Amp repair?!

[Dimi Pana]

OK, sorry for the late reply, got busy at work this past week. However, I was able to do some more "digging" and this is what I found. Also, keep in mind that all following measurements were taken with standard 125VAC grid voltage (e.g. without using the step down transformer) and with the tubes inserted, speaker connected, no guitar connected and volume all the way down. I did not let the amp stay on more than 1-2 minutes, I took the measurements and powered-OFF. Nothing "bad" happened, tubes were tested and found OK, no ...smoke or funny smells, you know what I mean eh?! 

So, here's what I found:

#1. The PT outputs 7.8 VAC which feeds the tube heaters circuit. As correctly pointed out by Phil, the heaters of a 12ax7 and a EL84, combined, should draw about 1A. However, an actual measurement showed that at power-ON the current "rushes" to a little below 3A but it quickly starts falling and settles around 1.8A after about 30 seconds or so (I am assuming this is when the heaters have reached full operating temperature). Now, if I measure the heaters' voltage at that point, I get a consistent 8.5VDC at both of them.

#2. Through trial and error I found that the amount of resistance needed to bring the heaters voltage (measured at the pins with tubes inserted) down to its optimal range (e.g. +/- 5% of 6.3 VDC) should be between 0.5 and 1Ω. How so you may ask? Well, I had a few 5W wirewound resistors (5, 2.2 and 1 Ω) and tried them all. I started with a 5Ω but it brought the voltage too low (about 1.85VDC) the tubes did not even glow, then I tried the 2.2Ω, voltage went up but still low so tubes were barely glowing, then I tried a 1Ω and the voltage went up to about 5.38VDC, the tubes did glow but I ...chickened and did not try to play through the amp, I was afraid I might do some damage.

So, now I have two questions:

A. How do I (accurately and confidently) calculate the proper resistance value in order to achieve smack 6.3VDC at the pins of the heaters?

B. Assuming this is accomplished, can I consider this a solid/final solution or is it a necessary workaround/hack that works but might cause problems down the road. If it is not a, then what do I have to do to bring the voltage down without using the brute force solution (in series resistor). Someone I asked mentioned something about building a simple zener diode voltage regulator. I googled that and from the little I read it sounds like a viable solution. Can anyone please weigh in on this?

And as a ..."bonus question" can anyone explain why instead of about 1A, I measured 1.8A drawn by the heaters circuit? Also, Phil, by "block connectors" you mean that in the pic right?

[phatt]

Yes that is what i meant

You can bolt them onto the chassis and run wires,, here's a pic for example of how to do it. The big 10Watt R's are mounted on Terminal/Connector blocks,, well that is what I call them.

OK you found 1 Ohm is very close then you need to halve that. Well try .5 Ohm
 If you can't get OR47 (.47) resistors then Two 1 Ohms in parallel will give you 1/2 an Ohm.

It does not have to be Exact,, anywhere in that 10% window is fine (5.5Volts up to 7Volts).
A Zener would need to go on the DC side and yes it works but another fail point down the track as Zeners fail more often than resistors. up to you.
The wall outlet will go up and down a bit and that alters the result,, depends on time of day you measure the voltage.
So EXACT will never happen.
Phil.

[Dimi Pana]

Success! 


Well, it is is not finalized yet but I managed to bring the heaters voltage down to 5.89 VDC, so then I was able to play with the EVJ for about an hour, clean, overdriven, pedals, straight into it, no funny sounds, high or low volume, no crackling, amp is not overheating, and most importantly it did not blow any tubes.

So the method Phil suggested has been tried and it works, I had to use a combination of four (!) resistors, 1, 2.2, 5 and 15 Ω (that's what I had available at home), in parallel, to achieve the lowest possible resistance value, which in theory is .58Ω but in practice my cheap DMM showed 0.8 Ω (well these were wirewounds at 10%).

Still 5.89 VDC is on the low side (not sure if that's good or bad) but sound/tone wise it had no perceivable change. I think an actual value between 1/2 and 3/4 of an Ohm should land me in the sweet spot of around 6.3 VDC, and actually I'd rather run a little on the "colder side" for tube longevity, am I right about that?

So a few questions:

1. As expected, the resistors got a little warm, still was able to comfortably touch them I mean after an hour of operation, they did not burn hot at all, so that's good. I could be wrong but the lower the resistance the warmer they appeared to be, the 15Ω was almost not warm at all. They were all 5W at 10% so what do you think for the final fix, should I get one at a higher wattage (thus bigger) or it is better to split the load (heat) across many of equal value. And I do not really need a heat sink, just bolt them against the amp chassis, right?  Also can you suggest where I can buy such resistors?

2. This is more of a philosophical question but it is funny how the PT the designers decided to use is putting out more voltage than what it is healthy for the tubes. Or you think the PT is developing a problem? From previous posts I understand that if the PT is on its way out, most likely it would have provided less voltage not more, am I right?

Thank you all and especially Phil for sticking with me along the way, so far this has exceeded my expectations, I really appreciate your help!

 

[Enzo]

You are WAY overthinking this, in my personal view.  The difference in tube life between 6.3v and 5.89 is negligible.   Phenomena like that don't necessarily "track" high numbers to low.  In other words having 9v there instead of 6 will have a large effect.  Having 6.9v versus 6.6 won't be detectable.

If trying to measure low resistance, always first touch the meter probes together to see how far above zero they result.  Then subtract that from any resistance readings you take.  If your probes have 0.2 ohm resistance, and you measure a 0.5 ohm resistor, the reading will show 0.7 ohms.

Transformers don't have a failure mode where their voltages drift.  They work on turns rations, and are simply wire wrapped around a form inside.  They don't wear out and lose voltage.

I doubt any designer sat there thinking he would run excess heater voltage through anything.  I suspect mains voltages change, yours, his, everyone's.  The load can change.  Perhaps the tubes he was using at the factory drew more heater current than the type you use.  In any case, I would assume the designer specified a transformer that put out about 6.3v at the designed load.


As to parallel resistors, Ohm's Law tells you the lower resistance will pass the most current, they won't share equally. 

FInd one resistor of the value you need.  yes, electrically ten resistors that add up the same will also work, but really...   Is 5w enough?  I don't know what is really flowing through your circuit.  Want to use a 10W?  Go ahead, teh cost is a few cents more, so why not.

[Dimi Pana]

You are correct, overthinking is a novice's natural reaction while treading unknown territory for the first time. IMO, I'd rather overthink than underthink, with high voltage amps it is also safer, I think.

Now, you have given me a lot of good advice, i.e. indeed the probes amount to about 0.2 Ω so the actual value I measured is about 0.6 Ω which is indeed within the theoretical 0.58 Ω. Ι looked up online, cost will not be an issue, any such resistor (.5 to .75 Ω @ 5 or 10 W or ...more) is going to cost way less than actually shipping it first class mail. Not worth ...overthinking that.

Can you also please help me with the following: Without inserting the voltage dropping resistor, there are two (orange color) wires coming from the PT providing 7.8 VAC for the heaters circuit and the tubes according to their data sheet should draw about 1A combined (in practice I measured 1.8 A but what do I know?). Anyway, according to Ohm's law the power should then be (that's easy P=VxI) a theoretical 7.8 W to an actual 14.04 W. Provided that my thinking is correct, a 5W resistor is insufficient in both cases. It is my understanding that the more power a resistor is rated at, the better it dissipates the heat. So what is the recommended value in my case?

Also, what is best: One resistor for the entire heaters circuit in series to one of these orange wires, or should I split the total recommended Ω value in half and add one resistor per each wire?

Regarding the PT voltage (and with my limited knowledge about amps) I doubt the designer screwed up so bad I mean the schematic I got says the PT provides 6.3 VDC but I am reading close to 8 VDC in mine, which after rectification you well know is going to be over 10, so as Phil said, someone screwed up big time or something is iffy with this particular PT. I still have not been able to borrow a friends identical EVJ to compare with mine, it will be interesting to see what's going on.

Anyway, thanks also for the info about running cold or hot, since I still have to get the proper resistor, I will try to tweak it just a tad higher, just to be OK according to the tube's data sheet. Btw, a few weeks ago when I started having problems with this amp I had no clue about heaters voltage etc. Now I understand that tube manufacturers say 6.3 V is ideal, a +/- 5% is acceptable and the maximum tolerance is up to +/- 10%. So, I hope around 6 V is OK and I heard somewhere that less heater voltage might help the amp distort a little later and the EVJ being as a simplistic design as it is, it does suffer from that also.

Again all this is just an excuse to play and learn with an amp that I can afford to ...destroy (not intentionally of course) so thank you all, and thank you Εnzο for the input.

[phatt]

Enzo has spent his life working on this stuff so I'd take his advice. 

Obviously .5 Ohms is all you need and gets you into the -5% mark which as Enzo has tried to point out is perfect.
IF heat worries you then use a 10Watt but two 1 Ohm 5 Watt resistors (in parallel) will give you .5 Ohms 10Watts
As long as the resistor/s are not giving you first degree burns then close the lid and it's done. HINT.

Regards mounting;
As all those resistors are only warm then mount them in a convenient place as long as it's away from the mains.

Think about the mounting point of your resistors before you cut the orange wire as you may have enough wire length in the orange wire to not need extra wire.
And No you don't need resistors on both sides of the AC.

Re transformers;
No the person who wound the transformer made no mistake.
The clown back at the factory who was under pressure to reduce hum due to bad track layout thought hey we can run the heaters on DC but forgot that doing so puts the heater voltage way over the limit.
I have a home built 10 Watt Valve Amp that has run on 6.9VAC heaters for over 15 years and it always runs flat out
Phil.

[Enzo]

There us no reason to use two resistors.  The current is AC and flows both directions.

Power is determined by voltage and current.  A resistor will dissipate whatever those data indicate.  A 5 watt or 10 watt or 20 watt will dissipate the same watts.  Larger ones will just be a less concentrated hot spot in the chassis, that is all.

Ohm's LAw pertains to the resistor.  Remember, the entire 7.8v is not dropping across the resistor.  The resistor is there to drop 1.5v.  So at 1A, that would mean 1.5 watt by my math.  1.8A?  1.5 x 1.8 = 2.7 watt.  So 5w should be fine or 10W if you have the space.

At some point you mentioned something to the effect you had the "standard" 125v mains.  Well 125v is not a standard even if common, look at old schematics.  120v.  I grew up with 117v.  115v, and older still 110v.  A transformer could have been wound for any of those standards.  If you run a 110v transformer on your 125v, then the secondaries will all be 13% high.  Now what results from an increase of 13% at 6.3v?  7.2v

[Dimi Pana]

Thank you both for the feedback, I do appreciate the free tutorials, and as a formerly clueless end-user, I can now confidently assert that hey, when it comes to expert opinion there is no substitute, you have definitely opened up my eyes to matters that I was completely unaware of. 

So,

#1. The amp is fixed and packaged back to its original state and it works REALLY well. I was able to find and ..."borrow" a few low Ω resistors from work, and after some mucking around, guess what? A .27Ω resistor (5W @10%) brings the heater voltage to its ideal range. Actually, at start up it shoots up to about 6.38V then goes down to around 6.28V and finally stabilizes exactly at 6.3V. Sorry for being pedantic, I was just playing with different values and just got lucky! I used some spade terminals soldered to the resistors leads dressed the connections nicely with heat shrink tubing, found a nice spot away from the mains and used some existing holes in the chassis (the ones where the heater wires enter from the PT) to thread a zip tie that fastened the resistor against the chassis with enough space for air to circulate around it. Like already said, even after of plenty of time of playing the amp, the resistor is comfortable to the touch, warm but not hot, actually while still in the open (e.g. chassis not back in the cabinet) a laser thermometer showed about 78F and after I put it in the enclosure that spot (touched from the outside) never felt hot, actually it was barely warm, perhaps because the chassis now dissipates the heat better(?!?). So it is done and THANK YOU SO VERY MUCH both/all!   

#2. I installed a new JJ EL84 and instead of a 12AX7 I opted for a new 5751 also JJ. The sound is sweet, it distorts much later and although it is probably the psychological aspect of it, I think it sounds better than before. So, looks like my $75 investment at a recent garage sale paid off. Even better, now I know a lot more about how this thing works (my first tube amp, btw) I did not spent a dime for the repair (other than time and readily available parts) and guess what, all this reading made me realize the EJV is probably one of the most modded low-power amps out there, to the point that I now feel ..."dangerous" enough to attempt some of this mods. I know what you're thinking, we'll never get rid of this guy, he "will be back"!!!   

#3. I was also able to borrow that other EVJ from my friend and guess WHAT?!? As far as voltages, it is EXACTLY THE SAME !!! The PT provides 7.6VAC for the heaters and with my house's 125 VAC mains (btw, is that normal, or a bit high?), and tubes installed, when measured at the pins, I read a ..."nice" 6.93VDC. My friend said he had this amp for about 5 years now, he does not use it for gigging, just home practice/jamming, and he has never changed tubes at all!?! Again it is probably my ears, but my amp is quieter at any volume level and actually if cranked mine sounds more defined, the other is getting "muddier" earlier. Anyway, ... So you were of course right all along (not that I doubted any of you) still I am not sure why mine had the problem I experienced which led to the discovery of the "elevated" heater voltage, but honestly, I am not going to ...overthink this any more/longer.   

#4. Regarding the wattage measurements, I used the 7.8VAC coming from the PT, and the 1A nominal current the two tubes draw per their data sheets.  But then I measured with my DMM the actual current and found it to be 1.8A, thus the widely varying figures I stated in an earlier post. I now understand how it really works, thanks Enzo and Phil. I am not sure why though, currently with the fix that draw has gone down to about 1.35A so there is still something on top of the filaments that draws the .35A current. Not worrying about it, after all the schematic does not show the heater circuit so I do not know if there is something else in there other than the heater's filaments.   

#5. This is not a question but I would be ungrateful if I did not mention again how appreciative I am for your help. If you guys ever happen to visit the "Windy City" PLEASE drop me a line, I'd like to return the favor.

 

Cheers to all and take care!       

[phatt]

No worries, glad it worked out.

If your mates amp reads 7.6VAC and the heaters read 6.93VDC then it maybe a different model/revision.
There are several ways to rectify AC into DC and they all have different outcomes.
It may only have a halfwave rectifier which would yeild; .9 times the AC.
So 7.6VAC x .9 = 6.84VDC.
A much better outcome for the heater voltage and half wave Only requires one Diode,,, but Full wave gives better noise results.
It's debateable if you would notice the difference in buzz or hum.

http://www.hammondmfg.com/pdf/5c007.pdf
Note the maths below each type of rectifier setup.

Re heater current draw.
EL84 draws 760mA
AX7 about300mA
There will always be deviations due to manufacturing processes. So don't expext perfect results.

And don't forget the meter you use may not give the exact same readings as another DMM.
Oh and be careful reading mains voltage with cheap DMM's as some are not really up to the task.

Re mains voltage fluctuations;
I think **Enzo** has already made comment on that?
It fluctuates all the time and the amount depends on where you live,, how far from the power station,, how close is the booster transformer,, are you at the end of a long line,, and so forth.

Re zip ties;
They are plastic and will melt or at least stretch if that 5 Watt resistor gets hot.
Which is why I mentioned block connectors as you can screw them down.
You use a double on each end and that gives you one mounting hole on each end of the resistor. This gives the resistor a solid mount at each end.

The noise problem explained,,
And why DC heater still won't fully fix bad hum:
I quote from this page
(http://web.archive.org/web/20071024154401/http://users.telenet.be/svokke/valve%20junior%20mods.htm)
--------
"The layout problem is mainly in two places. At number one there is Crosstalk between the heater and the trace from the volume pot wiper to V1B. At number two there crosstalk between heater line and the trace between anode of first gain stage to anode load resistor (R3). I can't believe that they didn't discover this at epiphone. After all, they should be the pro's. There is also some crosstalk between heater and the trace between anode of second gain stage to anode load resistor. On this location it is less important though because the gain following this stage is low."
--------
Same old problem as my mates Fender Pro Junior.
Cross talk between tracks.
Obviously people in front of a cad program who have no idea what happens when you put high voltage tracks right next to noise sensitive hiZ tracks.
they hit the auto route and go to the pub,,meantime the consumer has to put up with
The HUUUUUUUUUUUUUUUUUUUUUM.

This particular model of Pro Junior was a shocker and I fixed that by cutting tracks and running shielded cable direct to valve pins.
hum gone with no DC heater needed

Phil.

[Dimi Pana]

Excellent points, now I have lots of reading to do. Hmmm, I did not think about the zip tie getting loose because of the resistor heat, it's not much but it does warm up. I wanted a quick solution so I can put the amp back together and play. I will revisit this hopefully in a few months, I intend to do a few mods to this amp, the double block connectors is a solid solution I will do that. My friends EVJ is indeed a newer production (judging by the serial number) but still the exact same version 3 (identical PCB, components, etc.). So I am not sure what's going on but for sure this amp was designed by mistake/accident who knows, with heaters voltage higher than what it should be. Anyway, I am so glad mine now works and (to my ears) it works well. Thanks again, cheers!