New Project Required

[barnesdavid20]

Hi John
Understand what you are saying, and yes there are slight differences in the layout, they are not a mirror image, both inputs are at one side as are the loudspeaker terminals at the other side, I did take the mains switch from the front panel and put it next to the mains input socket to avoid running mains down the side of the the LH channel. Have done an awful lot of mods which has involved moving things around, bigger caps to fit in etc, but at no time did any thing have any effect on the hum level on the right, ironically that is the side with the shorter input leads as the input sockets are on that side, and of course the star earth for both channels is at this point to, so leads running from the left are that much longer so you would think that the left would have slightly more hum, crazy!!
I think I will just give up and enjoy the music,and accept that it is well within the spec intended for the amp,
I have got an oscilloscope and signal gen, how do I connect it up to measure the Hum, have not yet had the nerve to try it out on the amp yet.
Thanks for your help John
David.

[John Caswell]

Hi David,
Measuring residual hum and noise with a scope is fairly simple and will give reasonably accurate results but not as good as a dedicated audio analyser.
All you need to do is fit a dummy load of say, 8R, (2Watts is adequate) across both the LS terminals of the amp and then connect the scope across the dummy load(s) Then fit s/c phono plugs to the amp input
Turn the amp on and adjust the scope to maximum gain on the "Y" channels, trigger the scope from line (ie 50Hz)
There will be a trace of some sort displayed which is the residual hum and noise of the amp.
Because the scope displays the full signal, the measurement is peak to peak (p-p)and you can measure this using the graticule on the scope screen (I hope your scope has one)
To convert this p-p voltage to rms, which is what an a analyser displays on its meter, you divide it by 2√2 = 2.828.
So if the p-p is 5mV then the rms is 5/2.828 =1.76mV
1mV is -60dB wrt 1.00V so this will probably be audible in some loudspeakers. Try it and see.

John

[Richard]

Another way is with your multimeter set to AC mV and a probe on each speaker terminal with the amp on and a CD player connected and selected and switched on (but not playing!).This is approx as it may be 50Hz or 100Hz hum or a variety of noise frequencies that you can hear and the meter will just display them combined. The scope is much better to lock onto the wave and check its frequency etc. As a guide 1mV is pretty quiet for a valve amp though you will still hear it close to the cone. The other thing to bear in mind is one speaker may sound a little noisier if it's in a corner or closer to the wall(s) than the other one.

[barnesdavid20]

Hi John/Richard
Thank you so much for your advice/instructions, sounds simple so will give it a go today both with the scope and with the multi-meter.Will report back with results.
David.

[John Caswell]

Hi all,
If trying to trace hum on an amp you really don't want anything else connected to mask the results as you may have an earth loop, so stick to the bare bones.
Although a DMM is very useful it really doesn't have the resolution of a 'scope and also you can't see what the predominant hum waveform is.

John

[barnesdavid20]

Hi John/ Richard
Something really strange going on here, I am totally baffled.
I measured the milli- volts across the speaker terminals as Richard said using an RS calibrated Fluke 179 and the results were as follows LH speaker 0.9 mv
RH speaker 2.3mv so i thought great, at least that shows why i can hear more hum from the RH speaker, however I swapped the 7025s around and low and behold the 0.9mv moved to the RH speaker and the 2.3mv moved to the LH speaker, thought I had cracked it until I found that the hum was still coming from the RH speaker!!!, This suggests to me that the millivolts across the speaker terminals are not related to hum levels. What is going on?
Thought I would tell you this before I do anything with the scope.
Sounds as if I need to try some different 7025s anyway, they are obviously not balanced very well, although can't hear the difference.
Still needing help
David

[John Caswell]

Hi all,
Hence my comments re using a scope to see what is going on! Have a look to see if you can deduces what is happening.

John

[barnesdavid20]

Hi John
Well I have had my first attempt to use my oscilloscope, I have done what you said with the following results. The LH side had a sine wave with higher frequency's imposed on the top of it, peak to peak was about two divisions on the graticule with the scope set to 5mv/div, which i worked out to be about 3.5mv RMS. One cycle covered 4div at 5ms/div which I worked out to be 50cycles. but this of course is the side that is virtually silent, and measured 2.3mv 0n the Fluke. The RH side showed a flatter response again with high frequency imposed on it but showing the top half of the 50 cycle sine wave, but overall a lower RMS value, which was in keeping with the fluke result.
I then swapped the 7025s over and things changed, the left hand side now had the flatter response and the RH side had the larger RMS value but not such a nice sine wave as it was on the LH side.
Although not very clear what is happening, it seems that I have a 50 cycle signal there which I suppose is what I am hearing, but why it only appears loader on the RH side I can't imagine.
Where could the 50cycle signal be coming from, is it normal to see any signal at 50 cycles on the output of a valve amp?
Cheers David

[barnesdavid20]

A further thought, when I changed the 7025s over and got 0.9mv on the multimeter on the LH side where the slight hum is, that this is quite normal for a valve amp, changing the 7025s could possibly give me 0.9mv on each channel. The 2.5mv that I am getting at the moment from one of the valves suggests that only 0.9mv of the 2.5 is of an audible frequency otherwise the hum would go away when I changed them around. Am I thinking correctly.
Have you any advise on what valves to buy to replace the original sovtek 7025s that I am currently using.

David

[bikerhifinut]

A little light just switched on and I wondered?

In another post you say the 5881 amp has 75mV sensitivity for full output?
That's hellish sensitive. It won't take much of a hum/noise component to get noise on one or both channels. So maybe no amount of tweaking is going to reduce what is, from what you have said, already an acceptably low amount.

You'd likely get silence if the amp could be configured for less overall gain, say down to half a volt. as per the WD88VA. That's not the same as bunging an attenuator at the input. The amp will still have 75mV sensitivity after the attenuator.

This isn't a trivial bit of design, if you reduce the open loop gain, you also have to alter the feedback network to suit in order to keep the overall feedback levels the same.

I hope I didn't get hold of the wrong end of the stick here.

A