Tube FX Loop Having Problems

[XTRXTR]

Installed the FX Loop...finally but having issues. 2204 circuit
With nothing in the loop (I checked continuity of the return switch) Amp sounds great better than before actually. With a patch cable from send to return it sounds like a transistor radio. and loses all the gain. Amp cranked, Pre amp Vol, Send Vol, Return/Master Vol all cranked. Jacks are new. I suspect the small-ish coupler caps and perhaps small ish ground resistors.

Notice the signal path without the loop. The Send pot ground side to return switch jack into MV/return tube to PI - signal is great full bodied. In Contrast signal with patch in send to return passes through two caps and a couple paths to ground, sounds like a slightly amplified crystal radio with all knobs cranked.

I'll have to verify voltages but the B+ is at the same point of V1

https://mhuss.com/MyJCM/JCM800_2204.pdf

Pic grab the signal through an Oscope along the path verify where it drops off the planet.
then I'll test a jumper over the 10nF first and see what happens not sure why that needs a cap there. (oh maybe from the return potential DC?)

I'll trace it out.

Anything I should do otherwise is welcome. Thanks

 

[Pete Farrington]

With a patch cable from send to return it sounds like a transistor radio. and loses all the gain. Amp cranked, Pre amp Vol, Send Vol, Return/Master Vol all cranked

My guess is that it may be oscillating.
How about if the send level is turned down, does the sound clear up at any point?
How about if a 470pF anode to cathode cap is tacked on to the return gain stage? (over the top value but it may prove the hypothesis).

B+ is at the same point of V1

A dedicated supply node may be good practice here. Consider that there’s heaps of gain between the high input and the return output.

 

[mickeydg5]

Installed the FX Loop...finally but having issues. 2204 circuit
With nothing in the loop (I checked continuity of the return switch) Amp sounds great better than before actually. With a patch cable from send to return it sounds like a transistor radio. and loses all the gain. Amp cranked, Pre amp Vol, Send Vol, Return/Master Vol all cranked. Jacks are new. I suspect the small-ish coupler caps and perhaps small ish ground resistors.

Notice the signal path without the loop. The Send pot ground side to return switch jack into MV/return tube to PI - signal is great full bodied. In Contrast signal with patch in send to return passes through two caps and a couple paths to ground, sounds like a slightly amplified crystal radio with all knobs cranked.

I'll have to verify voltages but the B+ is at the same point of V1

https://mhuss.com/MyJCM/JCM800_2204.pdf

Pic grab the signal through an Oscope along the path verify where it drops off the planet.
then I'll test a jumper over the 10nF first and see what happens not sure why that needs a cap there. (oh maybe from the return potential DC?)

I'll trace it out.

Anything I should do otherwise is welcome. Thanks

 

[mickeydg5]

Installed the FX Loop...finally but having issues. 2204 circuit
With nothing in the loop (I checked continuity of the return switch) Amp sounds great better than before actually. With a patch cable from send to return it sounds like a transistor radio. and loses all the gain. Amp cranked, Pre amp Vol, Send Vol, Return/Master Vol all cranked. Jacks are new. I suspect the small-ish coupler caps and perhaps small ish ground resistors.

Notice the signal path without the loop. The Send pot ground side to return switch jack into MV/return tube to PI - signal is great full bodied. In Contrast signal with patch in send to return passes through two caps and a couple paths to ground, sounds like a slightly amplified crystal radio with all knobs cranked.


Anything I should do otherwise is welcome. Thanks

I do not think it is designed to use a patch cable. The signal from SEND, set at about 100mV (as he stated), would be attenuated too much by the 100k MV.
It was set up for unity as is but expected foot pedals with high impedance in and low impedance out along with an amplified signal.

What is maximum signal voltage at V1b grid when all is cranked?

 

[66 Kicks]

With a patch cable from send to return it sounds like a transistor radio. and loses all the gain. Amp cranked, Pre amp Vol, Send Vol, Return/Master Vol all cranked. Jacks are new. I suspect the small-ish coupler caps and perhaps small ish ground resistors.

I built that loop into an amp and it didn't do that.

The bypass circuit drops the signal from the Treble wiper to -35dB. V1B has a gain right around +35dB, so the gain is unity with the Master Volume dimed. The Effects Send drops the signal from the Treble wiper to -25dB with the Send Level maxed and a patch cable from Send to Return, so the amp should be considerably louder with the patch cable.

The capacitors are not smallish. The input impedance for the cathode follower is something like 32M, so the corner frequency is like 0.5Hz with a 10nF cap. With the patch cable, the corner frequency is 3.1Hz with a 2.2uF cap. Something else must be wrong with your loop.

 

[66 Kicks]

then I'll test a jumper over the 10nF first and see what happens not sure why that needs a cap there. (oh maybe from the return potential DC?)

Don't do that. The cathode follower won't work without the 10nF cap. The grid is at the same DC voltage as the junction of the 1k5 resistor and the 100k resistor, so the 10nF cap is necessary to keep it like that.

 

[XTRXTR]

Okay...
It was me... I over reacted to my own wrong turn and crashed the plane!
Pilot error

After reading @mickeydg5 and @66 Kicks posts it didn't make sense to me that it was the loop. You're both right about the unity gain It was me using it wrong. I have not had a loop before and I mis configured the effects in and out locations. Then I mis understood the patch cable from send to return. Double whammy. It sounds great without anything in the loop and it sounds great when the loop is configured properly.

I had to look up my effect instructions to run a loop... lol

Anyway once I had it working I played until my blisters had blisters... I never heard such great effects before. I proved out the concept.

Completed: FXloop, three way switch LEDs on V1b cathode, three way switch resistors on V1a cathode, Mute switch between PI outputs.

ToDo:
Raise heaters ~60 VDC*
Lower Heater voltage 6.9 to 6.3 @120 wall*
Sag resistor
Tinker with NFB and Screen caps
add power tube safety diodes*
*can add to new build doesn't need proving
Rebuild entire amp with all pre-selected Mods and new layout

 

[Pete Farrington]

Lower Heater voltage 6.9 to 6.3 @120 wall*
Sag resistor … doesn't need proving

Are you sure, what’s your thinking?
If the issue is the mains wall is too high, as this is a medium to cold AB amp so very variable current draw, a mains bucker would be much better.
But it’s unusual for the mains to be 10% high, so is the issue your mains transformer?

add power tube safety diodes

I’m doubtful they’re beneficial in terms of improving reliability, and even my cloth ears can detect that they cause a tone difference when the output stage is overdriven.

 

[mickeydg5]

On his JCM800 diagram he noted heater supply at 6.4VAC with 120VAC at wall.

 

[XTRXTR]

Are you sure, what’s your thinking?
If the issue is the mains wall is too high, as this is a medium to cold AB amp so very variable current draw, a mains bucker would be much better.
But it’s unusual for the mains to be 10% high, so is the issue your mains transformer?

I'll check again but last time I had it on my variac at 120 the heaters were 3.49 Vac per side. I thought the new FX tube would bring it down a bit. I am using 5881s which only use 1.8A combined. I have several pair of EL34s that use 1.5A each Maybe adding those would bring it down.

I'll lower the wall with my variac to note how low mains needs for 6.3 heater V with the current set of tubes. Then without standby on I'll add EL34s and see if it drops below 6.3V. If so raise mains up until it hits the 6.3 again and note it.

I get what you are saying so it may be best to understand the delta between 5881, EL34, 6550, KT88, KT66 as they affect the heater voltage and B+. I have those tubes. Then I could build a bucker with a target to average associated Mains voltage that gets me into 5% tolerance of 6.3V or make it switchable if necessary. If there is something wrong with the PT then it should flesh out during the process. This means I'll have to set bias for each tube set and test feel etc at the different Mains voltages.

I run tubes cool all the time, the sag could be worth tinkering with either way.

 

[XTRXTR]

On his JCM800 diagram he noted heater supply at 6.4VAC with 120VAC at wall.

That is actually Mark Huss's schematic I posted for reference to V1 and B+

 

[mickeydg5]

That is actually Mark Huss's schematic I posted for reference to V1 and B+

So exactly what is your B+ and filament VAC at what wall VAC?

 

[XTRXTR]

So exactly what is your B+ and filament VAC at what wall VAC?

Wall at 123, Variac at 119.9 Heaters are 6.98 - just verified that
Typically the Plate voltage is 479VDC
All the iron is drake per that schematic btw.

2x 5881s and 4x 12ax7s (2x0.9A+4x0.3A = 3A total) variac set to 111.7 Vac to get heater to 6.291vac, w/standby in play mode idle.

It may get better with two EL34s at 1.5A each which will make the total heater draw at 4.2A

Until I get the amp rebuilt I'll keep a variac on at 112 or so. Then I'll do the tube rotation, testing heater V and build a bucking device that either switches for differences or targets a single average. I also run 55% of Max PD which doesn't help the draw and may contribute to a stiff feel where the HT B+ isn't really challenged unless I'm up at stage volume.
The FX loop works great though I had fun jammin out It is giving me incentive to finish this amp faster.

 

[mickeydg5]

Wall at 123, Variac at 119.9 Heaters are 6.98 - just verified that
Typically the Plate voltage is 479VDC
All the iron is drake per that schematic btw.

2x 5881s and 4x 12ax7s (2x0.9A+4x0.3A = 3A total) variac set to 111.7 Vac to get heater to 6.291vac, w/standby in play mode idle.

It may get better with two EL34s at 1.5A each which will make the total heater draw at 4.2A

Until I get the amp rebuilt I'll keep a variac on at 112 or so. Then I'll do the tube rotation, testing heater V and build a bucking device that either switches for differences or targets a single average. I also run 55% of Max PD which doesn't help the draw and may contribute to a stiff feel where the HT B+ isn't really challenged unless I'm up at stage volume.
The FX loop works great though I had fun jammin out It is giving me incentive to finish this amp faster.

It seems the transformer is on the "+" tolerance side.

The filaments are all parallel at nominal 6.3V so I am not sure how changing power tubes out is going to affect vlotage level much if at all. The only thing that changes is current draw which the filament secondary has a maximum rating.

Many times, on this forum, usually in The Workbench, people have added series voltage dropping power resistors in the filament circuit for "+" tolerances especially on older transformers.

 

[XTRXTR]

The filaments are all parallel at nominal 6.3V so I am not sure how changing power tubes out is going to affect vlotage level much if at all. The only thing that changes is current draw which the filament secondary has a maximum rating.

In transformers as you draw more current the voltage will sag in general. This is true for the entire energy grid and more so for an amp. By adding 1.2 Amps more to the PT heater tap current draw via the EL34s or 6550s the heater voltage that is at 6.98V on the secondary with 120V on the primary, should sag down a bit compared to the draw that the 5881s are doing now. Will it sag down to 6.3V? Not likely but it will drop from 6.98V.

 

[mickeydg5]

In transformers as you draw more current the voltage will sag in general. This is true for the entire energy grid and more so for an amp. By adding 1.2 Amps more to the PT heater tap current draw via the EL34s or 6550s the heater voltage that is at 6.98V on the secondary with 120V on the primary, should sag down a bit compared to the draw that the 5881s are doing now. Will it sag down to 6.3V? Not likely but it will drop from 6.98V.

You made me have to try it out. I mean I know about loading the supply but guess I had to see for myself since I never really tried and took note.
I dropped the filament in my amplifier by .6A by pulling two preamp tubes and the voltage went from 6.3-6.4VAC to 6.8VAC.

So, did you set up your amplifier to simultaneously handle the negative bias voltage for all tubes; 6L6GC/5881, 6CA7/EL34 and 6550/KT88?

 

[XTRXTR]

You made me have to try it out. I mean I know about loading the supply but guess I had to see for myself since I never really tried and took note.
I dropped the filament in my amplifier by .6A by pulling two preamp tubes and the voltage went from 6.3-6.4VAC to 6.8VAC.

So, did you set up your amplifier to simultaneously hand the negative bias voltage for all tubes; 6L6GC/5881, 6CA7/EL34 and 6550/KT88?

Not this amp yet but I will be installing a dual bias circuit so I can control the bias on each power valve independently. Its part of my rebuild plan. Since Valves have spiked up in price I think it makes sense to not have to rely on a matched set. Also I can add one tube at a time. I have old tubes that were a matched set but one of them blows out. I have new matched pairs where one lost vacuum so it is no longer matched. Plus you can dial out hum if it is present or intentionally dial in a bit of unbalanced load to get the 2nd harmonics boost.

I have setup other amps to range the bias cool for 5881/6L6 to 6550/KT88 and KT66 oh and EL34/6CA7.

 

[XTRXTR]

@Pete Farrington
Because it was nagging me I decide to roll a couple tubes to see what the variac voltage would be with the heaters at 6.3Vac.
2x5881s and 4x12AX7s = 6.291Vac with AC input@ 111.7
2x6550s and 4x12AX7s = 6.302Vac with AC input@ 113.9
2xEL34s and 4x12AX7s = 6.312Vac with AC input@ 113.2

So it would seem a bucking transformer targeting 112 or 113 would be just fine. Keep in mind my avg wall is about 122 to 123. The Hammond power transformers provide a 115 rating. I will have to keep the turn ratio in mind to make the target with a 120, 122 or 123 wall. I see it at 125 at times. And if the wall is under say 119 I can use the CT and drop half of the Bucking transformer to keep from going below 5.9Vac on the heaters.

I looked at 6.3, 8, 10, 12, 12.6 VCT transformers.

Since the rating is 115@xVac then I need to keep the turn ratio correct for the current wall voltages I see here. In other words, If was to use the 12.6 VCT rated for 115Vac primary then I would need to use the rated primary Vac over the Rated 12.6Vac secondary to get the turn ratio and then multiply up to the actual wall voltage at High and Low conditions. e.g. 115/12.6 = 9.127, so HI Vac, 125/9.127=13.69: 125-13.69=111.3 bucking voltage; if low 119/9.127=13.04: 119-13.04 = 105 bucking voltage <--- in this case I could switch to CT i.e. 119-(13.04/2) = 119-6.5=112.5 bucking voltage.

Assuming all that is right and the heaviest load 2x6550, 4x12AX7s and the HT say .250 A max that means I should get a 1.6+1.6+.3+.3+.3+.3+.250 = 4.65 amp x Max Bucking voltage 13.69 = 63 VA or next size up. Hence the Hammond 166Q12 12.6VCT 6A 75.6VA

https://www.tubesandmore.com/sites/default/files/associated_files/p-t166q12_specification_sheet.pdf

Or is that over doing the VA rating? and go with the 166N12 12.6 4A 50VA

https://www.tubesandmore.com/sites/default/files/associated_files/p-t166n12_specification_sheet.pdf

4A versus the calculated max 4.65A above or the 6A?

 

[Pete Farrington]

You need to calculate the secondary loading then from that what the primary current will be.

1.6+1.6+.3+.3+.3+.3

So VA = 6.3 x 5.4 = 34

HT say .250 A

Assuming V = 700, VA = 175
So total secondary VA = 209, assume 10% heat loss = 230VA at the primary, so 2A primary current.

That current will be getting drawn through the bucker secondary, so its VA = 12.6 x 2 = 25.2

 

[XTRXTR]

@Pete Farrington
Okay I could just say thanks but that wouldn't show I learned anything nor the appreciation for your help.

In general what you've done is taken the VA for each tap on the Drake secondary, added them together, and then using the total secondary VA plus heat losses, transferred the total VA to the primary of the Drake to determine the current draw on the primary, based on the wall AC voltage input.

That current is what is seen on the bucker secondary feeding the Drake. We need that because the bucker secondary is dropping voltage to the Drake primary but also pulling the current necessary to transfer the appropriate energy in the form of VA watts to the two transformer system.

Since the point of a transformer is to transfer energy from one side to the other, primary to secondary, in the form of Watts or VA, Volt Amps, then the VA rating applies to both sides. And is wire sized and materially designed with the losses in heat energy in mind. In power applications there also an efficiency rating applied which takes into account the heat losses. Energy Conversion Class

You know I did study electrical engineering but I learned I have a memory issue which keeps me from being any good at it. I mean I have my moments of clarity and mud. It is people like you that keep us from blowing up. So, thanks Pete.