Reactive load - adding an XLR out and fan

[nevetslab]

Say no more. Mostly it was a question to solve the what seemed a simple problem yet perplexing all of us. I suppose modeling that circuit, and going thru selectively shorting out portions of the bridge would finally yield the problem you had.

[GainFreak]

I suppose modeling that circuit, and going thru selectively shorting out portions of the bridge would finally yield the problem you had.

I don't think it's worth the effort. What is important the circuit is correct and working properly.

[Chuck H]

I don't think it's worth the effort. What is important the circuit is correct and working properly.

I don't agree with discarding analysis on the basis of the problem no longer being present. There's no "understanding" in that. But then, since there was never any clues as to what the actual problem was, and what seemed to fix it doesn't analyze as causing it, it's probable that any model created won't be duplicating the actual circumstances.?. So in that light light I'm less inclined to take the time. But on the subject of analysis...

I think it's strange that the ground lift on your box is correcting a ground loop. At least as it is shown in the schematic. Because the actual ground for it is achieved via speaker cable contact with the jack sleeve to the amp. And that remains regardless of what position the switch shown in the schematic is in. Unless you have the input jack sleeve isolated from the chassis. Do you?

EDIT: It occurs to me that you probably don't have the input jack sleeve isolated. Because if you did you would lose the XLR sleeve when the ground lift was engaged, causing different noise problems.?. At least as shown in the schematic you provided for analysis. An exception being if the terminal signal end of the XLR is also sleeve grounded. Since you indicated plugging this rig into PC based digital media that seems unlikely, but I don't know.

[Helmholtz]

I think it's strange that the ground lift on your box is correcting a ground loop. At least as it is shown in the schematic. Because the actual ground for it is achieved via speaker cable contact with the jack sleeve to the amp. And that remains regardless of what position the switch shown in the schematic is in. Unless you have the input jack sleeve isolated from the chassis. Do you?

A ground lift is typically used to interrupt an undesirable low impedance ground current path between different mains connected equipment.
In this sense the switch as shown in the schematic should work as it separates the grounds of the guitar amp and the equipment connected to the Line Out.

[GainFreak]

Unless you have the input jack sleeve isolated from the chassis. Do you?

The input jack is isolated from chassis. It has a separate chassis sleeve which contacts the chassis and further connects to the PCB chassis plane (fan circuit, input jack and ground switch are on this PCB). Input jack ground is separate from the PCB chassis plane. The ground lift switch connects both (or not).
The line out circuit is mounted on another PCB where XLR combo socket ground/chassis pin and pin 1 connect to chassis. This means the chassis is always grounded either at the input or output. If both ends are grounded a ground loop is expected

[Helmholtz]

I have to say that Nick's proposal is an excellent addition to a plain reactive load. It's simple, passive and cheap (though I don't exactly understand why it works).
It is often overlooked that while good a reactive speaker load can make the amp behave as if it was connected to a real speaker, the signal at the amp's output is hardly suitable for direct processing, recording or feeding a monitor.
What is missing is the electro-acoustical filtering by a real guitar speaker. I've seen elaborate active filter circuits to emulate a speaker's response, but Nick's solution is much simpler and for free.

[GainFreak]

What is missing is the electro-acoustical filtering by a real guitar speaker. I've seen elaborate active filter circuits to emulate a speaker's response, but Nick's solution is much simpler and for free.

I'll be feeding the signal to an impulse response player so this problem is solved. It can be a PC based one or an external pedal like Mooer Radar for example or similar. IR's are the most accurate speaker/cabinet simulations to date. The reactive load signal can be fed also to a small SS amplifier and this way you practically get an attenuator.
If the amp feeding the reactive load is a tube one then the second amp must be SS otherwise you'll get the impedance curve at the speaker one more time.

[nickb]

Interesting. I see you can very quickly spend a lot of $$$ on impulse "packs".

But, there is a problem here. Take the Mooer Radar, it expects a flat frequency response on the both the input and the output. Using a reactive load and taking the amplifier output voltage to the DI is not quite right to the end of a flat response. You actually need a resistive load. That said the error is quite small at around 1 dB but there is significant extra cost in the hardware and all it did for you was introduce an error.

OTOH it's a bit silly talking about precision in the land of guitar amps

[GainFreak]

Interesting. I see you can very quickly spend a lot of $$$ on impulse "packs".

There are free packs which are quite good There are millions of IR's out there but in the end you settle on couple of them and that's it. Also you can make your own impulses.

Generally there are two types of IRs: 1/captured with an SS amp and 2/ with a tube amp. The first type is a "flat"type reproducing the actual speaker/cabinet response. With tube amp IRs you get the "smile" impedance curve and the tube amp response. So if you have a reactive load (which as close to a speaker load as you can get) you should use the first type otherwise you'll get two impedance curves - the one from your reactive load and the other one captured in the impulse. If you're running a preamp only and want the tube amp response and the speaker/cabinet impedance curve than you should use the second type IR's.
Using resistive load with a tube amp is good only for measurements. Soundwise it's just mediocrel. That's why most resistive attenuators usually suck big time.
Note that a tube amp will start distorting at impedance peaks which will generate harmonics (mainly in the low mid region) which you can not add in any way after that if you use a resistive load.