is 316k and was a value I could not find so I settled for two series connected 158k that were available from RS. Perhaps one day I will get the use of a swept audio spectrum analyser just to see how close to the ideal the RIAA curves are.
Les

[oldeurope]

Originally Posted by Top Cap

Hi Darius, The original WAD amp used 1% in places, presumably they are part of the RIAA characteristic circuitry. In fact one resistor is 316k and was a value I could not find so I settled for two series connected 158k that were available from RS. Perhaps one day I will get the use of a swept audio spectrum analyser just to see how close to the ideal the RIAA curves are.
Les

Hm, build or buy an "inverse RIAA". The whole thing must give a linear frequency response.

Please have a look at R11/12 and R13/14. They get there signal from Test point 9/12 plate of V1b/V2b. Do you know the resistance there?
S is maybe 1,5mA/Volt, micro is about 100. 1/S = 600 Ohms
So the valve resistance is (1K2 + 600) Ohms x 100 = 180K
180K II 100K II 1Meg = 60K Ohms I guess the tolerance may be 20%.
This not exactly known value is in series with 2 x 158K Ohms 1%.
Conclusion the 1% makes no sense any more. And if the resirance is higher tolerance, why must the cap be 1%. You definately need some adjustment.

Kind regards,
Darius

[Top Cap]

Hi Darius, well when I decided to build a phono preamp for my Shure V15 Mk3 it was a choice between the Shure (M65 I think) or the WAD. The WAD had some very good reviews and so I went in that direction. Unfortunately the company closed down before I could order a kit so I did the home built clone or as they call it over here in the UK a 'Chinese Copy'. There was a write up about the circuit on their web site but alas it cannot be accessed now. I did some manual plots in 5db steps, taking a frequency measurement at each step from +20 down to -20dB using a constant 100mV input. The curve looked pretty good and virtually in step with the attached curve I found on a website, but then I was doing it all with very basic test gear.
Les

[oldeurope]

Hi Les, I sent you a PM.
Darius

[mickjjo]

Originally Posted by Top Cap

There was a write up about the circuit on their web site but alas it cannot be accessed now.

Is this the one Les?

http://web.archive.org/web/200512260...s/phono11.html

Most things can be brought back from the dead here . :-

http://web.archive.org/web/*/worldaudiodesign.com


Regards, Mick. .

[Top Cap]

Hi Mick, Yep that's it, many thanks for finding it. Well I am not so happy now with the results from tapping the cathode . The PC soundcard sounds fair enough but when I feed the valve amp the reduced output is very noticable. I have to set the main amp input control to maximum and the mid range seems to have far too much prominence now. A change in freq response was what I most feared.
I Guess I shall have to fit a selector switch on the phono amp or try another method like reducing the gain of an earlier stage without disturbing the RIAA or overall freq response.
Les

[oldeurope]

Originally Posted by Top Cap

Hi Mick, Yep that's it, many thanks for finding it. Well I am not so happy now with the results from tapping the cathode . The PC soundcard sounds fair enough but when I feed the valve amp the reduced output is very noticable. I have to set the main amp input control to maximum and the mid range seems to have far too much prominence now. A change in freq response was what I most feared.
I Guess I shall have to fit a selector switch on the phono amp or try another method like reducing the gain of an earlier stage without disturbing the RIAA or overall freq response.
Les

Good morning,
I am a bit confused with the RIAA roll over frequencies. I must read and google...
Yours should have 75us and 3180us.
Using the ECC82 gives 14dB loss.
ECC82: R19/20=3K3 R17/18=100K R21/22=289K (270K)

Darius

[Merlin]

I've been looking at the full schem with interest. I think it odd that they left the cathodes undecoupled- this retards noise performance. I suppose it does increase linearity, but is that more important than noise?
I suspect the change in tone you hear when plugging into the amp is due to the fact that you've increased the output impedance of the phono by tapping off the cathode resistance of the ECC82. As a result you'll probably just have to go back to the way it was and just add an output volume control.

You could easily afford to reduce R17/R18 to 100k. That should bring the gain down to a more sensible level of close to 1000. You'll need to increase R21/R22 to 226k to compensate for this.

Really I would redesign the whole thing, probably with lower gain triodes- you really only need a total gain of at most 1000. This phono has a gain of over 1700! and with the poorer noise performance it seems a waste of all that gain.