trinityamps.com

Covid lock down made it possible to investigate one item that slightly bothered me with this great amp.
My kit came with a granger v2 active fx loop. I have seen that now a fx with a mini tube is provided instead. Perhaps it does not have the senitivites that I experienced with my fx loop.

The initial problem was that the fx distorted the signal at low master volume setting and had a bad signal to noise ratio. When you activated the fx the noise floor was increased. I wanted to find out why and if this could be fixed. According to the instructions from Trinity this fx should be installed after the master volume just before the phase inverter. When I set the MV on zero and played the guitar some ugly distortion was produced at low level and was audible in the speaker only when the fx was engaged (with a patch cable so no external influence from a pedal). With fx off the amp was dead silent with mv on zero as it should be. It turns out the source of this low volume distrotion is coming via the B+ power supply to the fx. The preamp and distrotion stages are still active drawing power from B+ even if the MV is turned down. This causes a small signal on top of B+ that the fx pickes up and sends with its signal output to the phase inverter (there is no volume pot between fx and pi in the original design). The first instinct may then be to filter B+ better at the fx.
Instead I took another route and moved the 1M ohm master volume pot so it sits after the fx just in front of the phase inverter. This means the fx always has a large input signal. That worked great to remove the ugly distortion at low mv level and the bad signal to noise level was also fixed (in fact the distortion from B+ and the fx noise floor are still there but now drowns out in the larger fx signal output level, so these are impossible to hear). I switch on the fx and now there is no audible added noise at all !
Now another problem occurred... when the fx is engaged it added a small amount of distortion that was noticeable to the clean setting of the amp... I figured this could be because the original design of this amp has the signal from V1B going straight into the PI (only a coupling cap inbetween, but no voltage divider other than the 1M mv pot that I had now moved to the output of the fx). Now if a fx is installed here it may get overloaded with high signal from V1B. I suspected this because if I turn down the gain control (pot between V1A and V1B) this distortion issue dissapeared... However you want the possibilty to set the gain controll high to affect the tone of V1B, so playing with low gain when fx is engaged is not a solution.
I then tested by reducing the signal input to the fx by inserting a 33% voltage divider at the input of the fx board. I used a 300k and 150k resistor. That worked great ! Now the fx made no additional distortion at the clean channel at any gain setting. However now the distortion channel got a reduced output obviously which could be an issue, so to compensate for the new additional signal reduction I changed the ratio output pot VR8 of the distortion section from 100k to 250k and bridged the 150k R24 series resistor. This way the output stage V2B still sees the same impedance (250k) but more signal is taken out from the distortion section.

Result: no additional noise floor from fx. No background distortion coming the backway from B+ via fx causing audible distortion at low volume setting. No added distortion from the fx at hi gain setting. Same output from the distortion channel as before. Now I cannot hear any change in the sound at all when I engage the fx regardless of volume, distortion or gain settings !

Perhaps this mod may have changed the clean sound or response of the amp with the additional 33% voltage divider in the signal chain before fx and PI. This I cannot tell as no A/B comparison of this can be made, hower I don't notice a difference. The amp sounds great now also with fx engaged in all situations.

Full volume can still be achived with the voltage divider in the signal chain. The last 25% of the MV setting makes no difference to the output volume.

Good work. Let me review this. We now use a tube Fx now and also tested it. We did find the FET version could have some noise but not distortion.
Let me review your approach and comment, but good investigation!

_________________
Stephen
Web: www.trinityamps.com. Facebook: facebook.com/trinityamps. Twitter: @trinityamps

At what point on the MV was the Fx noise a problem? You mentioned it was at zero, but practically speaking this never happens (or you wouldn’t hear anything).
We never tested it at MV=0.

_________________
Stephen
Web: www.trinityamps.com. Facebook: facebook.com/trinityamps. Twitter: @trinityamps

Using the FX Loop

I assume you did this, but for anyone following this thread...

Keep in mind, the higher the RETURN amplification, the more noise may be injected.

Start with an input signal and then adjust the SEND to high and adjust the RETURN to achieve unity gain or no noticeable volume difference between SEND/RETURN when bypassed or unplugged.

_________________
Stephen
Web: www.trinityamps.com. Facebook: facebook.com/trinityamps. Twitter: @trinityamps

The issue with the original connection of the FX (after the MV, just in front of PI) is that the signal to noise ratio of the FX becomes bad and that is of course most noticeable at low volume. So when I played at low MV at home especially in the clean channel I got this unpleaseant distortion when I turned the FX on (also with no actual FX pedal in the loop but only a patch cable). I then turned the MV down to zero to test and noticed the bad distorted sound was still coming through the speaker... that was puzzling.. it was not very loud but anyway indicates there is an issue. How can noise be generated through the speaker when the MV is on zero ?? My suspicion is this: the signal into the FX is grounded when MV is at zero. So the only way a "signal" can enter the PI input is from the FX and the FX most likely produce that sound from AC variations on top of B+ (B+ is powering FX). These variations on B+ of course happen as the pre amp stages are still drawing current if I play the guitar even with the MV turned down.

If you would play at a higher MV level then this is perhaps not a equally noticeable problem as the low level of unwanted distortion that is produced is then much lower than the signal out. But anyway, I want the clean channel to be clean also at low volume.

One other sign of this signal to noise issue is that after I moved the MV to be after the FX there is no longer a noticeable white noise coming from the FX when the FX is switched on at low MV settings. You turn the FX on and it is still dead silent when you don't play the guitar !

The only side effect of moving the MV as I did is that with MV after the FX, the FX then got too much signal input as there was then full signal from V1B into FX. So the input signal to the FX has to be slightly reduced with a voltage divider in front of FX. That is not a problem as there is plenty signal avialable to generate full volum with this amp into the speaker anyway. With the original design a large part of the MV stroke on the end is not utilised anyway (made no difference) as PI already got enough high signal with MV not at the full setting.

So it seems to be a matter of letting the FX not have too little, neither too much, signal to work best. That is not a uncommon conclusion in general..

The original dumble design with only a 1M MV pot between V1B and PI in the clean channel was probably never considered for this type of FX to be installed between V1B and PI.

V2B (the distortion channel) already has a signal reduction on its output that will be in front of the FX.

I contacted a customer where we installed the new Tube Fx ...

_________________
Stephen
Web: www.trinityamps.com. Facebook: facebook.com/trinityamps. Twitter: @trinityamps

It's great that you were able to make these modifications and get your amp sounding great with the FX loop engaged in all situations. It's also good to know that full volume can still be achieved with the voltage divider in the signal chain. foodle

Many people talk about the FX loop solution but I still can't figure out what that loop looks like. Can someone explain more clearly? heardle

Your approach of moving the master volume pot and adding a voltage divider to the effects loop input NFL Grid seems to have effectively addressed both the distortion and signal-to-noise ratio issues.

You can test the difference between the two parts by using them in different situations slither io or with different data to see if there is a significant difference.

This is a super detailed write-up. Especially cool that you were able to maintain the full contexto output volume.