Friedman SS loop hiss?

[psychepool]

I made a Friedman mosfet loop and attached it to an amp I'm making.
The amp has clean/crunch/lead 3-channel with the Mesa Rectifier lead channel.

The loop worked very well. Very transparent, and there was no distortion of the stompboxes, so I really like it.
Reducing the output volume at the back of the loop was a great help in using the amplifier at home, and I was satisfied.

A bit unfortunately, when the loop is turned on, a little hiss noise is added compared to the loop bypass.
Turning on the delay or chorus amplifies the noise a little more. (I used H2O. I haven't tested anything else yet)
Even the master volume of each channel is set to 0 and the guitar volume is set to 0 and the noise gate is applied, it does not disappear.
So I'm guessing this is due to the amount of amplification of the return(recovery) part.

Of course, this noise is completely inaudible when the amplifier is used at normal volume.(It's 35W amp with 2 x 6L6)
Even if the attenuator is set at -4dB, it is almost inaudible, and if set at -8dB, the noise is disappears completely.
Of course, if turn down the loop master volume for use at home, it won't be heard at all.

Even if I use it as it is, there seems to be no other problem, but I want to check if it is in a normal state.
Is this circuit originally has some hiss noise?
If not, which point should I check?

The circuit is as follows

ss_loop_sch.png

Made with 1/2W metal film resistor and Panasonic ECQ cap as shown in the photo.

ss_loop.jpg

I haven't used the shielded cable for input/output/send/return because it is still temporarily attached. Will it disappear if I use the shield cable?
However, to assume that it is a cable problem, there is no noise problem when bypassing the loop, and the wire connected to the send/return terminal is too short.

[pdf64]

How about putting an overall master vol after the loop?

[R.G.]

You might be getting RF oscillation heterodyned down to audio. MOSFETs are prone to this.
I suggest putting a resistor of 100 ohm to 1K in series with the gate as close to the body of the MOSFET as you can possibly get it. Q1 has that 2.2M resistor, but unless it's right on the gate, it may not be helping. Likewise, the 1.5K on Q2 would do the job, but it needs to be very close to the MOSFET body. The inductance of even very short conductor distances between a signal source and the gate on the package can tune the MOSFET to oscillate at several hundred MHz.
I would also put a protection zener on the first MOSFET. Won't hurt, might help.

[psychepool]

How about putting an overall master vol after the loop?

I started the master volume set to maximum.
With the master volume set to the maximum, I rotated the internal trimmer until set the level same with loop bypass.
Afterwards, in order to reduce the noise, I turned down the master volume and inter trimmer turned in the opposite direction to the extent that the effector is not distorted.
Currently, the master volume is set at around 3 o'clock. Anything lower than this will require the trimmer to be set too high, causing distortion of the effect.

You might be getting RF oscillation heterodyned down to audio. MOSFETs are prone to this.
I suggest putting a resistor of 100 ohm to 1K in series with the gate as close to the body of the MOSFET as you can possibly get it. Q1 has that 2.2M resistor, but unless it's right on the gate, it may not be helping. Likewise, the 1.5K on Q2 would do the job, but it needs to be very close to the MOSFET body. The inductance of even very short conductor distances between a signal source and the gate on the package can tune the MOSFET to oscillate at several hundred MHz.
I would also put a protection zener on the first MOSFET. Won't hurt, might help.

Do you mean to lay out the resistors connected to the gates of each MOSFET physically close to the MOSFET on the board?
Is the purpose of shortening the line from the resistor to the MOSFET as short as possible?
This board is smaller than you expected, so the line connected to the two MOSFET gates is less than 1cm. Can this distance be a problem?
Currently, the leg of the MOSFET is not cut short and is connected to the socket. If so, is it good to shorten this too?

[pdf64]

The loop has 2 gain pots, P5 sets the send level, VR15 sets the return level.
Turn up P5 until the fx just starts to clip, then back it off a bit.
Then adjust VR15 for the desired loop return level, eg unity.
I suggest to put a master volume after the loop, ie between the loop in-out switch and the power amp.

‘Close’ means to minimise the physical distance between them. So move the body of the resistor closer to the device gate, and cut the device’s leads so that they’re no longer than is necessary.

[psychepool]

The loop has 2 gain pots, P5 sets the send level, VR15 sets the return level.
Turn up P5 until the fx just starts to clip, then back it off a bit.
Then adjust VR15 for the desired loop return level, eg unity.
I suggest to put a master volume after the loop, ie between the loop in-out switch and the power amp.

The content mentioned in the text is the result of testing after setting up as you said.
The master volume after the roof is already installed because it's exist in the circuit (VR15 : 250KA).
Thanks for the answer.

[pdf64]

From the schematic provided, VR15 looks to be labelled ’return level’, it’s part of the fx loop, the whole of which can be switched into the signal path, or bypassed. The return level control would typically be mounted on the back panel, or perhaps even a preset trimmer.

My reading of your post is that your amp differs to the schematic, eg VR15 is a front panel master vol?
If so, is the issue that the loop is hissy even at low settings of VR15?
As that puts it in a rather different light.

[dorrisant]

What he said... The schematic this was pulled from shows switching for two master volumes right before the Channel switch and then to SW9.

VR15 (Return Level) is only in the loop when SW9 is engaged.

What Pete is saying is that you may want to put a master volume after SW9, between pin 5 of SW9 and C35. Also meaning, VR15 would not be used as a global master volume, just a return level control for when the loop is engaged.

At least that's how I see it as it was originally intended.

[psychepool]

From the schematic provided, VR15 looks to be labelled ’return level’, it’s part of the fx loop, the whole of which can be switched into the signal path, or bypassed. The return level control would typically be mounted on the back panel, or perhaps even a preset trimmer.

My reading of your post is that your amp differs to the schematic, eg VR15 is a front panel master vol?
If so, is the issue that the loop is hissy even at low settings of VR15?
As that puts it in a rather different light.

What he said... The schematic this was pulled from shows switching for two master volumes right before the Channel switch and then to SW9.

VR15 (Return Level) is only in the loop when SW9 is engaged.

What Pete is saying is that you may want to put a master volume after SW9, between pin 5 of SW9 and C35. Also meaning, VR15 would not be used as a global master volume, just a return level control for when the loop is engaged.

At least that's how I see it as it was originally intended.

The'return level' on the circuit diagram actually mounted as a pot on the rear panel of the Friedman be100, and the manual says that it can be used as a master volume when the loop is on.
I'm also planed to use this as a master volume, not an internal trimmer.
This is because I don't think it is necessary to attach two same master volumes series to the end of the circuit.
Moreover, I plan to omit the loop bypass because I like the transparency of this loop.

The important thing is that the noise is quite annoying when the 50K trimmer(P5) adjust to the limit at which the effect does not clip, and adjust return level to set unity gain.
As mentioned above, it is a negligible level of noise at normal loud volumes, and when playing at home, if I reduce the master volume (return level) at the back of the loop to the bedroom level, it does not heard at all.
Therefore, there may not be a big problem in use, but compared to the bypass state, since it is a not small level of noise, I wanted to ask if this is a characteristic of this loop, and if not, it means that there was something wrong with it, so I wanted to ask how to fix it.

[psychepool]

While searching, I found a thread like this.
https://ampgarage.com/forum/viewtopic.p ... op#p402657

The circuit diagram below is posted at the bottom of the page.

Metroloop_Rev4_v1_4.png

There are some differences from the loops of the BE100 I posted.

1. Added Zener diode between the gate and source of the send part MOSFET
2. A 33K resistor is added through the switch at the send part output (between the coupling cap and the send jack).
3. Send level trimmer changed to 200K
4. A pull-down resistor (2.2M) is added in front of the input capacitor of the return part.
5. A 3.3K resistor is added through the switch near the source of the return part.
6. Omit the return level pot

I think it's a dip switch mounted on a new loop PCB. (-20dB/+4dB)
The content mentioned in #2 seems to have been applied incorrectly. The coupling cap and send jack are already connected, so connecting 33K between them seems meaningless.

As advised, I'll try to apply the contents of the circuit above while correcting the position of the resistor near the gate and cutting the MOSFET legs short. In particular, I am curious about the result of adding 3.3K of the return stage.

As far as I know, this circuit was developed for non-master volume Marshall amps by Metro amps.
In the case of a Marshall amplifier with a master volume, it is known that it is recommended to mount it between the treble pot and the master volume.
Therefore, it is speculated that there may be excessive attenuation of the send part and excessive amplification of the return stage accordingly.
I wonder if the contents of the circuit above are a measure to adjust it, so I want to apply the contents of the circuit above once.

[pompeiisneaks]

I also reverse engineered the friedman loop I had on a butterslax from this thread:

https://ampgarage.com/forum/viewtopic.php?t=31775

Here's the schematic:



~Phil

[psychepool]

I also reverse engineered the friedman loop I had on a butterslax from this thread:

https://ampgarage.com/forum/viewtopic.php?t=31775

Here's the schematic:



~Phil

I also saw your schematic. Thanks for creating the thread.
I was a little worried because nothing was caught on the source of the send side MOSFET, so I think I will try to work with the above circuit first.
I was wondering what to do because I don't have a 200K trimmer, but according to the data you reversed, I was using 50K, so I decided to leave it as it is. Your schematic was very helpful!

[R.G.]

Do you mean to lay out the resistors connected to the gates of each MOSFET physically close to the MOSFET on the board?
Is the purpose of shortening the line from the resistor to the MOSFET as short as possible?
This board is smaller than you expected, so the line connected to the two MOSFET gates is less than 1cm. Can this distance be a problem?
Currently, the leg of the MOSFET is not cut short and is connected to the socket. If so, is it good to shorten this too?

Yes, as physically close as is practical. If you simply can't get the end of the resistor body next to the MOSFET gate pin/pad, put it a close as you can get it. MOSFETs have usable gain well into the fractional-GHz range, and it's easy for them to self-oscillate in the UHF radio bands if they have a chance.
The culprit is lead/trace inductance. Every wire has a certain self-inductance, and the longer the wire, the bigger the self inductance. A long wire on a MOSFET gate adds some capacitance, which doesn't matter much compared to the gate-source capacitance, but also that inductance of the wire. The resonance of the gate capacitance and wire inductance plus stray signal from the source can excite this inductance and make the MOSFET oscillate.
Adding capacitance doesn't help much, but adding damping does. A resistor directly in series with the gate lead damps the RF resonance and stops possible oscillation. This is necessary (in my opinion) for every MOSFET used as a linear amplifier for audio. RF designers have to make their own black-magic circuits! In my mind, it's much like the addition of a gate protection zener. For me, if there is no gate protection zener, I put one in. I tend to use just one 12V zener from gate to source. The zener breakover protects the gate in the active direction, and the diode forward conduction of the zener protects against transients in the reverse direction. Some circuits may need bidirectional zeners, but unless you already know you need the gate to be driven in the off direction, a single zener works fine.

[gregr]

I made a Friedman mosfet loop and attached it to an amp I'm making.
The amp has clean/crunch/lead 3-channel with the Mesa Rectifier lead channel.

The loop worked very well. Very transparent, and there was no distortion of the stompboxes, so I really like it.
Reducing the output volume at the back of the loop was a great help in using the amplifier at home, and I was satisfied.

A bit unfortunately, when the loop is turned on, a little hiss noise is added compared to the loop bypass.
Turning on the delay or chorus amplifies the noise a little more. (I used H2O. I haven't tested anything else yet)
Even the master volume of each channel is set to 0 and the guitar volume is set to 0 and the noise gate is applied, it does not disappear.
So I'm guessing this is due to the amount of amplification of the return(recovery) part.

Of course, this noise is completely inaudible when the amplifier is used at normal volume.(It's 35W amp with 2 x 6L6)
Even if the attenuator is set at -4dB, it is almost inaudible, and if set at -8dB, the noise is disappears completely.
Of course, if turn down the loop master volume for use at home, it won't be heard at all.

Even if I use it as it is, there seems to be no other problem, but I want to check if it is in a normal state.
Is this circuit originally has some hiss noise?
If not, which point should I check?

The circuit is as follows
ss_loop_sch.png


Made with 1/2W metal film resistor and Panasonic ECQ cap as shown in the photo.
ss_loop.jpg


I haven't used the shielded cable for input/output/send/return because it is still temporarily attached. Will it disappear if I use the shield cable?
However, to assume that it is a cable problem, there is no noise problem when bypassing the loop, and the wire connected to the send/return terminal is too short.

I don’t understand how the ring of the return jack can be connected to the tip of the send jack. It seems to me that a standard plug will ground the input of the send jack.

[sluckey]

I don’t understand how the ring of the return jack can be connected to the tip of the send jack. It seems to me that a standard plug will ground the input of the send jack.

Those jacks are mislabeled. They are not TRS jacks. The jacks are tip, tip shunt, and sleeve, like a Switchcraft 12A. The lug he has labeled "R" is really the tip shunt switch. So, with no plugs inserted, the send is connected to the return.