Need help sorting out new build (5E7) Success! Clip Added...

[ER]

Update-Thanks for the help everybody!

Check it out here; http://www.youtube.com/watch?v=_G0g7P9CzCw

original post;

Well I finished my 5E7 build over the weekend, it was scratch built with a d-style "universal gainstage" added just before the PI and a single input (just tied one leg of the 270K/270K split at stage 1 to ground).

Well I sorted through some obvious boo-boo's but now I'm stumped, All the voltages seem good (440v at the 6l6's, 300v at the PI and just before the tone stack, 190's on the gain stages and 1.57v at the cathodes).

When I check the voltages I get a nice "thump" at the speakers from the PI onward, and toward the front end just a weak "tap". The problem seems to be somewhere around V2 where the grid (pin 2) is tied to the plate (pin 1) with a 100k and the cathode (pin 3) goes to ground with 100K just before the tone stack. there seems to be a noticeable drop in signal going from the plate to the grid here.

Everything seems to be doing what it's supposed to from the gain stage on, but the stock path input signal is weak and a little distorted, something is wrong in the signal path on the front end, can't figure it out, the thing just won't boogie but I can tell it wants too...

Shouldn't the "thump" get louder as you move toward the input? Is my problem the way I incorporated the single input?

Any ideas?

[img:975:583]http://www.davidsonamp.com/sf/images/bandmaster5E7.gif[/img]

I'll recheck R values and solder joints this evening, but I'm stumped...

Thanks, ER

[rfgordon]

Go over the V2 carefully and make sure the cathode follower is done correctly. Also, you can see on the schematic the local negative feedback loop from the output of V2b is fed back to the grid of V2a. If that resistor (10 Meg) is smaller, you'll squash the signal.

The reality is that if you leave that one off, you've got a very gainy amp, even without an added gain stage.

[M Fowler]

ER,

The problem seems to be somewhere around V2 where the grid (pin 2) is tied to the plate (pin 1) with a 100k and the cathode (pin 3) goes to ground with 100K just before the tone stack.

Your quote is puzzling because when looking at V2 pin 2, pin 1 and pin 3 the layout does not look like what your describing?

Mark

[rfgordon]

One more thing--the 100k cathode resistor on the cathode follower may be too high for some modern production 12AX7s. The voltage there will kill them because they have lower-than spec tolerance for heater-to-cathode voltage difference.

[ER]

I didn't follow the schematic exactly, V1a = first gain stage, V1b= second stage, V2a= cathode follower, V2b= switchable gain stage and so on...

[Phil_S]

On the 5E7 schematic...

Both triodes of the 12AY7 (V1) are the first gain stage. The two halves are at the same gain potential, each taking input from the instrument jacks. The 12AY7 is followed by a mixer, a pair of 220K resistors.

The mixer is followed by the first half of the V2 12AX7. This is the second gain stage. The next half of V2 is the cathode follower (CF for short).

The first half of V3 12AX7 is the third gain stage. If feeds into the phase inverter (PI), which is the second triode of V3.

The job of the PI is to feed signal in alternating cycles to each power tube, so that one power tube gets the top half of the sine wave and the other gets the bottom half. This is a rather imperfect arrangement and the explanation is way over simplified, but good enough for now.

There are many type of PI. Your is a cathodyne PI (a/k/a concertina or split load; all three names are interchangable). The cathodyne PI only needs half of a 12AX7. It puts out half the signal on the plate and the other half on the cathode. Voltage output is not balanced, but current (mA) output is supposed to be balanced.

The PI sends the signal to the power tubes for final amplification and output to the speaker.

I'm not clear what you've done differently. I think maybe you modified the 3rd gain stage? I hope this clarifies the discussion for you.

[ER]

I used a 12ax7 instead of the 12ay7 and have a single input, so v1 covers the first (a) and second gain stages (b), V2 is cathode follower (a) and OD (b)post EQ, V3 is final gain stage (a) and PI (b).

OD is 100k resistor into 100K drive pot, 100k/1500 5uf plates/cathode, and 100k level. with overdrive switched out circuit is (relatively) stock other than 12ax7 in first position and slightly higher voltages.

This is very similar to what I built other than the placement of the OD (post EQ) and some of the OD stage values and no master volume;
[img933]http://www.kilback.net/homebrewtweaks/a ... od3_r3.gif[/img]

[tubeswell]

Hi ER

Apart from double-checking the wiring on your DC coupled pair (the voltages look slightly odd - the CF cathode looks too high for the given HT voltage. Are you using a 12AX7 in there or another tube?)

What is the idle voltage at Pin 6 of the PI tube? The cathode voltage seems high there. Have you tried the PI with an ordinary (5E7) PI cathode bias? (i.e. taking out the Paul C mod)

Also you are attenuating the signal heavily after v1. The 220k/220k voltage divider is in parallel with the bottom of the 1M pot. This would make the input impedance of the 1st stage of the DC-coupled pair very low indeed (too low). Try dumping the extra voltage divider and just having the 1M pot. 2CW

[ER]

I didn't have the paul C mod in mine, I went pretty much stock other than the OD and single input.

I followed RFGordon's advice (Thanks RFGordon!) and checked the feedback resistor (the only one I DIDN'T check before putting it in because I was so happy I found a 10M resistor at the surplus place!) well it measured .339M!!! Well I clipped it out and the amp is very gainy as far as drive (I'll have to find another 10M resistor somewhere, a good one this time), and slightly louder but still no Rock and Roll. Sounds about as loud as lobby music at the doctors office...All the other R values are surprisingly close to what they are supposed to be (better than 1%!).

I started wondering if it was the 4-12 test cabinet I threw together was doing something weird, so I hooked up my 40 watt vibroverb since it's about the same power level I'm expecting from this bandmaster...nope, not the cabinet other than needing some better screws for the back panel.

TUBESWELL: My voltage divider is 270K/270K and that's the one thing I'm really fuzzy on; I followed the circuit on the skiltone scheme for that part but used the stock fender values (270k instead of 220K). Are you saying just ditch the voltage divider and run the wiper to the grid of the following stage?

Thanks for the help everybody, I'm learning a lot from this project and I'm looking forward to how this amp will sound under full power.

I started this project as a scratch build using some iron from a baldwin/leslie booster amp and looking for something to get that early Montrose tone (without the big muff). The voltages are a little higher than stock (still within 20%) and are just about what you would expect plugging in a 110V late 50's fender into a 120V wall socket which is what I was going for...

Thanks, ER

[tubeswell]

TUBESWELL: Are you saying just ditch the voltage divider and run the wiper to the grid of the following stage?

You would get more gain out of it without them there. Up to You

[Phil_S]

I'm not clear about the 10M resistor. I don't see it on the schematic, but there are lots of parts and it's easy to miss something.

Anyway, have you lifted the whole NFB/Presence out of circuit? Disconnect it at the cathode end, not the speaker end. I have no real basis for it, but I'm thinking that loop is re-biasing the last gain stage in an undesirable way. If it isn't loud, signal isn't making it to the concertina, which has 0 gain. Look before that. I think that's why I'm suspecting the NFB/presence.

[ER]

Phil

No I left the presence circuit alone, the 10M feedback loop runs from just after the treble pot back into the grid at stage 2. I cut it to see if that was the problem with the low gain, it wasn't. I'll try disabling the Presence and see if that helps.

It sounds pretty wild and loose without the local NFB, so it will go back in once I get to the bottom of the gain problem.

I may visit the presence circuit at a later date for tweaking since I'm using an 8 ohm OPT and speakers, and the original used a 4 ohm OPT and 2.8 ohm speakers (3x10), but it shouldn't be too far off since this was a similar circuit for the tweed super etc.

When I test the voltages at the concetina, and even the final gain stage the speakers really jump, so the whole power section seems to be good, but the presence comes in right at that point, so that's a good clue thanks.

I'll try that first before going throught the trouble of removing the voltage divider after the volume pot this evening, I'm hoping that all this does the trick.

Thanks, E

[M Fowler]

I usually remove NFB from my amps as I see no benefit/need in applying feed back. The amps respond well without it. That very well maybe the problem area.

[ER]

Well I removed the feedback loops (local and presence) and it really opened things up tone wise but it still lacking umph, it sounds like a really nice 5 watt recording amp...still very quiet with out OD engaged, the OD adds more volume but not enough to play with a band, does a nice "clean to mean" with the guitars volume knob.

I pulled the 270K voltage divider after the volume control and it didn't do much other than introduce hi freq. occilation turned up anywhere past 2 o'clock or so.

The volume knob is only effective in the first 1/4 of it's sweep, the drive knob (OD input) doesn't seem to do much of anything other than affect the OD tone, the level (OD output) works great and is the only place to get any volume added by turning it up.

Might be time to go back to the drawing board on this one, I already have a 5 watt recording amp that does this trick...

-ER

[tubeswell]

Hmmm... this sounds wrong. That amp should have about 35W. What are the idle voltages thru-out (if the schematic you posted earlier is not what you built)?