Triple Rectifier 2 channel-Clean Channel issues (distortion)

The Boogie Board

Help Support The Boogie Board:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

MasterOtono

Member
Joined
Sep 19, 2015
Messages
9
Reaction score
0
My friend asked me to help with his Triple Rectifier 150 W dual channel Amp.
recently he had an issue with a V3 tube.
Symptoms - volume drop on all channels. After replacing - am worked fine.
One month later he had another issue with a send-return loop switching.
Repaired - One dead optocoupler and one dead transistor in switching circuit.
But now there is another issue, clean channel distorting at peaks.
All knobs at noon, Orange gain at about 5-10%
When i'm playing slightly - it sounds like a regular clean channel, when i make hard picking or ply chords- it sounds distorted even at lower gain settings.
I connected a cable to a "Send" jack of this rectifier and connected it to a "Return" jack of another amp.
It sounds exactly the same, with distortion.
When i connected an external preamp into a "Return" jack - it sounded good without any artifacts.
So the power tube issue can be excluded.
Channel Style select switch is in the middle position, Orange channel gain is set to "Clean Rhytm"
All tubes have been changed to a new ones, i tried two different sets of 12AX7 tubes.
Heater voltages are OK - 6.2V
Then i tried to connect my sound-test probe to the each stage and to check the sound From V4 to V1 stage by stage. I found that the issue appears at the V3 stage.
After the coupling capacitor of the V2b stage the signal is fine, but on the V3a anode sound is being distorted.
I tested all optocouplers and they are looking pretty good.
about 100 Ohm open ant more than 2Meg closed.
Then I measured signal with Generator and Scope.
I applied 150mV peak-to-peak voltage at the amp input and checked voltages and gain stage-by stage.
Signal at he LDR8 and LDR9 is allredy distorted, I disconnected theese LDRs and it doesn't help - so the tone stack issues can be excluded.
With gain setting at about 50% I measured about 9V p-p signal at he V3a Grid, which is really big value, considering the -1.6V Grid Bias.
That means that this stage can accept maximum 3.2V p-p signal without distortion.
When the signal level is higher, Grid to Cathode voltage goes to positive value and the Grid - to - Cathode diode begins to distort sound.
In this amp input signal at the V3a stage is 3 times higher than this stage headroom.
My thought was that something went wrong in previous stages and it caused overall gain increase.
Maybe a shorted optocoupler could add the cathode bypassing capacitor to the one of these stages, which could cause gain increase.
I started to calculate and measure each stages gain.
and got the following results:

qXK-n4o8rdc.jpg

IMNE1b-R7qE.jpg

yoidG_UUgyU.jpg

With 15-20% of gain setting I can see clean sine wave at the V3b Cathode,
With 50% of gain i have distorted sound after the V3a stage.
Calculated and measured gains have really close values, excepting the V3a and V3b stages.

My questions are:
1) Is it ok that with 150mV input signal and 50% Gain setting - we are geting these 9V at the V3a Input?

It looks like a bottleneck, but I heard Mesa DR clean sound and it was pretty fine without clipping till the 60-70% of gain.
Either the gain in V1a -V2a - V2a stages still is higher than it should be or V3 stage gone mad.

Could you please advice?
 
Check LDR 10. It should be open for clean. I think it will distort the stage when closed. It changes the internal resistance of the tube for DC and reduces gain.

As far as the voltage in to V3a: From V2b to V3a, the grid resistor is after the load resistor. So the first half of the potential divider will be formed by the resistance of the anode, (Ra + ra) = R1. If the Load is R2, then:

Voltage to grid resistor = (R2/(R1+R2))*V2b-out (I'm sure you know this math, but I don't like to assume).

Any voltage on the plate will make a voltage swing larger, but it will cut off and clip when it hits the edges of the predetermined headroom. I don't think the grid is the issue for the gain increase, since the cathode resistance sets the quiescent bias point for the gain conditions set by the anode resistance.
 
I checked all LDRs separately.
Also i tried to run V3a without 1uF capacitor (simply desoldered it to prevent any impact on gain)
LDR3a-b works well. when I shorting its leads with a screwdriver - I can hear a Gain increase.

Internal Plate resistance is about 62K in parallel with anode resistance it will be about 38K.
So (330K/(38K+330K))*9V =0.89*9V = 8.0V
The voltage drop is about 10% which doesn't help in this particular case (

I meant that the grid Bias is setting the input headroom.
If we have bias at -1.6V headroom will be 0-->1.6V-->3.2V
If we will apply 5V p-p input signal it will vary from -2.5...0...+2.5V
When the positive period of the input signal will be in 1.6 ... 2.5V range voltage between grid and cathode will be positive too. Electrons will start flow from grid to cathode and it will act as a simple Diode.

Previous stage will be loaded to a Grid Cathode resistance of about 1KOhm, so we will get a voltage divider.

I still can't understand - how ???? this clean channel has a clean sound in a working amp, when the output stage V3a has headroom 3.2V and the input signal is much higher.
All calculations look very close to the measured values on a broken amp.
 
Sorry for my English.
I'm not a native English speaker.

I desoldered V1 V2 V3 sockets and cleaned them in solvent.
Also I cleaned all tracers and solder joints with solvent.
After that I dried PCB and sockets for about 2 hours.
Then I soldered all sockets back without using of additional Flux - to prevent any additional conductance.

There is a bottleneck with 12AX7EH, datasheet says the following:
Heater to cathode Voltage
Positive - Not More than 100V
Negative - Not Less than 200V

Considering that heater is tied to the ground using two 100Ohm resistors
and Cathode voltage is 198V that means that it is Positive Heater to Cathode voltage 198V
Am I right?

I also have a pair of the JJ ECC83S
their datasheet says the following: Uk/f = 180 V

And a pair or Sylvania ECC83
Heater Negative With Respect to cathode:
Total DC and Peak..................................200V Max
Heater Positive with respect to cathode:
DC.......................................................100V Max
Total DC and Peak..................................200V Max


So I will try to change the V3 Tube.

Also I found a thread, where guys discussed about using EH in CF circuits:
http://forum.grailtone.com/viewtopic.php?f=24&t=48810
It seems that it is not Ok to use them in CF

P.S. Thank you for the interesting link regarding Soldano Schematics.

Also there is an interesting thing:
Without the cathode follower there is severe clipping as the driving triode's grid voltage approaches zero and more gradual clipping as the triode transitions into cutoff. The directly coupled cathode follower evens out the score. Soldano's fourth stage thus produces a more symmetrical output waveform. Second-order harmonics are reduced. Third-order harmonics are enhanced

If this stage will work correctly - it will clip more symmetrical.
Maybe it will sound more natural and without much overdrive.
 
I'm back.
I tried two more tubes to see if it helps to resolve this problem.
Measurements:
sfVesZBFCCo.jpg

All 3 tubes are working with hard enough clipping.
Sylvania distorts hardly than others.
The other strange thing that for all 3 tubes Ug2 > Uk2 about 4...10V Difference
But in the original schematic Ug2<Uk2 and difference is 3V

Also on the full gain I can hear a squeal from the PCB side.
It is more significant at 3.5 KHz
I can hear it on all tubes but it is hard to understand where it is exactly on the PCB.
When I touch V3a Grid with a scope probe it is fades out a bit.

Gain -50%
Tube -JJ ECC83S
V3a Anode:
cIUKcXLYr10.jpg


V3b Cathode:
-o7wfoewyl8.jpg
 
Sweet. I'm glad it's gonna work out.

The squeal is likely from the cascading gain. Stage 3, with the anode cap and cold bias, is the tonic for any miller capacitance or otherwise harsh treble.

There is a bottleneck with 12AX7EH, datasheet says the following:
Heater to cathode Voltage
Positive - Not More than 100V
Negative - Not Less than 200V
Considering that heater is tied to the ground using two 100Ohm resistors
and Cathode voltage is 198V that means that it is Positive Heater to Cathode voltage 198V
Am I right?

-----
(edited) The voltage drop from the internal resistance of the heaters to the 100 ohm resistors will make the ground plane different from zero, I think. Since it's split to negative and positive in parallel, that small offset wouldn't make much of a difference. The resistors are to help current run to ground and prevent noise.

On the CF, the cathode is positive to the heaters. It basically would be 198 at rest.
------
From what I understand, the spiral filaments are fragile by comparison to some other tubes. They can't really handle CF's. For regular cathode biased stages, they're quiet and do their job well.

You are super organized and thorough. Your friend is lucky.
 
Hello,

Sorry for delay. I was ill and spent a week in my bed.
Now I'm back!

I excluded the V3a and V3b Stages from this amp. I soldered these stages on a separate tube socket and connected wires to a PCB.
(Ground, Power and Input)
This was made to exclude any impact from a PCB side.
Also I connected heater voltage from a separate power supply - to exclude Heater to Cathode voltage issue.
As a load for the Cathode follower I built a a separate tone stack.
qVdL6HZGlZE.jpg


I installed the LT-Spice software and modeled this preamp.
As I can see:

with 15% of Gain and 100mV input signal
uzg1lfyL4lw.jpg

(Green -V3a Anode; Blue - V3b Cathode)
voltage at the input ("X" point) is 6.35V
the anode AC Voltage on V3a is about 272V (Clean)
V3a cathode AC Voltage is 2.27V it swings from 0.6 to 2.8V


with 15% of Gain and 150mV input signal
LWqy62PssNM.jpg

(Blue -V3a Anode; Green - V3b Cathode)
voltage at the input ("X" point) is 9.94V
the anode AC Voltage on V3a is about 324V (distorted)
V3a cathode AC Voltage is 3.1V it varies from 0.0V to 3.1V

In real conditions I can see the following values:
Input - 100mV; Gain - 15%
Point "X" - 6V
V3a Cathode AC Voltage - 2.5V (slightly distorted)
V3a Anode AC Voltage - 225V (slightly distorted)

Input - 150mV; Gain - 15%
Point "X" - 10V
V3a Cathode AC Voltage - 2.75V (distorted)
V3a Anode AC Voltage - 230V (distorted)

All parts are new, Socket is new.
I tried about 6 different tubes with the same result.
I tried a new Ruby 12AX7-AC tube - no success.

So the V3a output is lower than the simulated one.
Real amp:
Plate DC voltage - 223V
Cathode - 1.7V
I got 230V p-p Max at the V3a Plate, but sym shows 320V p-p max. (distorted signal)
This is about 30% lower than the estimated value.
Such way it works with less headroom and begins to distort earlier.

I don't know why this happens(((((
 
I wish I could continue helping you, but I had an accident. Between the pain and the meds, my mind isn't very clear. If you continue to find problems, I encourage you to reach out to Mesa. They are brilliant.
 
afu said:
I wish I could continue helping you, but I had an accident. Between the pain and the meds, my mind isn't very clear. If you continue to find problems, I encourage you to reach out to Mesa. They are brilliant.
I wish you to get well and soon.

I already sent an e-mail to the Mesa support.
They asked me for a serial number to provide me a schematic.

I asked to forward my message to the technical stuff.
Still don't have a reply, but will wait some more days.
 
Update!

I got a reply from Mesa support.
They said something like:

We don't have time to answer on these technical questions.
First models of rectifiers have good drive and clean is not such good as drive.
If you want good clean - buy newer models of rectifier.

(this is what I understood from their message)

So the next step will be to compare this amp with other rectifier.
To see if other have the same issues with the clean tone.

I fell really disappointed, thought that it is not hard for Mesa Engineers to take a look on some voltage maps and some posts to help fix the issue or to say that this is not an issue at all...

I think that exists a troubleshooting guide or more detailed voltage map with amp test conditions.
In production conditions - they need to be able to test an amp with a scope and generator.
 
I'm sorry hear that you are disappointed with the response you got from Mesa - usually they are extremely helpful.

Perhaps there was some issue with translation?

Anyway, I think you may be expecting more from that old 2 channel triple than it can do - or was designed to do.

The early rectifiers were really not designed to have a great clean channel - they were simply designed to produce that brutal gain that set them apart at the time.

Later - when folks started to repeatedly ask for a better clean channel, Mesa answered the call with the newer multi-watt series.

Although I respect all of your efforts, I think to solve your issue you are going to have to buy a different amp.
 
Just to clarify, the actual gain pot value is around 250kohm instead of 1Meg. It doesn't make such a great difference, but might account for some oddities when comparing the simulation to the values measured from the amplifier.
 
Finally i finished this repair.
Hum has been caused by a missing contact with ground rail on one of the master potentiometers.
Gain pot on red channel is 500K. I left it in place to do not confuse amp's owner with different amount of gain.
Initial issue seems to be a EH tube in the V3 stage. When we tested it on the full volume - there were slightly distortions on a clean channel, but without hard fuzzy sounds.
All my actions:
1 - cleared and resoldered all tube sockets in the preamp section.
2 - replaced 3 cathode capacitors with new ones.
3 - Replaced V3 tube with a Ruby 12AX7
4 - Clean Gain replaced to it's original 250K Pot.

I think that actions 3 and 4 resolved this issue, but who knows.... )

Thank you for your help and advises!
 
Cool. I'm glad it worked out.

I like Ruby tubes, by the way. I use their 6L6GC tubes in my Recto and I'm thinking about a set of EL34 from them. I sense that their reputation is not quite as high as distributors like Groove Tubes or Tung Sol, but put in line with EHX or JJ, I don't believe there is any difference in quality control. I could be mistaken, but so far, so good in my experience.
 
Back
Top