Toroidal transformer bolt question & mobile phone interference

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have nearly finished putting my amp in a case now. I didn't know how to mount the LED on the front panel really, so I just cut that off for now. I can hear the soft start relay click on and off anyway.

I also have no stand off feet for the bottom yet, and I was just wondering if it is safe to let the bottom transformer bolt touch anything metal?

I know you can't let the top *and* bottom of the bolt touch the case at the same time, but is there anything wrong with letting just the bottom touch another piece of metal e.g sitting on top of my current amplifier?

EDIT: Also, I seem to get mobile phone interference coming out of the speakers connected to the GC. Could this just be the lack of ~100nf? capacitor on the disconnecting network?
 
Best advise for mounting anything with voltage potential greater than 70 VAC or 30 VDC = use insulators where possible.

The center bolt on a torrid transformer is generally a long ways away from any of the windings ... BUT ... if you can, isolate and insulate the big mounting. Plastic or fiber washers and/or rubber o-rings under the bolt head where it may meet the chassis metal usually work. There are actually plastic washers available with cute little rings for the chassis hole. You usually have to search the bins at a surplus store to find these kinds of things, but it is a good idea to try. Check with a volt/ohm meter for isolation. The rules for production manufacture of electronic devices are cryptic and very complicated, but common sence and common cents should rule these kinds of discussions.

Another good reason for insulating and isolating that big bolt with plastic or fiber washers = if the completed amp gets dropped more than a few inches, the plastic will take a lot of the momentum forces and shock and could save the day from a serious distortion or even breakage of your chassis / case.

(Many bigger torrid transformers are shipped with big rubber washers the same size as the torrid = use 'em if you got 'em for the same reasons as above ... both electric insulation and vibration / shock dampening.)

:smash:
 
The point is there must not be a shorted turn! This will happen if there is a completed conductive path around the torriod core - a simple way of this happening is if both ends of the bolt are connected together or grounded.

"Touching" anything is not the issue.


Just make sure that the top of the bolt and its clamping plate cannot be grounded, and/or isolate the bolt bottom from ground.
 
v-bro said:
I don't see why not, aren't toroids a completely closed electro-magnetic system?

No idea :)


FastEddy said:
Best advise for mounting anything with voltage potential greater than 70 VAC or 30 VDC = use insulators where possible.

The center bolt on a torrid transformer is generally a long ways away from any of the windings ... BUT ... if you can, isolate and insulate the big mounting. Plastic or fiber washers and/or rubber o-rings under the bolt head where it may meet the chassis metal usually work. There are actually plastic washers available with cute little rings for the chassis hole. You usually have to search the bins at a surplus store to find these kinds of things, but it is a good idea to try. Check with a volt/ohm meter for isolation. The rules for production manufacture of electronic devices are cryptic and very complicated, but common sence and common cents should rule these kinds of discussions.

Another good reason for insulating and isolating that big bolt with plastic or fiber washers = if the completed amp gets dropped more than a few inches, the plastic will take a lot of the momentum forces and shock and could save the day from a serious distortion or even breakage of your chassis / case.

(Many bigger torrid transformers are shipped with big rubber washers the same size as the torrid = use 'em if you got 'em for the same reasons as above ... both electric insulation and vibration / shock dampening.)

:smash:

Alright, thanks. My toroid is 500va, but didn't come with any rubber / plastic washers. It has some sort of foam pad for the top and bottom, but I think these might be pretty normal. I will look in to some plastic or rubber washers.


cliff said:
The point is there must not be a shorted turn! This will happen if there is a completed conductive path around the torriod core - a simple way of this happening is if both ends of the bolt are connected together or grounded.

"Touching" anything is not the issue.


Just make sure that the top of the bolt and its clamping plate cannot be grounded, and/or isolate the bolt bottom from ground.

Alright, cool :) Time to stick it on top of my current amp then :) I have just drilled some holes in the bottom for feet, but don't actually have any feet yet.

Thanks very much!
 
The phone system interference is most likely due to the speaker cables acting as antennas conducting the signal into the output stage of the amp where it gets rectified (i.e. detected).

First use the R||L network typically specified on the data sheet or an ap note for the chip. If that doesn't kill it, try putting a ferrite bead on the speaker cable close to the amplifier.

I_F
 
I_Forgot said:
The phone system interference is most likely due to the speaker cables acting as antennas conducting the signal into the output stage of the amp where it gets rectified (i.e. detected).

First use the R||L network typically specified on the data sheet or an ap note for the chip. If that doesn't kill it, try putting a ferrite bead on the speaker cable close to the amplifier.

I_F

Ah, thanks. My current commercial amp must have those on its outputs then I guess :) I thought that was only for stability with capacitive cables, so I didn't leave enough room to fit that on these PCB.

Made a fair few mistakes with this one. Considering doing another one with updated boards and things, but then I have no clue what I would do with the one I just built!

Oh dear! :)
 
Commercial amps almost always use those networks to prevent exactly the sort of problem you're having. Sometimes in their enthusiasm to make something better, amateurs leave out critical components in the belief that they are not necessary. In some cases they can get away with it, in others, they can't.

Try the ferrite bead. If putting it on the speaker cable doesn't do the job, try putting it on the input cable.

This sort of interference usually only occurs if the RF source is very close to the amplifier. It also usually requires a fair amount of RF power. Is there a cell phone repeater site outside your window or is your cordless phone base station sitting on the shelf next to the amplifier?

I_F
 
I_Forgot said:
The phone system interference is most likely due to the speaker cables acting as antennas conducting the signal into the output stage of the amp where it gets rectified (i.e. detected).

First use the R||L network typically specified on the data sheet or an ap note for the chip. If that doesn't kill it, try putting a ferrite bead on the speaker cable close to the amplifier.
the Thiel Network on the output is there for two reasons.
To provide a load at very high frequency and to attenuate interference coming in.

The negative feedback loop takes any interference from the cabling right back to the inverting input of the amplfier.

Fit the full Thiel Network (R//L & R+C) not just the Zobel (R+C).

If the Zobel R is taken off the output and then the C connects to audio ground then the NFB tapping point can be from the junction of the R+C rather than the top of the R. This gives a 2pole attenuation of the interference and the extra 4r to 10r does not change the gain of the amp significantly.

BTW. the best place for the Zobel or the Thiel is ON the speaker terminals NOT on the PCB.
 
I_Forgot said:
Commercial amps almost always use those networks to prevent exactly the sort of problem you're having.

Sometimes in their enthusiasm to make something better, amateurs leave out critical components in the belief that they are not necessary. In some cases they can get away with it, in others, they can't.

If only I didn't care about the interference, I would have got away with it :D

Try the ferrite bead. If putting it on the speaker cable doesn't do the job, try putting it on the input cable.

Haven't got one of those to try unfortunately

This sort of interference usually only occurs if the RF source is very close to the amplifier. It also usually requires a fair amount of RF power. Is there a cell phone repeater site outside your window or is your cordless phone base station sitting on the shelf next to the amplifier?

I_F

Nah, I just put the mobile phone right up against the case of the amp :p Considering how often I actually use my phone, it's probably not actually much of a problem, but it's somehow annoying me just knowing it's there, if you know what I mean


AndrewT said:
the Thiel Network on the output is there for two reasons.
To provide a load at very high frequency and to attenuate interference coming in.

The negative feedback loop takes any interference from the cabling right back to the inverting input of the amplfier.

Fit the full Thiel Network (R//L & R+C) not just the Zobel (R+C).

If the Zobel R is taken off the output and then the C connects to audio ground then the NFB tapping point can be from the junction of the R+C rather than the top of the R. This gives a 2pole attenuation of the interference and the extra 4r to 10r does not change the gain of the amp significantly.

BTW. the best place for the Zobel or the Thiel is ON the speaker terminals NOT on the PCB.

Is there enough inductance on a wirewound resistor for this? Or does the inductance pretty much vary between manufacturers?

It seems sort of awk-ward to solder in series. I will give it a go though thanks!!
 
Hi,
Is there enough inductance on a wirewound resistor for this
no.
The resistors should be non inductive to allow them to work correctly at high frequency.

The resistor is parallel with the inductor.
The inductor passes the low frequency signals (audio) and the resistor passes the high frequency signals (and loads the amplifier with a resistive load at RF even though the cables may have become a capacitive load).
 
AndrewT said:
Hi, no.
The resistors should be non inductive to allow them to work correctly at high frequency.

The resistor is parallel with the inductor.
The inductor passes the low frequency signals (audio) and the resistor passes the high frequency signals (and loads the amplifier with a resistive load at RF even though the cables may have become a capacitive load).

Oh I see!

I saw some picture of a board somewhere that looked like an inductor wrapped around a resitor or something is all, I assumed this was a "wirewound" resistor and was sort of... an inductor and resistor in parallel / would do what I want in a single component.

I'll grab me appropriate value metal film resistor and inductor whenever I order stuff next then, cheers :)
 
Hi,
the easy way to make a parallel combination is to wind a coil around a carbon composition resistor using the resistor body as the former. The just solder the coil ends to the resistor leads and solder the resistor leads into the PCB.
It works and is adopted by most designers and builders.
But, it's not the best way.

Mounting a separated coil and resistor on the speaker terminals is better, the question will be "how much better?".

BTW,
most metal film and metal oxide resistors, even the power types, have a helix pattern in the resistive film route.
This helix will inherently behave like an air cored (or ceramic cored) inductor. The inductance is not high, but there are varying views on whether this is deleterious or not. Use carbon if you can find them economically (tolerance is not critical).
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.