Mullard 5-10 amplifier

[vampyretim]

Originally Posted by GrimJosef

You're welcome. While I was writing Mike (Boulevardier) at post #96 requested a quick check just to make sure the hiss really is coming from the amp and not from the pre-amp. If you're not certain that it's from the amp this check could save you a lot of time and fuss.

Cheers,

GJ

Ahh, I forgot to mention in my last post that I've disconnected the preamp and it's definitely coming from the power amp.

[PJL]

Shorting the input is still a worthwhile test. The EF86 anode looks reasonably clean with some 50Hz pickup but there is also some HF noise evident in the width of the trace.

[GroovyG]

Originally Posted by vidjoman

I have a feeling that the resistors Ra & Rb after the pot are to be set according to which power amp you have connected - 5-10 or 5-20. Think it was something to do with getting the correct input sensitivity. Long time (about 55 or more years) since I read the Mullard book.

I have the assembly data booklets for the Mullard 5-10 and the 2 valve control unit from Sterns Radio (both units built from kits by my Dad in the early '60s) and can confirm that RA and RB after the volume pot (pic in post #91) should be, for the 5-10, 680k and 150k, and for the 3-3, 270k and 180k respectively. (When used with the 20W amp, they should be omitted altogether, and R24, the HT dropper, changed from 22k to 56k.)

[vampyretim]

Hi all,

With the preamp connected or disconnected from the BNC input the 5-10 hisses but with the preamp disconnected and the 5-10 BNC input shorted to deck the hiss practically disappears. I have tried replacing the 2.2M resistor but it made no difference. I'm a bit confused, does this make sense to anyone?

[Hartley118]

Originally Posted by vampyretim

Hi all,

With the preamp connected or disconnected from the BNC input the 5-10 hisses but with the preamp disconnected and the 5-10 BNC input shorted to deck the hiss practically disappears. I have tried replacing the 2.2M resistor but it made no difference. I'm a bit confused, does this make sense to anyone?

This finding suggests that the power amp is blameless and you're just hearing the thermal noise of the 2 megohm input resistor.

Any resistor generates a thermal noise voltage equal to 4kTRB, where:

k is Boltzmann's constant
T is absolute temperature (degK)
R is resistance (ohms)
B is bandwidth.

Being a trifle lazy, I used the thermal noise calculator on this website
https://www.daycounter.com/Calculato...lculator.phtml
which indicates you'll be getting around 25 uV from 2 megohms. That's probably about 85 dB below the max output of your preamp, so will be audible.

So....sorry, you're hearing thermodynamics at work.

Once the low output impedance of the pre-amp is connected, that will shunt the 2 meg resistor and reduce the power amp's thermal noise a lot. However, your noise level will now be dominated by the pre-amp circuitry, which is probably noisier than the power amp.

Incidentally, one of the reasons that solid state amplifiers tend to exhibit lower noise levels than valve amplifiers is that the resistor values are generally lower.

Martin

[vampyretim]

Thanks Martin,

I guess that I'd better start looking at the preamp then.

[Boulevardier]

Originally Posted by Hartley118

This finding suggests that the power amp is blameless and you're just hearing the thermal noise of the 2 megohm input resistor.



Martin

Which 2 Mohm resistor do you mean Martin - I can't see one in either the 2-valve preamp or the 5-10 without controls.

Mike

[Hartley118]

I'm afraid I'm working without a circuit diagram and jumping to a conclusion from:

"I have tried replacing the 2.2M resistor but it made no difference."

If the input resistor is actually less than this then the noise should be reduced pro-rata.

The evidence that with "the 5-10 BNC input shorted to deck the hiss practically disappears" suggests that it's thermal noise in the input resistor.

Martin

[Boulevardier]

The preamp output is through a 250k log volume control with the slider connected to top of a 680k +180k potential divider, bottom to chassis and junction of resistors is preamp output.

I don't know what the OP is referring to by "2.2 M" resistor.

Mike

[vampyretim]

Originally Posted by Boulevardier

The preamp output is through a 250k log volume control with the slider connected to top of a 680k +180k potential divider, bottom to chassis and junction of resistors is preamp output.

I don't know what the OP is referring to by "2.2 M" resistor.

Mike

There's a 2.2M grid resistor for the EF86 it's attached to the control grid and the other side goes to ground.

[Hartley118]

That's how I read it.
M

[Boulevardier]

Assuming that you're talking about the input of the 5-10,and that it goes direct to the control grid of the EF86, then Mullard specify the value of that grid-leak resistor as 1 Mohm (see the text underneath the 5-10 circuit diagram in the Mullard book - page 40). However, I doubt if the wrong value would cause hiss - but it's worth trying replacing it with the right value.

Mike

[Hartley118]

In theory, changing the 2.2 meg to 1 meg should drop the hiss by 6dB with an open circuit input, but it would make negligible difference when the preamp is connected because then the preamp output impedance dominates what the input is seeing.

Martin

[vampyretim]

Hi all,

It seems that the schematic that I have differs from the ones that you're looking at, the one that I've been using was the one that came with my 5-10 and presumably was the one that mine was built from.

In the schematic that I have the 2.2M resistor is R1.

[Hartley118]

That confirms that the hiss is likely to be mainly the thermal noise from that 2.2 meg resistor. Worth noting that the EF86 stage itself is pretty quiet compared with a 2.2 meg resistor. That’s just physics.

As I’ve indicated, the power amp noise level with the preamp connected will be a lot lower than with the input open.

Martin

[Boulevardier]

Yes, that schematic is different from the original Mullard one. No wonder we've been getting confused. I haven't really been involved with valves for 40/50 years, so I'll leave it to those with more knowledge to say whether that resistor could be the cause of the hiss - which would of course vanish when you grounded the input at the socket as this would bypass the resistor to ground. From your photo that substituted 2.2 Mohm doesn't look exactly new, so could it also be faulty? I can't think why Stern-Clyne changed its value - unless they had a stash of that value they wanted to get rid of. My next move would be to change it for a new/newer 1 Mohm, and see if there's any improvement. They are fine amplifiers when working properly, considered in my youth to be in a similar class to the Quads, etc of the day (IMHO).

Mike

[EDIT - my post just crossed with Hartley118's, and he's saying much the same.]

[vampyretim]

Thanks for all your help guys, I only had a 3W 2.2M replacement and it was a bit of a tight fit hence why I put back the grubby looking original. The original did measure 2.2M so I didn't think any harm with putting it back. I've got some half watt metal film 1M resistors which I will put in its place tomorrow.

I'm getting all sorts of trouble with the preamp, it's humming and I'm getting 50hz at the output, I wanted to get the power amp up and running before I tackled it.

Thanks again, I've learnt a lot from this amp and even though the sound of it may well be coloured I love how it sounds.

[Boulevardier]

Here's a copy of the original Mullard circuit. In the caption to this schematic, it says: "For version without controls, the signal is taken directly to the control grid of V1, and R32 is replaced by a fixed resistance of 1 M ohm."

Mike

[Herald1360]

Given the output resistance of the preamp is around 150k (680k // 180k), you could happily drop R1 to 470k. That should quieten the hiss a bit further without any other effect on performance.

[vidjoman]

I notice that the feedback components are altered according to the speaker impedance shown in the above diagram.
That could alter the gain of the amp and therefor the noise/hiss level.