5F6-a tuning...

[Jester]

I'm still trying to work out how best to set up my 5F6-a clone. I'm a little confused and wonder if someone could clarify a few points for me.

1: When setting up the bias, should I really be just setting the current to 25ma for cold, 35ma for medium or 45mA for hot. I have read these figures a number of times for 6L6GC tubes, and I guess they are for a fixed plate voltage, is that right? Would I be right in saying what I'm really after is a range of plate dissipation powers? Say around 30 Watts for a 6L6GC (what range do you guys work with?). If that's the case, then I guess if I change my rectifier for a solid state device, I will increase my voltage, and to keep the same plate power, have to reduce the current. Is that right?

2: When I bias my amp, I put a 4.1 Ohm 25 watt resistor in between the cathode and the ground of both the power tubes (so I get the sum of both currents through this resistor - I have an order in for 2*1 ohm resistors, but in the mean time I use the lowest resistor I have). I then measure the voltage across this and tweak the bias pot to get the right current. So please correct me if you see me doing something wrong, but I think when I get 0.282 volts across my 4.1 ohm resistor, I have 34.4mA through each tube. (ie 0.282/4.1 = 68.8mA from both tubes, ie 34.4 mA each.) Since at this bias level I have 480Volts on the plates of my tubes, I think this bias is set up for 480*0.0344 = 16.5 watts.

FYI to attain this bias current I have -48.9vdc at the bias point shown on the original schematic.

I guess you will now see why I'm so confused, and thinking I've got something above wrong (okay I realise it might be more than one thing - so please help out a thickie, who's trying to understand it). Or maybe I just haven't biased the amp enough, and this low power setting is why I get early break-up from the bass notes, when played through a closed back cabinet.

3: I'm of the impression that higher plate voltage, when biased correctly, will give greater headroom (ie. more clean sound). Is that right?

4: I believe increasing the main filter caps. helps firm up the bass response of the amp. What effect do you hear if you increase the pre-amp filter capacitor?

Thanks for your help here guys.

A little confused Jester.

[MWJB]

1) assume 25W max dissipation for most current production 6L6GC, JJ is 30W. If you bias to 70% plate dissipation (17.5W) with the SS rectifier you should have a happy enough compromise when switching back and forth with a GZ34. Typically 5F6A are biased to 15-17W (30-35mA).

2) Yes, sounds good. Neg voltage is just a means to an an end - set by plate current. You do not have the amp set at "low power" - the amp's power will not differ tangibly if you add or subtract 10mA, the tone will however. We are talking here about the tubes' DC plate dissipation W, not the amp's AC output W.

Do you think that this early bass break up is part of the 5F6A's character (seems unlikely) or is it a malfunction? If it's a malfunction you need to fix it rather than start changing things (bias & filtering would seem to be covered pretty well). What's your layout like? What do other amps sound like in the same cab?

3. Correct - experimenting with different brands of tube may see you yeild +/-10 or 15vdc at the plates. Obviously some tubes characteristics better lend themselves to tight & clean (e.g. NOS 7581A), this is probably the more important area to look at, you are limited in ways that you could push up the plate votage (you'll probably get max plate voltage with SS recto, Sovtek 5881WXt biased on the cold side, 20-25mA, but you may dislike the tone - so it makes sense to use a tube with characteristics you like and regard plate voltage specifically as a secondary issue).

4. Don't go over 20uf at the preamp, try 40-50uf at the screen supply instead.

[Jester]

Thank you MWJB for your helpful response.

1) assume 25W max dissipation for most current production 6L6GC, JJ is 30W. If you bias to 70% plate dissipation (17.5W) with the SS rectifier you should have a happy enough compromise when switching back and forth with a GZ34. Typically 5F6A are biased to 15-17W (30-35mA).

I understand - Thank you. Can I ask where the 15-17 watt number came from? I only have one other reference for setting plate dissipation, and it's for a Champ style amp, using a 6V6 and their the plate dissipation is recommended at 12-14watts, and from the data sheets I had this is the maximum plate dissipation (in fact the upper range appears to be two watts above). I assumed this maximum dissipation was relevant, and thought it might translate to other valves used for guitar amps too. Clearly I'm wrong, so if you happen to be able to steer me in the right direction, I'd be thankful.

2) Yes, sounds good. Neg voltage is just a means to an an end - set by plate current. You do not have the amp set at "low power" - the amp's power will not differ tangibly if you add or subtract 10mA, the tone will however. We are talking here about the tubes' DC plate dissipation W, not the amp's AC output W.

Thanks for the confirmation.

Do you think that this early bass break up is part of the 5F6A's character (seems unlikely) or is it a malfunction? If it's a malfunction you need to fix it rather than start changing things (bias & filtering would seem to be covered pretty well). What's your layout like? What do other amps sound like in the same cab?

I don't really have a good base line on which to judge this. I have a book called the sole of tone, celebrating 60 years of fender, and it includes a cd with sound recordings of a 59 bassman played with the tone and volumes set at 5 (I realise this number could confuse due to different log scales on some audio pots). It's bass notes are clean at this volume. I realise I might get a different sound playing through a hi-fi, but it's all I have to compare to. I must confess what I hear on the CD is what I think I heard when I played this type of amp in the guitar shop months ago and is what I'm aiming for. This is why I think I might have a problem. Finger picking / strumming is fine, with clean sound, it's just plectrum strumming when it starts to break up, and I don't like that sort of sound - if the amp were close to full volume, then I would accept it as normal.

As for layout, I have tried to keep control grid wires as short as possible, and have wired the input resistors directly from the jacks to the tube sockets. In fact most items are connected directly to the tube sockets, and I have not used a cct board for the main components, just for the various electrolytic power filtering caps. I have the signal flow from left to right, and not going back on it's self. I have the power supply, choke and o/p transformer all in different planes and at least 4cm apart. The three pre-amp tube etc are at least 10cm from the nearest transformer. What sort of layout do you think would increase bass breakup?

I have a selection of speaker cabs to play through, and the Celestion and Tone Tubby 12" alnico speakers are both closed cab fold back cabs (I do intend to build some proper cabs soon). These show break up first. The other speaker is a celestion 12" vintage 30 in an opened back Marshall combo. This open back configuration gives the cleanest sound with regard to break up, but it is still noticeable. The Champ clones that I have built use these cabs and don't appear to exhibit the same problem, even thought they have a good clear and solid bass response.

3. Correct - experimenting with different brands of tube may see you yeild +/-10 or 15vdc at the plates. Obviously some tubes characteristics better lend themselves to tight & clean (e.g. NOS 7581A), this is probably the more important area to look at, you are limited in ways that you could push up the plate votage (you'll probably get max plate voltage with SS recto, Sovtek 5881WXt biased on the cold side, 20-25mA, but you may dislike the tone - so it makes sense to use a tube with characteristics you like and regard plate voltage specifically as a secondary issue).

This is a good lead, and maybe something I need to try out. Thanks

4. Don't go over 20uf at the preamp, try 40-50uf at the screen supply instead.

Okay, thanks.

PS: So, after reading all your suggestions, I gained the confidence to try a few things. I had been worried about using an ss rectifier (because of my flashes from using the standby switch, and having taken to power up with the switch made), but thought I should give it a go. So I put one in and measured a few voltages, and also played around with the bias, using my ears to decide what I liked rather than the multimeter (yes, I checked I was in sensible ranges all the time). Eventually I decided to settle on a bias setting of 36.25ma at 500V on the plate (yes it might look as though I'm being a bit accurate here, but in fact it's a fairly course setting on my pot, so this is me aiming at 35ma. I really don't think I can hear the difference between a couple of ma, as i am a beginner). Anyway, I am much happier with this set up, and the sound is now very nice - I'm not noticing the early breakup, and I will play it like this for a while and get use to it, before changing o/p valves.

Thanks for the advice, it's very useful and helped a lot

Jester

[MWJB]

"I understand - Thank you. Can I ask where the 15-17 watt number came from?" It simply came from multiplying plate curent by plate voltage...I don't set out to achieve 15-17W, it's just that when the plate current is set to where it sounds right, to me, in milliamps, that this is typically what you get. Set bias/plate current in milliamps, THEN look at plate dissipation to make sure you're not going to burn anything up.

A Champ is SE, Class A and cathode biased - many, many SF Champs are in fact dissipating over 20W! It's not a good idea to bias a push-pull, class AB, fixed bias amp like your 5F6A at such high plate dissipations. Unlike that champ, your plate current will rise appreciably when hit with a signal, you want to allow for this.

"What sort of layout do you think would increase bass breakup?" From what you say, you appear to have given the layout a good bit of thought, I'd have to see it to make any further comments.

[Jester]

MWJB,

Thanks for your reply. It's good to know that the values for plate dissipation come from the ear, rather than the data sheet (which is what I was trying to get a feel for).

It's also good to hear that many champ clones run at 20 watts. I was worried when one of mine did, even though it sounded great. I change the power supply to drop the voltage, and got another fine sounding champ clone.

Thanks for helping a newbie, and answering my question.

cheers
Jester

[Enzo]

I am curious, since you have fixed bias, why did you select 4.1 and 2.1 ohm resistors for your bias current sensing? If 4.1 is the only low resistance part in your drawer I can see using it, but you plan to ORDER a 2.1 ohm. WHy not use a 1 ohm 1% resistor like everyone else does? That way the milliVOLTS across it are exactly equal to the milliAMPS through it. IN other words 34ma through the tube reads 34mv on the meter.

ANd while you need a high power resistor when using it for cathode bias - 250 ohms or whatever - consider that the 1 ohm resistor with its 34mv across it, or even the 68mv for two tubes, will only be dissipating less than 5 MILLIWATTS. Even your 4.1 ohm resistor only dissipates less than 20 milliwatts. SO there is no need to order 25 watt resistors for this use. Yes, if the tube fails there would be a lot more current, but then even a 25 watter could burn up.

I usually use 2w 1 ohm resistors, not for theur power handling, but that size is large enough to be mechanically sturdy and easy to get my meter probes on withiout being large enough to get in the way.

Just a thought.


Filter caps. The first filter cap is for the power tube plate and the second for the screens, at least usually. They serve the power amp and can affect its response. By the time the B+ gets to the screen cap, it is pretty well filtered. The later filter caps, the ones for the preamp stages, are in my view more about decoupling than filtering per se. The preamp is all class A circuitry so the current draw is pretty steady. Once the caps are large enough to sufficiently decouple the stages, increasing them doesn't do a lot.

[Jester]

In answer to your questions Enzo:
I tried to say I've ordered 2 times 1 ohm resistors, not a 2.1 ohm resistor. I assume I put one in each leg, and can see if they are balanced (though I don't know what I'd do if they are not - as I only have one bias cct.) I did not want to assume that just putting one on one of the tubes was okay and leaving the other. I realise I could join both of the cathodes together, then ground them through one resistor, but then I'd get twice the bias current through it right? So I've tried to keep it simple. Do you just do one?

I did not select the 25w 4.1 ohm resistor from a great selection. It was either this power or some other resistors with 100s ohms at 1/2 watt. So it was just the best available while I wait... I don't leave it in cct, so I just hook it up when I bias, then re-ground the cathodes when I'm done. I'll leave the 1 ohm resistors in there -this one's just a bit big, and as you say massive over kill. Does this resistor effect tone - ie do people use Carbon Composite, or just the cheaper carbon film in this position?

Thanks for explaining what you use.

Thanks also for your comment on filter caps.

A learning Jester.