New JTM45 build-low power output, early distortion

[mfreqmaster]

Here are the pins 2 & 7 of the three preamp tubes:

V1 p2 2mVAC p7 106 mV (plugged into the bright channel for all these tests)

V2 p2 1VAC p7 37VAC

V3 p2 1.6VAC p7 .76VAC

[g1]

It looks like you have fairly healthy signal at V2 pin8, but not much after the tone stack at V3 pin2.
Do all the tone controls seem to be working?
Maybe try bypassing the tone stack as a test. Disconnect the wire from V2pin8 at the board end. Disconnect the wire from the middle lug of the treble pot. Connect those 2 wires together.
Any improvement?

[mfreqmaster]

I disconnected the wire on pin 8 of v2 and connected this to the wire originally on the middle lug of the treble pot. Output seemed to be slightly louder, but more flubby & certainly not cleaner than the original. Actually the first time I tried this the volume was way down, I figured something had been moved during my desoldering. Everything is put back the way it was originally now. Would an audio file be of help?

[Chuck H]

The colors is your photos aren't entirely clear, but it looks to me like you may have some incorrect resistor values in there. Can you take an extra close and clear shot of the preamp area?

EDIT: It also looks to me like there are a few questionable solder joints. Two eyelets look like they don't even have solder in them!?!

[g1]

Agree with Chuck, you are using precision resistors with 4 band colour codes. I've never got too comfortable with those and always double check their value with resistance meter.
And a resolder is in order.

[g1]

I disconnected the wire on pin 8 of v2

You left it connected to pin8 and disconnected it at the board end, yes?

[mfreqmaster]

Here is hopefully a better close up of the preamp section. You are probably no doubt looking at the funky series parallel input resistor combination I have for V1. I didn't have any 68k resistors on hand so I made some combinations that got me pretty close. Did check these with an ohmmeter. I know it looks crappy & was going to eventually replace with proper valued resistors, but thought since I got the value close could still proceed with the amp to check it out.

Resoldered the preamp area of the board since taking the photo this evening. One thing about this amp that is different from others I've built in the past is that it seems to lose voltage on the power capacitors real quick after turning the amp off. Seems like an amp should be able to shock the piss out of you if you touch it in the wrong place, even with the power off but this one measures about 12V on the caps right now, amp hasn't been off that long.

Yes, the wire was connected to pin 8 but taken off the board where it connected the 100k resistor/250 pF capacitor.

[Chuck H]

You are probably no doubt looking at the funky series parallel input resistor combination I have for V1.

Nope. I was looking at other, single resistors and their locations. Something looked funny. I see color codes on your board of brown/yellow/black where there should be 100k values which I know to read as brown/black/yellow. There are some other funny things too but I didn't trouble to nail it down. No reason you can't do that since you have the amp right in front of you. Many solder joints look like they could use a re flow too.

[g1]

I see color codes on your board of brown/yellow/black where there should be 100k values which I know to read as brown/black/yellow.

The blue ones are precision resistors using 4 band color code (5th band is tolerance). So the third band is another digit, and the multiplier works on the 3 digit number. So 100K is brown black black orange.

I disconnected the wire on pin 8 of v2 and connected this to the wire originally on the middle lug of the treble pot. Output seemed to be slightly louder, but more flubby & certainly not cleaner than the original. Actually the first time I tried this the volume was way down, I figured something had been moved during my desoldering.

This doesn't make sense to me, it should be way louder.
When you did your AC signal measurements in post #29&31, you had 27VAC at V2 pin8, but only 1.6VAC at V3 pin2. The tonestack bypass from post #32 should have put 27VAC at V3 pin2 (instead of 1.6VAC). That should have made a huge difference.
If this makes sense, it may be instructive to retry the tone bypass, then redo the test measurements from posts #29&31. See what kind of increases you get at V3 and the power tubes when you have 27VAC going into V3 pin2.

[Justin Thomas]

In general, certain values of those resistors are tricky - I've put in several that I had read backwards. When your tolerance bands get into red, brown, etc, it's a real PITA. And since it's usually the 1/2W metal films that do it, there's not much room on the body to leave a gap between value bands and tolerance band.

Chuck, I'm reeading those (plate resistors?) as brown black black orange, which in 5-band is 100K. But yellow on blue makes orange, so I initially thought 1M. Too hard tp tell... But agree - all those 5-band MF (Mother #$%&er) resistors should be metered out.

Re: losing voltage quickly, all my amps do that, yet are not lacking on power. They go down to about 16V in seconds. If I leave them overnight or a few days, they'll build back up to about 30V, maybe 40. I don't think that's the issue. That'son all three builds and my numerous old Fenders. Now, with the tubes out, it's a whole new ballgame. Zapped myself a couple times with the tubes out... power on with no tubes and watch out.

Justin

[Chuck H]

Thanks for the correction on the five band coding G. It still looks like yellow rather than orange to me on those "100k" circuit positions. The voltages don't seem far enough off for that to be the case though. Test them anyway. No one gets hurt.

Justin, you said yellow on blue makes orange. That's not exactly true. If the yellow offers any transparency at all then yellow on blue makes green. One reason yellow on blue may LOOK orange is that orange and blue are directly contrasting (Please, no physics geeks need to tell me about light wavelengths. I'm discussing pigments now which are SUBTRACTIVE) If the same yellow used on brown resistors (which has a tad of red in it so it won't show green) is used on the blue body resistors and is relatively solid it would look somewhat orange from contrast.

[Justin Thomas]

Well, in MY world, yellow & blue makes orange! Musta been something funny in that sushi last night. Thanks! Sometimescyou just have to say your thought instead of asserting facts, maybe?

Justin

[Chuck H]

Spectra work differently for an additive vs. a subtractive scenario. I use to be a color tinter for a couple of paint stores. Not the goober mixing Mrs. Brown's counter order though. I made colors from samples, not formulas, before there were these new fangled photo spectrometers. And I've had an earful from guys that treat color strictly as a property of light. My job was to understand what colors will not be reflected by a pigment and is only peripherally related to the analysis of rainbows. I did need a some awareness of the additive nature of light because indoor lighting varies a great deal. So I had different light stations set up for viewing results and would use every one for each color. It's not much more complicated than an artist mixing paints on a pallet in reality though. It was my consideration of the other principals involved that made me better at that job than most.

I know how exciting reading this must have been. Congrats to anyone who stuck it out.

[mfreqmaster]

I tried again removing the wire on pin 8 of V2 at the junction of the 250pF cap and 56K resistor, attaching this to the wire previously on the treble pots middle wiper. I ran a 100mV sine wave through it and took AC voltage measurements at valves 2 and 3. Also, maximum AC voltage at output now only 2.8V. Possibly due to changing to 100mV input signal? Here are the AC voltages below:

V2 pin 1 94V, pin2 2.4V, pin 3 .7V, pin 6 19mV, pin7 94V, pin 8 1.14V

V3 pin 1 3.7V, pin2 .7V, pin 3 1V, pin 6 3.9V, pin 7 1.1V, pin 8 1V.

Plugged the guitar in, again the tone was distorted and not consistently loud. I searched for any loose or poor soldered connection in that area & could not find anything. When I touched my probe to V2's pin 8, the volume did get louder (meaning back to the engaged tone stack level of loudness) and went away as I took the pressure off. I made sure that wire had a solid connection & I thought, socket problem. These are new sockets and appear fine.

Put everything back together as original. Amp sounded the way it has in the past. Compared loudness with my ears only, switching back and forth between this amp and my 5E3. Both about the same overall volume, 5E3 is slightly louder and cleaner. The 2 power tubes appear to be mismatched and the issue follows the tube not the socket. V4's voltage at pin 1-8 about 10mV less than V5. As mentioned earlier I put a set of 6L6's in here just to try out, no real change in volume. I'm almost to the point of tear everything out, get different components to populate the board, and try again.

[g1]

Somewhere your data is wrong.
You speak of changing to 100mV input signal.
We are trying to compare to posts 29&31.
At that time you had 27VAC at V2 pin1. So these new measurements can't really be compared.
You really need to choose some set levels and not change any settings.
Readings without jumper, then readings with jumper installed.
How could there now be 94VAC at V2 pin7, then 1.14VAC at pin8 ?
In posts 29&31 those readings were 37V and 27V.