Linear L50 Valve amp - biasing issue

[cjhudlin]

Hello guys! I'm new to the forum here and I have been doing lots of reading from this site - please forgive me if this post is in the wrong area.

Anyway, I have an issue with one of my amplifiers. I have a pair of Linear L50 valve amplifiers. I have completely refurbished them internally and one of the amps sounds great, but I get a reasonable amout of hiss from them both.

Also, my faulty amplifier has a biasing issue (I think) I think this is because the voltages are out. (I have attatched the wiring diagram I have used to wire my amplifiers for better understanding) firstly, I'm mixed between if my output tubes are cathode, negative or fixed biased... so if someone could clear that up for me, that would be great!

Secondly, my amplifier is push-pull using two EL34's. Now, when I turn on the amplifier, the power transformer hums mechanically quite loudly and this hum will go through to the speakers, but only minimally. If I remove the output tubes, the transformer is silent. this lead me to remove the anode wires from the output valves (trial and error fault findng) and the left EL34 1k resistor overheated. I then removed the red and black wires from the junction box between the two EL34's and the transformer hum had disappeared. This lead me to believe that the output transformer was possibly shorting to earth somehow, so I removed the transformer and tested it by using the continuity setting on the DMM and tested the connections from wire to output trans chassis and found nothing.

Please bear in mind that I have not taken the current (mA) and I'm not a great electrician but I'm trying to make the best valid attempt. I'm trying to make these amplifiers into two seperate - clean mono amplifiers for Hi-Fi listening with the equipment I have!

Thanks for any suggestions.

[Edward Huggins]

I remember these amplifers very well. These were primarily used as (very low cost) PA and/or Guitar amps and you may be expecting just too much of it in a "Hi Fi" context.

The hiss and transformer (lamination resonance?) sound you hear may be relatively typical of the quality of the components used. In those days 50 watts was very powerful and such minor imperfections might have been tolerated as a kind of trade-off given the power desired.

[Station X]

Circuit here:-

http://www.chambonino.com/carlsbro/conc30a.html

Check that there is no POSITIVE voltage on pin 5 of the EL84's

It uses cathode bias with individual 470R cathode resistors for each valve. Check that they measure 470R (amp turned off) and then check the DC voltage across them. Use ohms law to calculate the combined anode and screen current.

I'm not surprised the 1k screen dropper resistors overheated when the anodes were disconnected. All the current intended for the anode would have been flowing through them! I hope you haven't damaged the valves.

[lesmw0sec]

From the circuit which Station X kindly added, I can confirm that the amp. uses cathode bias and not fixed. It is a great mistake to remove the anode connections whilst leaving the screen powered. This can lead to the screen grids drawing too much current and becoming red hot, as well as the overheating resistors which you discovered.
As X states - worth checking that there is no positive voltage on the control grids, which could be due to a leak of the coupling caps. If the voltages and resistor values are correct on both sides and hum persists, look at the smoothing.

Les.

[cjhudlin]

Thank you for the replies guys, really appreciate it.

Thanks for the advice about the anode thing!

Pin 5 to earth on the left o/p valve (according to the diagram I posted, with 200VAC and VDC setting on DNN) - 00.4VAC/00.1VDC and right o/p tube - 00.7VAC/00.0VDC.

I have also noticed that my anode voltage from this bad amp runs at 623 volts whereas my good amp only runs 610 volts, the bias voltages are pretty much the same. With the anode voltage being 13 volts higher. Caused by the AC voltage on pin 5?

Both of the resistors measure up at 470ohm. The left EL34 470 resistor has a DC voltage of 30V and the one on the right reads at 35V.

As for measuring the combined anode and screen current, can I do that from the voltages I've given you using ohms law or do I need to do something else?

Sorry for my incompetence! I'm new to the valve scene!

[Station X]

Your grid voltages on pins 5 look good, but there's no need to check for AC voltage here. Also I'd use a much lower voltage range say 10V or perhaps 2.5V on your meter so as to get a more accurate reading.

The combined anode and screen currents of each EL84 pass through its 470R cathode resistor. Now when current passes through a resistor, a voltage is developed across it. This is the 30V or 35V you're measuring.

Ohms law says that current is given by volts divided by ohms, so dividing 30 by 470 you'll get about 64mA. I'll let you work out the figure for the other valve. You can also work out the screen grid currents by measuring the voltage across the 1k dropper resistors (check their values first).

Try swooping over the valves to see whether they draw the same current in the other position. It's possible a cathode bypass capacitor is electrically leaking and shunting its associated cathode resistor.

The notes to the diagram linked to in post #3 suggest an HT voltage of 500VDC.

[Hartley118]

The Linear L50 and its 30-watt brother the L30 are good robust PA amplifiers. I have several examples in my collection including a fully restored one which I'm confident to lend out from time to time when neighbours' offspring want to produce a cheerful noise at parties from their iPod music. It's unburstable under abuse, unlike many transistor amplifiers!

Half a century ago I used the redoubtable RSC kit version, the A10, for many years for student entertainment in theatres, garden parties etc. It gave sterling service and those valve watts were plenty loud by the standards of its day. I remember building an A10 for the school in the 1950s which clearly covered an area of several football fields working into a horn speaker. It actually measured pretty well AFAIR giving full power down to 30Hz or so.

Back in the day, many were bought as guitar amplifiers though I suspect that the relatively small output transformer probably wasn't up to handling the low-frequency transients from a live guitar or bass pickup. On speech & music it was fine.

Its Achilles heel, which marks it out as a PA amplifier rather than a domestic hifi amplifier is, as you've pointed out, its quiescent background hiss level. This isn't a component quality problem, it's a gain structure issue. This is because the two input mixing level controls are placed right at the input of a high gain amplifier (2x EF86 pentodes + ECC83 phase splitter feeding 2x EL34s - or 2x 807s in the original RSC A10). So effectively you have a mic amp running at full gain irrespective of the volume control settings.

If you don't need the full gain of the mic input, you could reduce the background hiss level by around 10dB by changing the 10megohm feedback resistor (brown black blue bands) round the first EF86 stage to 3.3Mohm (orange orange green bands).

But before making any modifications, follow the sound advice from Station X and get both amplifiers working properly!

Martin

[cjhudlin]

My mistake, my DMM only goes down to 20vdc then 2000m and 200m, and thank you for the equation! this means that 35v / 470 = 0.0744 = 74mA.

Is the 64 and 74mA difference acceptable? or is there something I could do to even them out? On my other amp the bias voltages are consistently 32vdc each.

Anyway I re-measured in the 20vdc setting. The left EL34 reads 0.18, the right: 0.04.

if I swap the valves around left: 0.07, right: 0.13

The 1k screen resistors (which I'm assuming are the dropper resistors you're referring to) both measure at 1k and 998 (basically 1k), and the voltages are exactly the same for each valve.
voltage on the pins: 483vdc
voltage after the resistor 503vdc

I'm not entirely sure which voltage I should be using to measure the current so I read the two.

On my other amp, the voltages are 437 and 448 respectively. Is there any particular reason why these voltages are lower on my 'good' amp than my 'bad' amp?

Originally Posted by Hartley118

If you don't need the full gain of the mic input, you could reduce the background hiss level by around 10dB by changing the 10megohm feedback resistor (brown black blue bands) round the first EF86 stage to 3.3Mohm (orange orange green bands).

Thank you very much for the advice and will keep this in mind! I've always questioned how the high gain stage worked. I like having both of the knobs - just in case I want to use it for any reason, but the hiss is very annoying!
If I was to change this resistor, will it affect the volume from this knob also?

regards

[Edward Huggins]

As stated here these are not really designed for Hi Fi. I'm wondering what signal source/s you intend to use? What I am thinking is that if you are still determined to use these as Monoblocs, whether you could bypass the pre-amp/Baxendall tone controls and use these as just power amps driven at line-level into the first triode of the ECC83. This could be achieved if you were to use it for records with a RIAA pre-amp. Edward

[Station X]

When measuring the grid voltage on pins 5 you can afford to use the 200mV (0.2V) or 2000m (2V) range. Obviously you'll need to use a higher range to measure the cathode voltages.

I spoke of measuring the voltage ACROSS the screen dropper resistors. This means that you connect one lead of your meter to one end of the resistor and the other meter lead to the other end of the same resistor.

Doing it your way you have 503V on one end and 483V on the other. The voltage across the resistor is therefore 503-483=20V. Applying Ohms Law this gives a screen current of 20mA.

You haven't said what happens to the cathode voltage if you swop the valves over.

[cjhudlin]

Ahh I see, thanks for clearing that up. You're really helping out here, I appreciate that.
As for the cathode voltage:
Before swapping: left: 31 right: 36
After swap: Left: 35 Right: 32
Regards

[Hartley118]

Originally Posted by cjhudlin

i like having both of the knobs - just in case i want to use it for any reason - its there. but the hiss is very annoying!
if i was to change this resistor.. will it affect the volume from this know also?

If you change the resistor, the hiss level will drop substantially,but you'll also need to turn up the volume knob higher for a given input level. As long as you can then get enough volume from the high gain input, that'll be OK.

Martin

[cjhudlin]

Ah okay thanks Martin, I will keep this in mind and try out some resistors and tailor to my liking.

regards

[cjhudlin]

Originally Posted by Edward Huggins

As stated here these are not really designed for Hi Fi. I'm wondering what signal source/s you intend to use? What I am thinking is that if you are still determined to use these as Monoblocs, whether you could bypass the pre-amp/Baxendall tone controls and use these as just power amps driven at line-level into the first triode of the ECC83. This could be achieved if you were to use it for records with a RIAA pre-amp.

I know they arent, but if I directly wire the input into the first valve (these are wired up as pentodes in this amp [as far as I'm aware]) then I would just have to buy another pre-amp... Why would I do that when it has one built in?

[Station X]

Originally Posted by cjhudlin

As for the cathode voltage:
Before swapping: left: 31 right: 36
After swap: Left: 35 Right: 32
Regards

The fact that one valve has a higher cathode voltage, regardless of its position in the circuit, shows that the valves are not matched, but it probably hardly matters in this amplifier.

You could prove the point by swopping valves between the amps and seeing what bias voltages you get.

Going back to your original post I don't think there's anything wrong with your biasing arrangements.

As for the hum, the smoothing capacitors ie the 16uF and 32uF electrolytics are the prime suspects here. You say you have "refurbished the amps internally". Were the smoothers changed? Are any replacements fitted the right way round?

I will not get involved in discussions about modifying the amplifiers. Other forum members know much more about this than I do.

[cjhudlin]

Thanks, I'll try that.

Yes, all of them were done. All of my capacitors are 32uf apart from the coupling caps which are 0.1 and 47.

I have questioned myself as to if they are in the right direction but they all check out

[Edward Huggins]

Originally Posted by cjhudlin

I know they arent, but if I directly wire the input into the first valve (these are wired up as pentodes in this amp [as far as I'm aware]) then I would just have to buy another pre-amp... Why would I do that when it has one built in?

The first valve is a Pentode. I know there are sufficient amplifying stages in the existing L50 to give you enough gain - BUT - what I am suggesting here is that the noise/hiss that is annoying you may be being generated in the original pre-amp stage/s ahead of the ECC83 - and a modern MM phono stage will be much quieter.

[Hartley118]

Originally Posted by cjhudlin

I know they arent, but if I directly wire the input into the first valve (these are wired up as pentodes in this amp [as far as I'm aware]) then I would just have to buy another pre-amp... Why would I do that when it has one built in?

I know what you mean - a couple of EF86s should be a capable pre-amp, but these Linear amps were designed back in the day for crystal/ceramic pickups (low gain input) and crystal microphones (high gain input). Neither of the inputs has the RIAA equalisation that a magnetic pickup would need.

The nearest approximation would be to feed the magnetic pickup into the high gain input and turn bass boost to maximum and treble to minimum. That might be listenable, but not very precise!

If you're feeling adventurous, you could convert those two EF86 stages into an equalised pre-amp with input switching like the attached example. However, that would present you with two further challenges:

1. Obviously you'd end up with two independent preamps built into the amps without ganged stereo switches or tone/volume controls.

2. The negative feedback is currently connected from the output via those two 100k resistors to the cathode of the second EF86. That would have to move or it would work to cancel out the effect of the tone controls in the new preamp design. But where should it move to? With this sort of mod, you rather quickly get into the challenges of redesign!

Personally, I'd look for an external stereo preamp to feed the monoblocs.

Martin

[cjhudlin]

Originally Posted by Edward Huggins

The first valve is a Pentode. I know there are sufficient amplifying stages in the existing L50 to give you enough gain - BUT - what I am suggesting here is that the noise/hiss that is annoying you may be being generated in the original pre-amp stage/s ahead of the ECC83 - and a modern MM phono stage will be much quieter.

I think you are quite right. If I remove the ECC83 I get zero hiss. The treble knob obviously amplifies the hiss also.

Originally Posted by Hartley118

If you're feeling adventurous, you could convert those two EF86 stages into an equalised pre-amp with input switching like the attached example. However, that would present you with two further challenges:

1. Obviously you'd end up with two independent preamps built into the amps without ganged stereo switches or tone/volume controls.

2. The negative feedback is currently connected from the output via those two 100k resistors to the cathode of the second EF86. That would have to move or it would work to cancel out the effect of the tone controls in the new preamp design. But where should it move to? With this sort of mod, you rather quickly get into the challenges of redesign!

Yes it's seems a bit of a job I would like to keep the tone controls and eq if I could...

Would it be possible to have one EF86 pre amplify the signal before the volume knobs and then use the volume knobs to control volume after it, and use the other valve to control bass and treble?

Or use both EF86 to pre amplify the signal then use the volume knob treble and bass after?

An example would be the Quad 22 pre-amp. The signal goes straight to a valve and is then attenuated by the bass treble and hop knobs. Would something along these lines work? Or would it not make any difference?

Thanks for your help guys

[Hartley118]

Most of the hiss originates in the first EF86 stage because that has the most gain following it. It's likely that the thermal noise in the mixing network resistors is the dominant factor.

Martin