Using Helix with high gain tube amp - huge noise floor

[nhoven]

Hi Phil,

 

2 minutes ago, phil_m said:

My Bogner Atma is dead quiet when connected via the 4CM... I honestly don't get any sort of extra noise, even on the highest gain channel. A lot of this comes down to how the loop in  the amp is designed.

 

not to be too bold, but that's where you're wrong - alot of it really comes down to how the amp's preamp works (see below). Read the above reply again - he had the exact same experience as I did:

 

24 minutes ago, krisjohnson22 said:

My signal chain at this stage doesn't get any simpler than effect into amp input. I tried all the different outputs, instrument vs line level, guitar connected or not connected, volume knob in and out of the equation, volume pedal/gate in an effects block turned all the way down, the noise stayed steady and didn't fluctuate one bit. It's obviously the noise floor from the DAC stemming from the signal level being set quite low relative to Full Signal in the DAC making the noise floor quite strong relative to the anemic instrument level signal. They've provided an extreme amount of headroom on their outputs that makes no sense to me except....

 

No effects loop involved. I had 100% the same issue - Send from the Helix to amp input, hissing and noisy as hell, and not even thinking about four cable method. And the noise is coming from the analog domain, as it cannot be influenced by any of the Helix's volume controls or gates.

 

I have no personal experience with the Bogner Atma (and I could not find a schematic), so I can't comment on that. But both my amps and the Soldano HR100 are designs with four gains stages (plus cathode follower), which is ALOT of signal amplification (technically: That's four stages in series with a µ of around 70 to 100). The samples I heard of the Atma (and the fact that it only has three ECC83 including the phase inverter) suggest to me that some (if not alot) of the Atma's gain might be coming from FET / diode clipping, which is fundamentally different from clipping through a gain stage in terms of noise amplification... so yeah. That might explain why some amps with high gain sounds are fine, and in some designs the Helix's output noise craps the hell out of the sound.

 

Me, I have given up on hoping to get the Helix to work with my amp, at least as far as using the third channel. Noise level is acceptable with the second (Marshall 2203-ish gain level), but sub-par with the third one.

 

Kind of dissapointing that the output noise of a flagship unit like this one is so bad... or was "overlooked". I would be curious if the analog output circuits of the HX FX has been redisigned, or if it is just as bad with certain preamp designs.

[phil_m]

30 minutes ago, hoven said:

Hi Phil,

 

 

not to be too bold, but that's where you're wrong - alot of it really comes down to how the amp's preamp works (see below). Read the above reply again - he had the exact same experience as I did:

 

 

No effects loop involved. I had 100% the same issue - Send from the Helix to amp input, hissing and noisy as hell, and not even thinking about four cable method. And the noise is coming from the analog domain, as it cannot be influenced by any of the Helix's volume controls or gates.

 

I have no personal experience with the Bogner Atma (and I could not find a schematic), so I can't comment on that. But both my amps and the Soldano HR100 are designs with four gains stages (plus cathode follower), which is ALOT of signal amplification (technically: That's four stages in series with a µ of around 70 to 100). The samples I heard of the Atma (and the fact that it only has three ECC83 including the phase inverter) suggest to me that some (if not alot) of the Atma's gain might be coming from FET / diode clipping, which is fundamentally different from clipping through a gain stage in terms of noise amplification... so yeah. That might explain why some amps with high gain sounds are fine, and in some designs the Helix's output noise craps the hell out of the sound.

 

Me, I have given up on hoping to get the Helix to work with my amp, at least as far as using the third channel. Noise level is acceptable with the second (Marshall 2203-ish gain level), but sub-par with the third one.

 

Kind of dissapointing that the output noise of a flagship unit like this one is so bad... or was "overlooked". I would be curious if the analog output circuits of the HX FX has been redisigned, or if it is just as bad with certain preamp designs.

 

If the inputs of certain preamps are noisy, wouldn't that be an issue with those preamps moreso than than the Helix? As far as the Helix's (and HX Effects) converters, they're are the same converters that Fractal uses on the AX8 and I believe the Axe FX II (don't know about the Axe FX III). The Helix has a noise floor, as any sort of processor does, but the noise floor is really low, generally (not saying something might not be going on in some context that's raising it). I'm not sure what's causing the noise you guys are having issue with, but I'm a little skeptical about blaming it all on the Helix when it could be because of some sort of interaction.

 

In your setups, did you have the Helix's Send set to instrument level? Does it make a difference if you use the 1/4" output versus the Send?

 

The other thing I'd add is if you haven't open a ticket up with Line 6 about this yet, it might be worth doing. Over on TGP, there were several people who had some noise issues with the HX Effects, and Line 6 looked into and came to the conclusion that their units had some components that were out of spec. So if you are having what you believe is excessive noise, you shouldn't necessarily assume that it should be working that way.

[nhoven]

It's an issue with certain preamp designs, I agree. My conclusion: The Helix (my Helix) simply doesn't work with certain designs. The issue is: Those are very classic and popular designs. The amps I've seen mentioned, both in this thread and elsewhere, share the same design principle of using multiple tube stages (usually four) with moderate gain to shape the tone instead of a few high gain stages (or adding distortion though semiconductors).

 

Yes, every device has a certain noise floor, and if you amplify that noise floor by 70 or 100, clamp it down, and repeat that four times, that noise floor becomes an issue, that's the nature of it. Also, I don't think the converters are the problem; I would think it's the analog circuitry after the converters.

 

I'm just wondering if nobody at Line 6 ever hat the idea to pair the Helix with an SLO, 5150, Knucklehead, an Überschall, or a Hot Rodded 2203. I can put a cheap Danelectro Vibe pedal in front of my high gain channel, and it produces less additional noise than the Helix. I can't believe no one noticed. Or, as I said before, maybe most people are simply accepting the noise as part of a high gain setup...

 

At this point I think there's nothing that can be done. The 4CM simply isn't possible with certain preamp designs without external gating.

 

Serious Question: Could this be related to the issue in the sticky post about the HX FX Audio Artifacts? I found this youtube video that seems to show exactly the same issue. Is it possible that there's a similar issue with the Helix and it was overlooked, because the 4CM isn't the primary use case (in contrast to the HX FX)? Is it worth filing a support ticket at this point?

 

Really, I'm not trying to shift blame or lollipop off anyone... I'm just bummed that things don't work as planned.

[phil_m]

17 minutes ago, hoven said:

Serious Question: Could this be related to the issue in the sticky post about the HX FX Audio Artifacts? I found this youtube video that seems to show exactly the same issue. Is it possible that there's a similar issue with the Helix and it was overlooked, because the 4CM isn't the primary use case (in contrast to the HX FX)? Is it worth filing a support ticket at this point?

 

Well, that's what I was getting at. That post is the issue I was talking about. It is certainly worth opening up a ticket, imo. The Helix should be useable in 4CM, and indeed it was designed with it in mind.

[nhoven]

okay, I wasn't getting what you were getting at :-). I kind of didn't really read your last paragraph... sorry :-). I will prepare a proper recording of the added noise levels and open a ticket, tehn. I'll keep you guys posted.

[Smooth555]

Please let us know what comes of it. I’m trying to run my Helix lt with a Mesa Mark v in 4cm and having the same noise floor issue. I’ve got the ebtech hum eliminator solving a ground loop in the effects loop, but the hiss is still there. The hiss is there on all channels but particularly prominent when I go to channel 3. I’m managing with a hard gate, it’s not so bad while playing, but significantly noisier than my pedals into the amp. 

[baron55]

Guys guess what. The HX effects guys have the same exact issue and Line 6 found that those units had "parts" that were out of spec. 

 

Here is the official statement.

 

Hi,

Thanks for your patience while we sorted this out.  In extremely rare cases (roughly 0.5%, or 1 out of every 200 HX Effects units sold), certain specific combinations of HX Effects units populated with out-of-spec parts and amplifiers and/or hookup configurations can manifest notable noise and/or other unsavory audio artifacts. Since your particular setup happens to exhibit this behavior, we will have the unit shipped here and have the fix implemented repair it.  We have also updated our manufacturing and testing process to ensure no future units exhibit this behavior, regardless of configuration.  Please update or confirm your accountâÂÂs shipping address and we can start the repair process.

 

[nhoven]

Yeah I got that. I've had the problem since before the HX was announced. My guess is that not enough people reported the issue since (my guess) it's far from being the primary use case for the Helix, so it "slipped by".

 

I've opened a ticket and got the response that Line 6 was aware of the problem and working on a solution... and to be patient.

[patdixon]

Just a fly in the ointment- the noise issue isn't limited to Helix (HX) or even Line 6- did it on my digitech RP300 also- I put a post up a few days ago asking the same question (before I bought anything) because my old X3L did it so badly that I had to put a vol pedal before the amp. I have a non-modified Blackstar Club 40 that I love and sounds great with FX processors but at idle it has a lot of noise. It's not the input, the guitar or how it's wired (does it even in tuner mode).

 

Really after a while I got used to it; it's no big deal- just place a vol pedal before the amp. Just don't forget to toe down before the next song starts. I put a mini switch at the heel of the digitech that shunted the output to grd  but I don't advocate doing that on your Helix. Didn't work on the X3L because of a pop every time the switch engaged.

 

All of these wonderful things like having 1,024 presets come with a price- maybe your ticket will fix the issue for all of us. To be fair I haven't tried the LT with the Blackstar yet;  just now got it but the HX had it before I returned it for unrelated reasons. One thing I haven't tried is putting a gate dead last in the signal path but if it still does it in tuner mode not sure that would work either.

 

Thanks for posting this!

[nhoven]

19 hours ago, patdixon said:

Really after a while I got used to it; it's no big deal- just place a vol pedal before the amp. Just don't forget to toe down before the next song starts.

 

Big deal if you're used to using the volume control. Even if it's a high gain channel I use it very dynamically. It's very annoying if the rig hisses like crazy in the middle of the soft part of a song... hard gating or shutting it down with a volume pedal might work in... I don't know, a metal of punk context, but not in mine.

[patdixon]

Last night I hooked it up to the BS, it's definitely quieter than the X3L in a side by side comparison, albeit still noisy. Also tried putting the noise gate last, no change.

 

I agree that it's complicated using a second vol pedal - another option is to use the amp's channel switching pedal to mute between songs (the clean channel isn't nearly as bad), but still another thing to keep up with and wouldn't work in quiet song sections. Our cover band is Classic Rock, no metal.

 

Hopefully they're finding a solution for all of us.

[nhoven]

Hi there,

 

I thought I'd get back to this topic and give you guys an update. I had been pretty *bleeped* off that a few technicalities (this topic being one, the bugged handling of MIDI CC messages between snapshots being another) prevented me from using this otherwise great piece of gear in the way I intended to when I bought it. So, I let it sit in my music room and didn't touch it for two months.

 

The support ticket I opened delivered no result; the person handling my ticket misread it and thought I had the HX FX. The short answer was: Yeah it's noisy. Live with it. Which sucks.

 

Anyway, I've decided to revisit this and give the 4CM another chance; my (last) plan would be to try it with the mains outs (which are a little less noisy) feeding the amp input and using a return and send to get the signal from the amp's preamp and feed the post FX back to the power amp. I'm guessing it would cost some flexibility regarding pathing and the potential to bypass the tube preamp, but we'll see.

 

If that doesn't work, I'm very ready to give up on the 4CM. At this point, especially since one of the last updates brought the BE-100, which sounds awesome and can be dialed in to sound very similarly to my amps, I'm willing to ditch my amps' preamps and either use the heads as power amps only, or even get a dedicated small SS power amp. My desire to get rid of my heavy rack gear has become far greater than my desire to keep my amp head's preamp in line. ('m getting too old for heavy gear).

[Smooth555]

I'll throw my results in because I hate when there's no resolution on these kinds of threads. L6 basically said that the noise was normal with my Helix LT. I ended up buying an EbTech hum eliminator and that removed my ground loop issues with the effects loop and send return. I still get some RF interference that's amplified on the higher gain settings of my Mesa Mark V but 4CM now works. On those high gain patches, I use a hard gate to kill any additional buzz and that's been quite enjoyable.

 

This ends up being a lot of cables for what was supposed to be a simpler set up (I've got short patch cables between the LT and the EbTech, longer cables to the amp and EbTech, but throw in that I am also using a voodoo lab control switcher to midi control my Mesa Mark V, with midi cable and 4 1/4" into the switches on the back, there's a lot going on) but since I don't play out so much, it's not so bad and it does sound really good.

 

I split my playing time between going through the Mark V and going direct through some Yamaha studio monitors. It's totally dependent on my mood of the day and what I feel like playing, even if I think the amp is still 5% better than the models in the LT. Unless I'm doing rapid fire AB tests, which aren't totally fair to the LT, it more than produces excellent tones that are inspiring to play. The Helix has completely replaced my pedalboard and the footswitch for my Mark V with the exception of a digitech whammy (the pitchshift is just not as good/clean/clear).  I still have GAS for a Friedman Dirty Shirley to compliment my Mark V, but I'm also thinking about getting an FRFR solution like a DXR10/12 or Powercab for the helix because it's very good at the pallet of tones that the Mark V doesn't do so well (mid-gain, vintage voicing, M style types of things and some of the different cab types). 

[nhoven]

Hi folks,

 

ich got back to working on this as by band is on a short break for three weeks. I tried the Helix with the different kind of 4CM-setup I mentioned earlier - using the main out (instrument level) as the output to the amp, feeding the send from the amp's FX loop into a Helix return and passing it back to the amp's return via a Helix send (isolating the send and return with an iso transformer).

 

I set up a pedalboard preset feeding path 1A into 1B and outputing to the 1/4", set the return as input to path 2A, feed into 2B and outputing that to the send. So I have path 1 as the pre FX path and path 2 as the post FX path with 16 slots for FX each, which is ample. I ran out of footswitches before I ran out of DSP :) .

 

Works pretty much as I expected, noise level is quite a bit better than using the officially suggested way feeding the amp input through a send. With the high gain channel, there's still a noteable amount of noise added compared to plugging straight in, so I will have to result in adding a noise gate block after the return, I think; I'll have to test that out at band volume.

 

With that setup, I loose the easy ability to switch the amp's preamp for a modeled one via snapshots or footswitches. Will have to think about if I can get that to work, but  that's not my primary use case anyway. If needed, I might have to do that on a per-preset basis rather than snapshots.

 

I also loose the ability to control the master volume with the big knob, but that I can live with because I bypassed it anyway. So maybe I'll get this to work after all. 

 

My summary to all this: There's been some really crappy hardware choices on the "flagship" device, the noisy as hell sends being one, not using relays for the external amp control being another. I don't get it. I mean, I do, clearly using external gear (especially real amps) obviously wasn't the focus for the Helix, but why offer a plethora of connectivity and then f*** it up with crappy hardware? This is an expensive unit; maybe I'm alone, but I would not have cared if the Helix would have been a few bucks more expensive in trade for relay switching and less noise in the analog circuitry.

[Gritch666]

I feel your pain man, l6 should reworke this noise issue. I'm pretty sure it's hardware related too. BTW I kinda work around my noise problem with a mxr noise gate set at the minimum between helix and the amp front input. Now it's super quiet, but it suck a bit to add an other pedal to my setup...

[robbieb61]

I'm now using a Mesa Triple Crown Amp. And I too found that trying to go out of a send in Helix to the amp input was horrible.

I guess because it's not true bypass.

I don't use any effects in front of the amp except wah..so I changed it up and bought a crybaby. Now I go from guitar to crybaby to amp input and then out of line out on the amp back to return 1 on the helix. I set input on helix to return 1 and set that to line level. 

Then I come out of my  1/4 inch to my frfr and xlr to front of house.

Using midi to control the mesa with my snapshots.

Sounds incredible.

[krisjohnson22]

A little follow up for anyone else who stumbles across this thread, I eventually did find/devise a solution that I'll share. It's not ideal as it take some MacGyvery, but it's definitely done the trick for me.

 

The problem I was finding was that the outputs seem to be using the same opamps or whatever circuitry for both the line and instrument level output and are just switching signal level (and somehow switching output impedance) when switched between instrument or line level outputs. So the the "normally" very low level hiss inherent in any digital signal or DAC isn't so low relative to the much lower level instrument level signal being sent once the output is set to instrument. Basically the noise floor stays at the same level, but the signal is a lot quieter in comparison to it when set to instrument output. Once this goes into the front end of an amp that is applying a ton of gain (level goes way up) and the signal is clipped (guitar signal reduced somewhat it level), the difference between the guitar signal and the noise floor gets even smaller and the hiss is intolerably loud.

 

So the challenge is keeping the signal being output well above the noise floor for the output that is driving the front of the high gain amp (the level feeding the loop return is not being gained up a ton after fact, not being clipped much to reduce the relative signal level). So i built at 20 dB pad that I placed right after the output and then ran to the input of my amp from there. Since the level is being reduced in the DAC, rather than in the analog domain, there is a ton of headroom to push an extra 20 dB out of the output using the send gain control to make up the loss from the pad (remember that I'm outputting a instrument level signal, so it's much lower in level that the line level signal that the output DAC can handle to begin with).

 

So both the noise and the guitar signal are reduced by 20dB, but the guitar signal gets cranked up by 20dB before being output and now appears at the same level as before while the noise floor has been reduced to a level that is a LOT more manageable through a very high gain preamp.

 

The trickiest part was building a pad that was as simple a circuit as possible to avoid tone suck and other harm to the signal, while maximizing the decrease in level AND keeping the input/output impedance of the pad withing a range that wouldn't cause issues for the Helix or the amp. After a quick prototype that worked splendidly, I built 2 more into an old 2 channel DI I had sitting around and velcroed it to the inside of the rack my helix lives in and ran to the inputs of my 2 amps (Boogie Roadster and Soldano). Since I was using the send block as the level control, I could tweak the the boost required to achieve unity gain with the pad (18.5dB) by putting the entirety of the helix in a switchable loop in my RG16 before the amp set to a clean channel and then turn the boost up and down until the level stayed the same when i toggle the helix in and out of the signal path using the loop. This also allowed for a comparison of any potential tone suck or issues from an impedance mismatch in the helix setup vs guitar straight into amp.

 

All in all, I'm pretty happy with the setup and glad I've finally found a solution to the problem I've had since I first setup the PodHD rack however many years ago! I've free to use the Helix for all of the fantastic stuff it does, I just need to remember that each patch needs to begin with the template I created with the boosts for my 2 amps on the sends already built in.

[zolko60]

8 hours ago, krisjohnson22 said:

So the the "normally" very low level hiss inherent in any digital signal or DAC isn't so low relative to the much lower level instrument level signal being sent once the output is set to instrument. Basically the noise floor stays at the same level, but the signal is a lot quieter in comparison to it when set to instrument output.

I can not confirm that. According to my measurements S/N ratio is the same for AD/DA pairs set either to instrument or line level. Full scale levels change  (11dbu instr vs 19dBu line) but the noise floor stays at  -110dBA. 

 

9 hours ago, krisjohnson22 said:

So both the noise and the guitar signal are reduced by 20dB, but the guitar signal gets cranked up by 20dB before being output and now appears at the same level as before while the noise floor has been reduced to a level that is a LOT more manageable through a very high gain preamp.

 

The problem is you usually don't have 20dB of headroom. Typical PAF style humbucker loudest transients appear after 11dBu full scale AD converter at -10dBFS level so clean, undistorted 20dB boost is not possible without clipping.
This is why I am skeptical about your noise reduction solution. It reminds me dolby  but "a form of dynamic preemphasis employed during recording, plus a form of dynamic deemphasis used during playback, that work in tandem" is missing. While kepping close to 0dBFS can be a good advice, it requires constant monitoring and at least limiter protection.

[ChristianBoddum]

For what it's worth I've tried using an AC powerbank to isolate ground issues, just to make sure that any noise i get isn't coming from "dirty" power.

NO! no difference.

I run 1 XLR with phantom blocker to FOH, and run 2 Orange Rocker 30 combos (I'm not sure of the name, but they are tube amps).

I run the combos on the normal inputs on the clean channels (they have no efx loop).

No ground issues, but a high pitched noise before the gates shut of.

The noise is a little reduced when I touch the bridge of guitar, grounding through strings and bridge.

But the noise is always present at some level.

Annoying....

[rd2rk]

44 minutes ago, ChristianBoddum said:

For what it's worth I've tried using an AC powerbank to isolate ground issues, just to make sure that any noise i get isn't coming from "dirty" power.

NO! no difference.

I run 1 XLR with phantom blocker to FOH, and run 2 Orange Rocker 30 combos (I'm not sure of the name, but they are tube amps).

I run the combos on the normal inputs on the clean channels (they have no efx loop).

No ground issues, but a high pitched noise before the gates shut of.

The noise is a little reduced when I touch the bridge of guitar, grounding through strings and bridge.

But the noise is always present at some level.

Annoying....

 

Reviving a zombie thread is probably not the best way to get help with this.

A high pitched noise says USB to me. Are you connected via USB? If you pull the USB connection, does it go away?

[ChristianBoddum]

No USB when playing with the band.

 

[rd2rk]

If the noise is coming through the tube amps, it could be a bad tube. Without hearing the noise myself, that's all I've got!

Good Luck!

[ChristianBoddum]

Definitely not bad tubes. Thanks for the replies :-)

 

[kellyrmartin]

On 6/21/2019 at 4:13 PM, krisjohnson22 said:

A little follow up for anyone else who stumbles across this thread, I eventually did find/devise a solution that I'll share. It's not ideal as it take some MacGyvery, but it's definitely done the trick for me.

 

The problem I was finding was that the outputs seem to be using the same opamps or whatever circuitry for both the line and instrument level output and are just switching signal level (and somehow switching output impedance) when switched between instrument or line level outputs. So the the "normally" very low level hiss inherent in any digital signal or DAC isn't so low relative to the much lower level instrument level signal being sent once the output is set to instrument. Basically the noise floor stays at the same level, but the signal is a lot quieter in comparison to it when set to instrument output. Once this goes into the front end of an amp that is applying a ton of gain (level goes way up) and the signal is clipped (guitar signal reduced somewhat it level), the difference between the guitar signal and the noise floor gets even smaller and the hiss is intolerably loud.

 

So the challenge is keeping the signal being output well above the noise floor for the output that is driving the front of the high gain amp (the level feeding the loop return is not being gained up a ton after fact, not being clipped much to reduce the relative signal level). So i built at 20 dB pad that I placed right after the output and then ran to the input of my amp from there. Since the level is being reduced in the DAC, rather than in the analog domain, there is a ton of headroom to push an extra 20 dB out of the output using the send gain control to make up the loss from the pad (remember that I'm outputting a instrument level signal, so it's much lower in level that the line level signal that the output DAC can handle to begin with).

 

So both the noise and the guitar signal are reduced by 20dB, but the guitar signal gets cranked up by 20dB before being output and now appears at the same level as before while the noise floor has been reduced to a level that is a LOT more manageable through a very high gain preamp.

 

The trickiest part was building a pad that was as simple a circuit as possible to avoid tone suck and other harm to the signal, while maximizing the decrease in level AND keeping the input/output impedance of the pad withing a range that wouldn't cause issues for the Helix or the amp. After a quick prototype that worked splendidly, I built 2 more into an old 2 channel DI I had sitting around and velcroed it to the inside of the rack my helix lives in and ran to the inputs of my 2 amps (Boogie Roadster and Soldano). Since I was using the send block as the level control, I could tweak the the boost required to achieve unity gain with the pad (18.5dB) by putting the entirety of the helix in a switchable loop in my RG16 before the amp set to a clean channel and then turn the boost up and down until the level stayed the same when i toggle the helix in and out of the signal path using the loop. This also allowed for a comparison of any potential tone suck or issues from an impedance mismatch in the helix setup vs guitar straight into amp.

 

All in all, I'm pretty happy with the setup and glad I've finally found a solution to the problem I've had since I first setup the PodHD rack however many years ago! I've free to use the Helix for all of the fantastic stuff it does, I just need to remember that each patch needs to begin with the template I created with the boosts for my 2 amps on the sends already built in.

@krisjohnson22 - Can you provide some details on the components you used to create the output pads?