Fender reverb clone - Distorted reverb

[Cornelius]

I've just built a Fender Reverb (6G15) from scratch with whatever i had lying around, except for the tank and reverb-transformer.

To my big surprise, it worked at first power up! But the reverb sound itself are distorted. I've used all the standard values from the schematic, except for the dwell pot; i only had a 500k. (schematic says 250k.)

It sounds as the distortion happens before going to the reverb-tank; if i have the dwell pot on 1, the reverb are clean, but anything above, it's fuzzy; not so much the fading reverb though. Turning on the bypass switch, the sound is clean (going just through the cathode-follower stage...)

Also, for some reason i didn't quite get the right tank i think. the transformer has an output impendance of 8ohm, and the tank's input are 10ohm.

I've checked all the voltages on all the tubes, and they are close to original schematic..

[rfgordon]

To help isolate the problem, you can connect a speaker to the RCA jack that sends to the tank. If that tone is OK at that point, the issue may be later on. Make sure the tank is connected right way round. Also a 10 ohm tank is fine, as long as the output impedance is at least 2k ohms.

[Cornelius]

I thought about that, but wasn't sure it was safe.

With a speaker connected to the 'out-to-tank rca', the sound was exactly as i heard it through the reverb; clean at 1 on the dwell pot, and distorted when turned up. Though, when playing lightly on the strings, there is no distortion.

The distortion are a hard-clipping, underlying sound...

I inspected my wiring more closely, and i see that i have swapped the in and out wires on the dwell pot (in from v1a on center, and out on edge); i'll try to swap those later...

Edit:
Yes, output from the tank are at 2,2k...

[Cornelius]

I swapped the wires on the dwell pot; no change...

I made a probe going to my amp, and started probing at the input and output stages while playing my guitar with my feets...

Input at v1a: ok.
Output at v1a (after cap): ok.
Input at v1b: ok.
Output at v1b (after cap): terrible.

So it seems that it's the v1b stage that distorts. The sound level difference at the input and output of v1a wasn't great; only slightly amplified. (I realize that inserting an audio output into the circuit might alter the impendance etc., but i think this test was valid enough...)

V1 are a 12AT7WA; i tried with a 12AU7, and the distortion are present with that one too, just at lower volumes.

Could the dwell pot at 500k, which should be a 250k do this? The circuit around v1b is simplicity itself, so it's not much that can go wrong there...

Now, this build is a terrible birds-nest, so i'll check the wirings again...

[rfgordon]

Double check all the connections at the tube socket. If the connection to the dwell pot, or the dwell to ground, is not good, you won't have a proper grid load. As signal increases, blocking distortion occurs at the input grid. It's one of the ickiest sounding distortions of all.

Also, you might take a tip from the Dumble forum, and, at the connection of the plate voltage side of the verb transformer connect a .1 mF/630V cap to ground. This increases gain and clarity. Do this after you've fixed the other problems.

[Cornelius]

Double check all the connections at the tube socket. If the connection to the dwell pot, or the dwell to ground, is not good, you won't have a proper grid load. As signal increases, blocking distortion occurs at the input grid. It's one of the ickiest sounding distortions of all.

I've gone over all the solderings, and there is no change. I'm not bad at soldering, but it didn't hurt going over the points, since all components, including the sockets are previously used...

After some more probing, i've concluded that the distortion are a clean preamp-overdrive sound.

Now, i did som more voltage measurements, and as i said earlier; the voltages are around the schematics volts... With some variances...

I'm running a bit lower voltages all around. The schematics says 250v to the preamp tubes; i'm running 238v.

Plate voltage at v1a: 115v
Cathode at v1a: 1,66v
Plate voltage at v1b: 115v
Cathode at v1b: 1,85v

Schematics volts are 105v, 1,8v; 120v, 2,0v.

What confuses me, are the plate volt at v1a. There is a voltage divider going to the plate, with a 100k and a 10k, where the signal are tapped between them. The volt before the 10k are 238v, after the 10k, before the 100k: 226v. Why isn't plate voltage lower for v1a? Measuring the 10k and 100k resistors; they are within 1%.

I start to wonder if not all fender reverbs have a distorted 2'nd stage preamp...

[rfgordon]

The plate voltages for the original are uber-low. You're gonna have near zero headroom on this. It may also put DC on the grid, so if you plug a guitar into this and find the volume pot sounds scratchy, that's what's happening. To prevent that, put a cap between the input jack and the grid of V1a. Low plate voltages also make those gain stages sound darker.

All in all, it'll sound better with plate voltages around 200VDC, but the PT you're using may not give you that, I don't know.

[Cornelius]

A thought:
I'm using some old 1/2W carbon resistors (from the '60) at the plates; the tubular kind...

They measures ok, but could these be the guilty part in my distortion problem?

[rfgordon]

The type of resistor (CC, CF, MF) will change the tone, but not create or eliminate distortion. Low plate voltage will allow a gain stage to distort with relatively low input. Headroom is proportional to plate voltage, in both preamp and power tubes.

Remember, those old Fender verb units were intended for use with low output Fender guitars (pickups measuring less than 7k ohms).

[Cornelius]

Ah, ok.
I just thought i read somewhere that old carbon resistors could vary its value with temp and voltage, and besides being noisy could add a distorted sound....

But plate voltage....
I'll try to raise it; i have 335V available after the first r-c net in the power supply.

[Cornelius]

I'm now running 174v on v1a, and 186v on v1b plates, and 1,3v/1,5v at the cathodes; it got a bit better. I can now turn the dwell control to 2 without distortion, and with the mix control all over to the reverb, there is so much reverb (dwell) that 1/4 of it would be too much...

I think i have to revise my questions here...

Now there is 2 pre-amp stages (V1 (12ax7)), and a 'power'stage (6v6gt). The dwell control goes between v1a and v1b. By probing, i find that it is v1b that distorts.

Could i skip v1b entirely, since i find there is too much reverb at 2 on the dwell pot? (Thus eliminating the distorted stage.)

I'm wondering if the 1 tube reverb design found here would be just as good. (I'm thinking of integrating the reverb into my little 5F2A clone (with el84 at the output)...

Thanks for all the help so far.