Not sure if my 6V6 or OT are bad...

[Raezzordaze]

So I just finished redoing a build that was basically a clone of the AA764 Vibro Champ but without the vibrato, so one 12AX7, one 6V6 and a 5Y3 rectifier. This is more or less the schematic. I modified the schematic I had since it was from a book of easy DIY home builds. I built it using point to point wiring and a homemade chassis and it turned out well. However, no matter what I did I couldn't get the amp to drive the 6v6 to full power and start breaking up. Even with a humbucker-equipped guitar.

I originally used a Classic Tone 40-18019 with a 325/0/325 HV secondary which I felt was high. Also, the guy messed up his the inputs so both hi and lo had some at least 6db attenuation. So I went to a single input with a 1meg leak and a 34k grid stop like this. I also swapped to a different PT that has a 300/0/300 HV secondary, the Hammond 272 BX. I also wanted a cleaner layout so I swapped to a turret board and used individual power supply filter caps instead of a can.

After I completed the build initial testing produced a somehwat reduced volume and very slight but noticeable high frequency squeal/oscillation that would disappear when volume or bass/treble were reduced, also disappeared when the power tube was pulled which.. I mean... duh. Also, turning off the feedback made the squeal slightly louder when it was present.

Initial trouble shooting showed me that I had a few connection that weren't right so I fixed those and the noise persisted. Tried moving a few wires to reduce feedback/induction/etc and no luck. Then I noticed that I had the 115V primary lead connected instead of the 125 volt. So I took some voltage readings and was really shocked to find something out. At the input to the first filter after the rectifier it was showing 370 volts (on the schematic I linked it shows 360 so not bad.) However, when I read on the plate itself, to ground, it showed 338 volts... a drop of over 30 volts across the primary of the OT which, btw, is a Classic tone 40-18030. Furthermore, I had added a 1k ohm screen resistor (and a 1.5k ohm grid stopped resistor) to the 6v6 to try to stop and oscillation, and I found that even though I had added that, my screen voltage was actually HIGHER than my plate voltage at 352 volts (the voltage on the other side of the screen resistor was 358 volts, so dropping 6 volts across it... guessing that the 6ma that the screen is pulling is high but in line with these voltages...) Additionally, I also noticed the damn bias at the top of the 470 ohm cathode resistor was 21 volts!!! I couldn't find any characteristic curves for a 6v6 that were anywhere close to these plate/screen voltages, but I was pretty sure that was a lot closer to cutoff then I liked.... so I guessed that the higher plate and screen voltages were pulling far more current than the original design called for...

Swapping to the 125 volt primary lead on the PT did reduce voltages to where I wanted them. The 370 earlier was reduced to 335. The plate is sitting around 305 now so still getting that large drop. Screen is at 320ish though so still higher than my plate. I checked the bias and it was still freaking 20 volts so I dropped that to a 220 ohm and got that down to 13ish volts. (I've been looking at some of my design books and articles and after crunching the numbers I do realize that is a bit low and causing the 6V6 to red plate a bit so I've ordered some 360ohm 2W ressies which should get the bias where I would like it.)

Side note here, the squeal is gone and I am getting full volume out of the amp now... and getting some deeeeeeeelicious SE class A growl at full volume. Feedback turned off to make it a bit juicier!

So I have 2 questions here.... number one, is that large of a voltage drop normal across these small OTs? From the tube amp design theory books and articles I've read, they generally say the voltage drop across an OT primary can be safely ignored in the design phase as it usually isn't more then 5-10 volts... so do I have a bad OT? Or is it a bad 6v6 with a lower than normal plate resistance causing a larger than normal current draw? I unfortunately don't have a spare 6v6 or access to any, and really don't have the funds atm to be buying parts willy nilly to test.

And second... should I raise that 1k 1W filter/dropping resistor to 2.2k or even higher, or the 1k screen resistor (actually, both of those resistors are 2W to be safe... didn't feel comfortable using 1W resistors in this...) to bring the screen voltage down to at least the same as plate to ground? Or reduced it to be a bit less? Or leave it as is? Is the high screen voltage causing the tube to pull more current than I was expecting and thus solely responsible for the higher cathode voltage?

If you've read this far, thanks! And if yer able to help, thanks again!

 

[Raezzordaze]

So I had a good think since I posted this and I think I understand why my screens are sitting higher than expected. The schematic I used was based off of the AA764 Vibrochamp, but with the vibrato circuit removed. The current draw that the extra 12AX7 added to the power supply was no longer there which meant the voltage drop across the 1k 1W dropping resistor isn't as high and thus the screen voltage is a bit higher. That, coupled with the increased voltage drop across the OT primary (which still bugs me) means that the screen ends up higher than the plate, even WITH the 1k screen resistor I added.

So I'm thinking I either need to increase the dropping resistor or the screen resistor to at least 2.2k. I know if I increase the dropping resistor it'll affect the preamp as well. Just wondering what the pros/cons of that might be... slightly cooler 12AX7? Little less pre-amp headroom, which isn't really a problem in such a low gain amp anyway, right?

 

[2L man]

Smaller cathode resistance value increase bias current on cathode biased tubes!!!

Measure voltage over cathode resistor and divide using resistance and you have cathode current.

Anode current cause voltage loss on OT primary when amp idles. This work only when amp idles. Voltage loss over primary can be used to calculate anode current.

When amp amplify signal the OT inductance has most effect and because of phase difference between currents and voltages this simple math stop working. Sometimes amp oscillate so high frequency that we can not hear it and then this simple voltage measuring does not work and an osciloscope show this.

 

[2L man]

Tube amp operation is based to loadline and this calculator is very good to show what effect it has. When you input data the anode voltage is voltage between anode and cathode so for cathode biased stages g1 voltage is deducted from B+1 and on cathode biased stages it is voltage over cathode resistor. Also voltage loss over OT primary must be deducted from inputted anode box value.

Input headroom which is about 50V less than anode and green area appear. Also predicted distortions. Pointing cursor to grid lines show their value. Their ratio against headroom produce stage gain.

There is one thing which can confuse. Some tubes have grid lines which go to positive values which mean A2 or AB2 operation which in practice can not happen using simple typical drive stages the instrument amps use. On right there are two tables which show max power which can have positive grid drive signals and max power at g1=0V so this later is what should be concentrated.

Universal Loadline Calculator

This calculator allows you to trace the anode characteristics graph, and to compute and tune loadline and operating point of various vacuum tubes.

www.vtadiy.com

 

[Raezzordaze]

I get all that... what I'm worried about is that it seems even at a cathode bias voltage of 20v (which seems to be around what the original 5F1/AA764s had their 6V6s set at) I'm seeing far more voltage drop across the OT primary then the schematics and other builders/designers note. This makes me wonder if the OT itself has a bad primary or if some OTs simply have more DC primary winding resistance. Or if maybe my 6V6 has a higher transconductance/lower plate resistance at then is normal at a given voltage and may be going bad (though I am going to take more readings after swapping back to my 470 ohm cathode resistor and calculate total plate and screen current.) Or if maybe I should use the 5k winding instead of the 8k that the 6V6 data sheets say should be used for load resistance on a 6V6 with 280ish plate voltage. or.. or... or.... etc.

Because, like I said, having the screen sit higher than my plate just.... weirds me out. Lol

 

[dan40]

This old thread from the Music Electronics Forum offers a little insight as to why this is often seen in these single ended circuits.

Screen voltage higher than plate voltage - Music Electronics Forum

 

[2L man]

Similar Hammond SE OTs are 1750C and 1760C? Their 8k primary coil resistances are 330 and 360 ohms so typical 6V6 SE bias 40mA leads to ~13V and ~14,5V voltage drops. If you did measure 13V over 220 ohms it is almost 60mA cathode current and these OT primary voltage drops approach 20V. Cathode current consists about 5% Screen current and technically it should be deducted to get Anode current.

Obviously Classic Tone OT has even higher primary resistance? Just measure it. If you don't get solid reading short output or connect loudspeaker. OT inductance can screw resistance measuring but shorting other coil usually ruin inductance and multimeter measure coil resistance better.

 

[Raezzordaze]

The problem is, from both direct measurement and calculating via voltage drop divided by plate current, I'm seeing the primary winding DC resistance around 670 to 700 ohms. And that is with either the 220 or the 470 ohm cathode resistor in place (I currently have the 470 ohm in place with the bias at -21v.)

What I am trying to ascertain is if this OT is going bad or if this is normal as the manufacturer, who has since gone out of business, cannot be reached and doesn't supply their winding DC resistances anywhere that I've found. Every other similar OT that I've seen in regards to wattage and primary impedances shows around half that DC resistance which suggest either Magnetic Components makes ****ty transformers or mine is going bad. This is all I am asking to be answered here really.

Well, that and if having my screen be higher than my plate is an issue and if so whether it'd be better to drop the voltage using the dropping resistor or screen resistor. Lol

Thanks for the answers too!

Oh, and the screen supply is tapped AFTER the initial dropping resistor, so it's not like that 5F2. Which makes it even more worrisome to me that the screen voltage measured at pin 4 is STILL higher than the plate, and even WITH a 1k1W screen resistor added which the 5F1/AA764s didn't have. Granted... it's only like 5.5ma of current through it so it's only dropping 5.5 volts... but that's on top of the 7ish volts that's being dropped across the first dropping resistor too. I think my next move is gonna be to change the screen resistor to 2.2k and see how that works. I know screen voltage has a much higher impact on plate current than plate voltage does which is why I'm fixating on the screen voltage so much... and especially since I'm still RIGHT at max for plate dissipation on this tube.. maybe even a bit over it. Lol

 

[2L man]

Just feed Mains voltage to the primary coil ends and measure input and output voltage and you know the winding ratio!

You can also connect loudspeaker to OT output and listen Mains signal. Speaker also loads the OT but american mains is not much drive for OT. Here mains is 230VAC and loudspeaker power comes about four times higher and it sound loud and would be too much for small OTs and the feed lower voltage to OTs using a Variac.

Naturally working with hard wired Mains must be cautious and assume that Variacs do not calvanically isolate Mains!

I have a scrap Mains cable which I tack solder to transformer lugs. When soldering to transformer primary wires I tape them against work table top and they don't cause shorts. Few minutes and I have test what the OT impedance options truly are and it function. I have found two different winding ratios and one OT which secondary wires were opposite than on datasheet which all would have had some effect to amp operation.

 

[dan40]

Did you read the thread that I linked to in post #6. It is not uncommon to find the screen running higher than the plate in these single ended amplifiers. A power tube running in class A mode is drawing 100% current at idle so there will be some voltage drop through the resistance of the OT.

Does the OT feel unusually warm after playing the amp for several minutes? If not, I would say that the OT is fine. The Classictone line of transformers were made by the parent company Marvel Electronics which has been making transformer for many of the top amp companies for many years now. It's not unheard of for a new transformer to be bad but their quality control has always been pretty good judging from all of the glowing reports of their transformers.

 

[Raezzordaze]

Did you read the thread that I linked to in post #6. It is not uncommon to find the screen running higher than the plate in these single ended amplifiers. A power tube running in class A mode is drawing 100% current at idle so there will be some voltage drop through the resistance of the OT.

Does the OT feel unusually warm after playing the amp for several minutes? If not, I would say that the OT is fine. The Classictone line of transformers were made by the parent company Marvel Electronics which has been making transformer for many of the top amp companies for many years now. It's not unheard of for a new transformer to be bad but their quality control has always been pretty good judging from all of the glowing reports of their transformers.

I did read that thread but it was about a design that pulls screen and plate supplies from the same part of the power rail so wasn't quite like mine which, like the 5F1 and AA764, pulls the screen supply from AFTER the first dropping resistor. So not only have I added a screen resistor that drops the voltage by about 6 volts, but the first dropping resistor also is dropping the screen (and preamp) supplies by about 7 volts. And STILL my screens, after dropping 13-15 volts from B+, is 15 volts over plate voltage which I was worried about since screen voltage has far more effect on plate current than plate voltage does.

Also, the only real suggestion in that thread was to add a screen resistor which I've already done. Though the bit about decoupling, which I've also seen elsewhere, makes sense and is something I'll be looking into as well.

Since I can't seem to get anyone to answer either of my questions I guess I'll just have to see how it performs. I'm right at max plat dissipation atm so I guess if the OT is failing it'll happen sooner or later. I mean, the amp definitely has output that sounds pretty good to me, I'm just overall worried that the amount of DC plate current is gonna saturate the OT or at least get it close enough that going full power is gonna sound like crap after a while.

Maybe I'll just throw in some temp ressies or clip in some leads and try a few different resistors and caps to ground and see how they sound and what my volts/milliamps looks like.

 

[dan40]

I believe that you will know pretty quickly if it's bad once you have a chance to play through it. If the amp sounds strong with good volume, it's most likely just fine. If it sounds weak and distorted, you may have an issue. Definitely try the feel test with your hand after playing for a bit. It should be warm from the radiated heat of the tubes but it shouldn't be so hot that it get's uncomfortable to place your hand on it.

 

[Raezzordaze]

That's what I was thinking. If it sounds good, just run it till it springs a leak. Just kind of threw me for a loops so see twice the DC voltage drop I was expecting and got worried that long term it was fry my plate and/or screen with the screen sitting higher than the plate like that, especially since I can't really gauge the effects of all that once it's actually transmitting signal.

Think I'm gonna still shoot to get that screen voltage down a bit though by increase that first dropping resistor a bit, and raising the second one so the pre-amp 12AX7 supply isn't dropped too low. That should still give me the same ripple filtering. Might even go ahead and up the filters caps to 40mfd just to counter the ripple a little more.

On a side note, anyone noticed negative affects on these champ-style circuits once a 1k screen resistor is added? Thinking about maybe including a 10mfd bypass cap from the screen to cathode to counteract the screen voltage being swayed by the changes in screen current as the tube conducts more or less, but I'm not sure if that's even noticeable with the screen resistor being so small.

 

[dan40]

I'm right at max plat dissipation atm so I guess if the OT is failing it'll happen sooner or later.

Have you tried a different 6v6 to see if the numbers change a bit? FWIW, many old Fender 5f1's are running their power tube at 15-16 watts of dissipation at idle and some of them are still running on their original power tube. At idle they run pretty hot but once you begin playing through the amp, the dissipation actually decreases in these single ended circuits. This is the reason that the power tubes don't redplate immediately. If this was a push/pull circuit, the dissipation would increase as you begin playing and the tube would die a quick death.

 

[Raezzordaze]

I don't have a spare 6v6 atm so not able to test if it's the tube. But 41 ma plate current seems pretty fair given a 305 plate vdc (to ground that is, not cathode) and a 321 vdc screen votlage, again measured to ground.

Gonna give it some testing tonight and see how things pan out, swapping out the dropping and screen resistors and maybe adding in one of my spare 22mfd electrolytics.

 

[2L man]

Screen current limiter between B+2 and screen protect the Screen. Without it B+2 filter capacitor keep Screen voltage high although B+1 voltage has already drop so low that Screen current begin to climb rapidly. This happen when OT/loudspesker load is not high enough on certain frequencys and loadline pass the g1 zero "below the knee". If on A-class operation this happen the Anode dissipation is well below max. Brief "overshoots" have almost no effect how hot Anode and Screen come but they can sound interesting

Simply saying if Screen voltage is effected it will increase distortion because screen voltage has almost direct effect how efficiently control grid control anode current and signal peaks "flatten"

I think you don't have to worry when Screen voltage is higher than Anode voltage when amp idles at bias point. OT primary resistance just is high!

When power tube amplify signal on SE circuit its anode voltage sweep almost double the B+1 and still producing quite clean output. That is possible when bias current store energy to OT inductance. I find this very fascinating when looking using an oscilloscope. Approaching this using loadline principle explain it great when bias current elevate loadline and loadline spread to the right where functioning anode voltage increase.