Peavey 6505/5150 biasing problem?

[markyb]

Hi,

I am fairly new to tube amp maintenaince but i'm quite good with electronics generally and understand most of the principles, so having previously modded and rebiased my EL84 equipped Carvin Nomad with absolutely no trouble, I have had a go at my (brand new) Peavey 6505 head, firstly to install an adjustable bias pot (as the 6505, and 5150 before it, have a fixed bias), and secondly to then bias it a bit hotter, as I've read that these amps are shipped with a very cold bias.

I replaced the standard 15K resistor with a 10K multiturn cermet pot and a 6.8K resistor in series, uneventfully. (The procedure for this mod is described in several online articles on 5150 bias mods). After getting some really weird readings when trying to bias the amp I replaced the stock (very poorly matched) Ruby tubes (which I understand are fairly cheapo) with a matched quad of JJ Tesla 6L6GCs, and now I'm at least getting consistent readings.

I am measuring the bias current using a Fluke 77 Series II meter, using the output transformer shunt method. The internal resistance of the meter is about 10 ohms.

OK, now the problem. I'm measuring a current of about 31 mA, which is a believable bias current, BUT this is meant to be for TWO tubes (cos it's a two-tube-per-side push pull amp), so it's really 15mA which is quite cold.

I have read that some amps (Marshalls - which much of the 5150 is based on) have a 'low' resistance in the output transformer primary, and that in these amps one should NOT believe a bias current measured with a meter with a 'high' internal resistance. In this context what is a 'low' transformer primary resistance, and what is a 'high' meter resistance?

I feel in my bones that 10 ohms (meter) is a significant proportion of 28-37 ohms (output tran primary) and so I'm reluctant to believe that enough of the current is being 'shunted' through my meter.

I have used another method of measuring bias current - measuring the output transformer primary resistances (with amp off) and then measuring the voltage drop across the respective windings with the amp on and warmed up fully. Using this method I get:

(red wire to brown) 1.105VDC divided by 29.5 ohms = 37 mA, and
(red wire to blue) 1.384 VDC divided by 37.5 ohms - 37 mA

divided by two (for two tubes per side) = 18.5 mA

If this is right, using the V=IR method, then there is definitely scope to increase the bias a little bit. I was planning on running the tubes a little colder than the usual 30-34 mA recommended, as the tubes will be getting some heavy work in use.

QUESTIONS:

1) Can I believe the (very low sounding) bias current as measured with the output transformer shunt method, bearing in mind quite similar resistance in primary windings and meter?

2) Can I believe the 'calculated' bias current (V=IR method)?

3) the plate voltage is supposed to be about 500V on the 5150/6505. But should I actually measure this as well?

Cheers,

Marcus

[Amp Kat]

You should always measure the plate voltage when doing bias . The actual voltage at quiescent multiplied by the current is the power dissipation product of the tube. The 6505 will not benefit from a hotter bias as the amp will get extremely musshy on you which is why it's biased where it is which is pretty cold and may have a slight bit of crossover distortion however this is over shadowed by the massive amout of gain and noise or hiss in the preamp section. You'll find out in the end that the bias is where it should be for the best sound of this perticular amp. -55 volts of negative grid bias is what it's supposed to be and that's what I'd set it to

[Enzo]

The tone of this amp really is in the preamp. But the mod you did should work. The series resistance from the rectifier of the bias supply is about 3k, then your pot and resistor finish the divider to ground. Turn the pot to zero ohms and I cannot imagine the bias votage won't be lower.

Those Ruby 6L6GCSTR tubes are really pretty good tubes, and I think they work well in a 5150.

You got 31ma one way and 37ma another way. Those are not all that far off from each other. Never hurts to have a bias probe around or make one for when the other methods leave questions. Those sound close for stock.

You most definitely have to measure plate voltage, otherwise you have no idea what power levels those milliamps represent. I find the 5150 generally runs at about 465 volts. The more current you have the power tubes drawing, the lower the plate voltage will sink. SO the two are a bit interactive.

[markyb]

bias problem update

I decided in the end to bite the bullet and make some adjustments. I measured plate voltage and set the bias current to give a power well within the maximum plate dissipation wattage. Bias current now about 30mA, plate volts 470V, giving a plate dissipation of about 14W, less than half what is quoted for JJ 6L6s, and less than 70% of the 'standard' dissipation for 6L6s.

I think the amp sounds better now. But I can always go back if needed.

The main difference is when using clean, 'crunch' or lower gain sounds, I noticed that before the bias was set, notes did not decay naturally, rather they cut off early with an irregular decay (as if a noise gate was on), but now they do not. Perhaps that 'cut off point' was the point at which the signal was mainly within the crossover distortion region. Certainly the clean channel is better sounding (while still by no means a Fender!)

I agree with the poster that on the 'full tilt' high gain settings which this amp is famous for, the power amp has very little effect on the amps tone.

Marcus

[Amp Kat]

This amp was actually suppossed to be a spin off of the Soldano an SLO 100 but their were some suttle differences and one major one and that was the amount of hiss that is present in the preamp at high gain settings which in this case is anything over 3. It is really meant to be an amp you can use at low SPL levels with tons of gain and with the master volume it really is but can use a noise gate in the loop for less noise when not playing. If you crank it and get power tube distortion which IMO is better with 5881's than 6L6GC's then you'd better be cranking that pregain down or all you'll get is a very high end sqeal. The clean is quite bright and yes very far from a fender but if you look at the schematic that Steve A redrew, as the Peavey one is a nightmare (thanks Steve !) you can see all of the frequency shaping networks that contribute to it's origin and removing them will only make it worst. One thing that is a strong point of this amp is the power section has balls for days and if you take a preamp and send it to the return it is very good and loud as hell but using the send and return can be troublesome as the send level is at -10dB and has no adjustment so some units may like it and some won't. The OT is also somewhat uneven which is why you got different readings from side to side. To compensate for this you could install two bias pots one for each side but I really didn't find a hum advantage to this but it did even out the sides. Overall the 6505 has it's place in mainly a metal amp but to be quite honest the XXX and JSX are much better amps of course depending what type of music you play.

[Steve Conner]

That "Noise gate" thing is indeed what you'd expect from an output section biased way too cold with masses of crossover distortion. It amplifies small signals much less than big ones. I've seen back-to-back diodes in series with the signal path in other Peavey circuits, to achieve the same effect (no signal below 0.7V can get through.) I'd tend to agree that it's a necessary part of the high gain sounds that people expect from those Peavey amps.

I experimented with the diode thing myself and found that it "tightened up" high gain sounds and reduced hiss and unwanted feedback when playing loud, but at the expense of making the thing sound just a bit "wrong" sometimes. Trying to clean it up with the guitar volume just made it fizz and sputter like it had a loose contact somewhere.