Homebrew reverb problem - Send side I think

[MattT]

I just built a homebrew amp and tried to shoehorn a custom reverb circuit into it. It uses a BF Fender type Send: a paralleled 12AT7 driver with a 2k2//25uF cathode, 022921 Reverb tranny connected to the plates (tranny red wire), 1M grid load. Input is through a 560pF coupler off of the second gain stage. Pan is an Accutronics 4AB3C1B. I have 382vdc supply (rev tranny blue wire), 376vdc plates, 7.6vdc cathode. Black wire is to ground, green wire is to the Reverb Send jack.

When I input a 880Hz/100mV signal (true rms) to the amp and turn up the Volume to a reasonable level I have ~160mVac signal at the 12AT7 grid. Then I have ~3.1Vac at the plates but then only ~52mVac at the Reverb Send jack (I can barely even measure it). Then I get nothing out of the pan.

Also, when I scope the 12AT7 plates the line is jumping/wobbling up and down (even with no signal, the horizontal line is jumping/wobbling...and not just a little).

Another thing...when I measure tip-to-sleeve of the Reverb tranny Input I get a very low value...like around 1ohm...its so low I can't really get a good measurement...I just know it's more than the leads by about 1 ohm. This is an unused pan I bought new years ago. It looks mint.

Last question: Is it OK to run this with no pan? It's an OT so I didn't think running it with no load was kosher...but I think I may have accidentally done it.

Any help appreciated. Thanks.
Couple of pics...a face only a mother (or father) could love.

[loudthud]

It's normal to read a really low resistance at the output of the reverb transformer. You could just connect a small speaker there as a test. Give the reverb pan a little bump, you should hear a "KaBong" sound thru the amp. Try injecting a signal at the reverb return jack and see if it gets to the speaker output thru the reverb control.

It's not a good idea to run without a load for the reverb but it is unlikely that you hurt anything.

[MattT]

Thanks loudthud. I expected a low reading at the OT but expected a slightly higher reading at the input of the reverb pan (when disconnected from the amp). It's actually closer to 1.5ohms.

Those are good things to try. Thanks.

[MattT]

Send jack into a little 6 ohm computer-type (unpowered) speaker I get sound...it's not loud but I get sound.

If I jumper the jacks; the signal gets through and gets amplified quite a bit in the Return stages (I did all this into a load and I'm just looking at it on the scope...not listening).

I guess what I need to know is the ballpark signal levels going into and coming out of the Reverb pan. I'm not sure if my Send signal is just too small or what.

[tubeswell]

Maybe test your reverb tranny back to front with a variac I think you should get about 56VAC off the primary from a 1VAC input into the secondary. (Impedance ratio is 3,125:1, therefore turns ratio is 56:1 - right?)

Did you say you had resistance in the pan? If the input or output is open, the respective transducer has haddit.

[MattT]

Thanks Tubeswell. I'm getting signal out of the Reverb OT..I would think I could measure this with the reverb tranny in-circuit. Wires are buried and solder joints are shiny...I'd hate to unsolder all this to measure it when I'm still not sure what I'm looking for. But I'm guessing the tranny is OK since I'm getting sound when I hook it up to a speaker.

[tubeswell]

Yep you can test the tranny if you unplug the pan and pull out the driver tube, and disconnect one of the tranny grounds (probably the pan end is easier) - you don't want the variac supply touching the amp chassis when you measure it this way tho. I was wondere if there was a partial short in the tranny. What was the situation with the pan inputs? (open/closed?)

[MattT]

When I measured the pan the output jack is the one that has it's ground common with the chassis. IIRC I measured something like 217ohms across the output transducer (tip-to-sleeve). The input transducer measures around 1 to 1.5 ohms tip-to-sleeve and is isolated from the chassis. No connection between either input lead (tip or sleeve) to the chassis or anything else...it appears to be totally isolated.

[tubeswell]

Those DC resistance measurements on the pan are about spot on - I've got a brand new 4AB3C1B sitting in the cupboard and I just checked it. There is no connection from the input sleeve to the pan chassis, and that is deliberate (to eliminate ground loop). (Remember the 8 Ohm input impedance and 2.25k out impedance are resistances to AC, so you can't measure that with a DC meter).

Similarly I have a new 22921 tranny sitting in the same cupboard and the primary DC resistance is 1.09k and the secondary DC resistance is 1.5R, and there is no reading between the primary and secondary winding.

Did you check your send/return cables for bad connections/circuit?

[MattT]

Thanks for your help. It works! It appears that when I measured the pan and messed with those little connectors inside that something changed (probably bad connection) and it now works. I didn't even play the amp after I measured the pan until today and the damn thing works.

I have to now deal with some hum when I turn up the Reverb Return control. There's two stages of recovery, typical Fender values 100k/1k5/25u except the second stage cathode is not bypassed. Just a .0047uF cap into a 100k-A Reverb 'Volume' control between the stages. .01uF cap into a series 220k 'mixer' resistor on the second stage (mixes with the dry 220k 'mixer'). 220k input/grid load at Reverb return jack. Zero hum when Return pot is on 0, hum increases with pot. Tried tube swaps.

[Enzo]

Move the pan as far from the chassis as possible. Grab a pair of longer RCA cables if necessary. Does the hum diminish?

Always mount the pan with its OUTPUT end as far from the power transformer as possible. And open side facing away. SOme people put a steel sheet over the open side as a shield.


And one tiny detail. THIS ABSOLUTELY DOESN'T MATTER, SO DON'T BOTHER TO CHANGE IT. Your primary wires are reversed. Red goes to B+ and blue to plate, if you want to be official.

[MattT]

Thanks Enzo. The hum is not pan related I don't think. It hums the same with no pan at all. FWIW, I do have the pan laying at the preamp end of the chassis.

Regarding the tiny detail on the Reverb OT: I had wondered about this but I read somewhere that the phase of the reverb signal doesn't really matter. I just hooked it up the way I thought would be correct. Also, this reverb circuit is different than a regular BF Fender combo reverb. I'm not sure if that affects this or not.

Below is a schematic of the preamp and reverb. This build was an experiment...I had tweeked the 6G6-B Blonde Bassman Bass channel circuit in another amp and then tried to shoehorn the reverb into it in this one. It was a case of: I want to try to build upon this circuit...I have an extra, unused reverb pan and tranny...I have this old homebrew that I want to rebuild...I've never built/experimented with reverb...so I just cobbled it all together. I even used resistors/caps from the 'leftover' box. The Deep switch is so subtle it essentially doesn't work. I'm thinking of redoing the Deep switch to something similar to the Mid-Boost except affecting the Mid cap instead of the Treble cap. The Mid Boost works great. The Bright switch is not really needed...it's plenty sparkley without them.

I also used a SF Fender inspired PI: 12AT7, 47k plates, 470R cathode, 22k tail, 470k grid loads, into 2 each cathode-biased EL34/6L6 (with 180k grid loads...probably could have gone lower with not much loss you think?). No NFB, low plate voltage (~395vdc), Matchless-style crossline MV and Vox-style Cut. It's sort of a strange combination but it actually sounds really sweet...clean and cranked. It's sort of a cross between a BF Fender and Vox AC30 tonally. Only ~21W clean and 30W distorted...just a tiny bit more than my 4xEL84 Rocket (which only runs @ about 305vdc plates).

[Enzo]

The phase doesn't matter in a reverb. The color codes on the transformers are to denote their function, that's all. it was just a little side note. For example in a push pull output tranformer there will also be red and blue wires, plus a brown one.