sorry just to see if i understand this correctly, are you suggesting that a guitar signal does not need a good quality or hi performance device with good transfer characteristics?
if your viewpoint is that there is no point in sweating over performance because it's a guitar signal i could not disagree more.
are you a musician by chance?
i think not...
Don´t overthink it.
You are talking **electric Guitar** use, so :
1) highest frequency of interest is 5 kHz
2) lowest frequency of interest is 82Hz
3) signal level is minuscule
4) distortion is a bonus, specially transformer distortion.
Not that you willl have much of it anyway, if at all, given the very modest requirements.
I've been playing guitar for 45 years:
Acoustic
Electric rhythm and lead
Bass (currently 5-string, previously fretless 4-string)
I have been a guitar amplifier designer for several years, and my designs are in quite wide use by guitarists around the world.
With guitar, we are dealing with a narrower frequency spectrum, and small amounts of distortion are also considered perfectly acceptable, even for "clean" sounds. All I'm saying, is you don't need to sweat over it. The low cost transformer will do a great job for your guitar, as your ears will tell you.
Highlighted terms above are subjective, meaning different things to different people.
Now if we "put some numbers into it" as I did in my post above, we see that Guitar requirements are VERY modest indeed, compared to minimum Hi Fi ones.
A mains isolation transformer to float a desktop computer or guitar amp, or a USB isolator that works with modern high speed interfaces such as that found on the Boss GT-1 are both more expensive than an expensive audio transformer.
In days gone past some guitar amps dealt with ground loops by having a unsafe ground lift switch.
MJB expensive USB isolators? not necessarily so. In my day job we use a lot of isolated USB dataloggers. Here is what we use for isolators:
Adafruit USB Isolator - 100mA Isolated Low/Full Speed USB ID: 2107 - $34.95 : Adafruit Industries, Unique & fun DIY electronics and kits
Adafruit USB Isolator - 100mA Isolated Low/Full Speed USB ID: 2107 - $34.95 : Adafruit Industries, Unique & fun DIY electronics and kits
The cheap 'Low/Full speed' isolators don't work with Hi speed 480 Mb/s USB.
The cheapest 'high' speed isolator is USD $90.
Hifime High-Speed USB Isolator
The isolator hardware that can do this high speed stuff has only been available for a few years.
Is galvanic isolation of Hi-speed USB impossible? - Electrical Engineering Stack Exchange
Last edited:
Optical isolation is rarely used because it comes with it's own problems. Poor linearity is one, excessive shot noise is another.
Photons, like electrons, are quantized - the signal arrives in "lumps", albeit small ones, so you get shot noise from a light beam just as you do from an electric current.
On top of this, you usually need conditioning electronics on both sides of the isolator, and that also means providing clean power supplies on both sides of the isolator.
This is why the tried-and-true transformer solution continues to be widely used. It's simple, and works very well. Distortion is usually far lower than typical opto-electronics. There is no added shot noise to worry about. Thermal noise from the transformer itself can be kept negligibly low.
And no, you don't need to spend a king's ransom to get a decent signal transformer - all that is just "woo-woo" marketing designed to trick you into spending ten or twenty times what you actually need to spend.
In a recent guitar amp project, I used a $7 Triad TY-250P ( TY-250P Triad Magnetics | Mouser Canada ) to provide a ground-isolated, balanced, output signal from a guitar amp, which could be run direct to the P.A. system.
Put a 1k or smaller load resistor across both primary and secondary, or otherwise make sure it's fed from, and loaded by, 1k. (Have a look at the datasheet to see why you need the resistor: 1k loading gets you bandwidth out beyond 40 kHz, ten times more than you need for electric guitar.)
In my case, I needed to pad-down the signal before sending it to the transformer, which I did with a simple trimpot wired as a variable resistor, in series with the 1k across the actual transformer primary.
This $7 transformer gets you bandwidth far more than you could possibly need for guitar. Distortion is far below what you get from your guitar pickups or guitar speaker. In other words, you will not hear any change in the sound of the guitar due to using this transformer.
The only negative to this transformer is that one single pin is at a non-standard spacing (not 0.1"), so it won't fit generic proto-board or strip board. The problem is easily solved with a pair of side-cutters: snip off the offending pin, you don't need it anyway, because it's not actually connected to any winding!
-Gnobuddy
Maybe I can specify a little bit more what I need.
When you connect a pedalboard to 2 different amps, to get ´stereo´, chance is that you will get hum.
Normally in that case i would go from a stereo reverb pedal with an output impendance of probably around 200R and the destination, the amp, will normally have en input impedance of 1 Mohms, so that is quite a difference.
Lehle of Germany makes pedals that isolate 1 output, but not the other to avoid a groundloop between the 2. He uses Lundahl LL1531 that are quite expensive.
In some other blog somebody pointed us at Oxford transformers oep.co.uk, that deliver quility transformers for a lot less money, about $20 for a shielded version. He pointed at the Z3003C which is a 10K : 10K line transformer that also exist in 600 : 600R with the productnumber Z3001C.
But I am in very much in doubt if any of those have the right impedance ?
I know the impedance will be reflected forwards and backwards, and that you sometimes need a zobel on the secondary, to reduce ringing, but I dont know which impedance is right. They make all sorts of other small transformers too, but they havn´t answered my question yet.
Some of you may know more that I do
How is the best way to fit the 2 different impedances together?
Best regards
Arthur.
When you connect a pedalboard to 2 different amps, to get ´stereo´, chance is that you will get hum.
Normally in that case i would go from a stereo reverb pedal with an output impendance of probably around 200R and the destination, the amp, will normally have en input impedance of 1 Mohms, so that is quite a difference.
Lehle of Germany makes pedals that isolate 1 output, but not the other to avoid a groundloop between the 2. He uses Lundahl LL1531 that are quite expensive.
In some other blog somebody pointed us at Oxford transformers oep.co.uk, that deliver quility transformers for a lot less money, about $20 for a shielded version. He pointed at the Z3003C which is a 10K : 10K line transformer that also exist in 600 : 600R with the productnumber Z3001C.
But I am in very much in doubt if any of those have the right impedance ?
I know the impedance will be reflected forwards and backwards, and that you sometimes need a zobel on the secondary, to reduce ringing, but I dont know which impedance is right. They make all sorts of other small transformers too, but they havn´t answered my question yet.
Some of you may know more that I do
How is the best way to fit the 2 different impedances together?Best regards
Arthur.
Last edited:
This is normal, and best practice for audio equipment. You do NOT want to match line out impedance and line-in impedance; you want a low output impedance, and a high input impedance.
You only want to match input and output impedances when you want maximum power transfer - such as when connecting a valve guitar amp to a speaker. In the case of line level signals, we are looking for maximum voltage transfer (not power transfer). This is achieved by having very low output impedance, feeding very high input impedance.
So there is no impedance matching problem to be solved. However, there is the ground loop problem that you mentioned, and you can solve that exactly the way I described in post #31. Let me outline it again, step by step to avoid any further confusion:
1) Buy two Triad TY-250P transformers for $5.39 USD each: TY-250P Triad Magnetics | Mouser
You also need four 1/4" mono jacks, four 10k resistors (1/4 watt or 1/8 watt carbon or metal film, 5% or better), and some hookup wire and solder.
2) Download and study the TY-250P datasheet: https://www.mouser.com/datasheet/2/410/media-346070.pdf
3) Wire pin 7 and pin 6 together on one TY-250P transformer.
4) Wire a 10k resistor from pin 5 to pin 8.
5) Wire pin 5 to "hot" of a mono 1/4" jack.
6) Wire pin 8 to "ground" of the same mono 1/4" jack.
Very Important: The jack you have just wired will feed guitar amp #1. (Run the usual mono 1/4" plug to 1/4" plug guitar cable from this jack to the input of amp #1.) Let's call this jack "Output #1".
7) Wire a 10k resistor from pin 2 to pin 4 of the TY-250P.
8) Grab another 1/4" mono jack. Wire its ground to pin 2 of the Triad TY-250P.
9) Wire the "hot" of the same jack to pin 4 of the TY-250P.
Very Important: This jack will be your "Input #1". Feed it from one of the two outputs of your stereo reverb pedal. Use a normal mono 1/4" to 1/4" plug guitar cable.
You have just made a ground-isolated signal connection from reverb pedal output #1 to guitar amp #1 input. You're halfway done.
Now we have to repeat the same process with the second TY-250P for the second channel.
10) Grab the second TY-250P. Wire its pin 7 to its pin 6.
11) Wire a 10k resistor from pin 5 to pin 8 of this TY-250P.
12) Wire pin 5 to "hot" of a 1/4" mono jack.
13) Wire pin 8 to "ground" of the same 1/4" mono jack.
Very Important: This jack will be your second output, "Output #2". Connect it to the input of your second guitar amp, amp #2, using a normal guitar cable.
14) Wire a 10k resistor from pin 4 to pin 2 of this second TY-250P.
15) Grab another 1/4" mono jack. Wire its ground to pin 2 of the Triad TY-250P.
16) Wire the "hot" of the same jack to pin 4 of this (second) TY-250P.
Very Important: This jack will be "Input #2".
17) Feed the second signal from your reverb pedal (reverb output #2) to this jack ("Input #2"), using a guitar cable.
You are done. Each of your stereo reverb pedals outputs goes through its own TY-250P, which provides ground isolation, and feeds its own guitar amp.
Total cost of $11 for the two transformers, plus the cost of four 1/4" mono jacks, plus four 10k resistors, plus whatever housing, paint, etc you choose to use.
It would be a good idea to use a metal (aluminum, most likely) housing to provide shielding from electrical interference.
VERY IMPORTANT: use insulated 1/4" jacks, so that they do NOT connect to each other via the metal case! That would ruin the ground isolation we are trying to achieve. I would suggest using plastic 1/4" jacks, such as this one (designed for PCB, use with a small proto board): NMJ4HCD2 Neutrik | Mouser
It would be a good idea to ground the metal case to any ONE of the four grounds in the circuit; for simplicity, let's say you should ground the metal case to "Input #1" ground.
Remember, don't ground anything else to the metal case!
Enjoy!
-Gnobuddy
Thanks a lot for your extensive answer . I understand the thing you say about not to do impedance matching, but how and why did you choose the transformer impedance?
Most of those are coming in 10K : 10K or 600R : 600R. The 600R came from long time ago, because it was used in old telephone systems, and the 10K seems to be used a lot in line transformers, but the one you point at is 1K : 1K.
My plan is actually to extend the concept to 3 ´channels´ wet Left, dry Center, wet Right, to keep the pick attack and clarity in the center channel, but then I just have to add one more transformer.
For people interestet in this concept The Pedal Show on YouTube explains and shows how it sounds very well.
It is this concept that Brian May from Queen uses with his walls of VOX AC-30s.
One also have to angle the 3 amps to reduce freq. cancelling, some say at least 25 degrees.
Best regards
Arthur.
. I understand the thing you say about not to do impedance matching, but how and why did you choose the transformer impedance?Most of those are coming in 10K : 10K or 600R : 600R. The 600R came from long time ago, because it was used in old telephone systems, and the 10K seems to be used a lot in line transformers, but the one you point at is 1K : 1K.
My plan is actually to extend the concept to 3 ´channels´ wet Left, dry Center, wet Right, to keep the pick attack and clarity in the center channel, but then I just have to add one more transformer.
For people interestet in this concept The Pedal Show on YouTube explains and shows how it sounds very well.
It is this concept that Brian May from Queen uses with his walls of VOX AC-30s.
One also have to angle the 3 amps to reduce freq. cancelling, some say at least 25 degrees.
Best regards
Arthur.
If you look at the data sheet, the Triad TY-250P is specified to work as a 1:1 transformer for any impedance from 600 ohm to 10k ohm.
It's true that the frequency response is flattest and widest with 1k on both sides - but 1k is a heavy load for your reverb pedal to drive, and the TY-250P has more than enough frequency response for electric guitar with 10k on both primary and secondary.
So that's why I picked 10k.
Exactly, you got it!
I might be interested if I had roadies who worked for free, and someone gave me a couple more guitar amps, also free.
As it is, I sometimes play my electric guitar through a few pedals and my mixer, simply to avoid carrying one guitar amp; even without the guitar amp I already have more stuff than I want to carry!
Aha, and of course, he does have roadies, and lots of amps, and so much money that they might just as well all be free.
Here is my favourite video featuring the "That pedal show" guys (but no pedals!), along with the very talented Mary Spender: YouTube
-Gnobuddy
Yes that is a great video
On the former calculation there is one point where I do not agree with you:
If you put a 10K across the input on the primary side of the transformer, you parallel it with the transformers input impedance (1k), which gives a resulting impedance of only 909 ohms, which will load the reverb to much.
If you need a 10K impedance you will need to put the 10K in series with the input signal.
Am I right?
Arthur
On the former calculation there is one point where I do not agree with you:
If you put a 10K across the input on the primary side of the transformer, you parallel it with the transformers input impedance (1k), which gives a resulting impedance of only 909 ohms, which will load the reverb to much.
If you need a 10K impedance you will need to put the 10K in series with the input signal.
Am I right?
Arthur
This is a 1:1 transformer - its input impedance will be the same as the load on its secondary.
In my recipe, I am using a 10k resistor on the secondary. Therefore the input impedance will be 10k, not 1k.
If you set up the transformer for 1k input impedance (by using 1k everywhere I specified 10k), and then put a 10k in series with the input, you will drop the signal level by nearly ten times (-20 dB). There will be a huge volume drop, and probably won't be enough signal level to drive your guitar amps properly.
Simply using 10k resistors on both secondary and primary fixes this; the reverb pedal will see a 5k impedance to drive (10k resistor in parallel with 10k reflected input impedance from the transformer). I think 5k shouldn't be too heavy a load for your reverb pedal.
I used this same TY-250P tranformer to provide a mic-level output in a guitar amp I built recently; in my case, I had a line-level signal in the guitar amp, and so I did use 1k resistors across the TY-250P primary and secondary, and a series resistor on the primary side to drop the signal down from line to mic level.
But you have a different requirement, you need to preserve the signal level at the output of your reverb pedal with minimal loss.
I don't think so, but you are thinking it all through, rather than taking it on faith, which is always a good thing!
-Gnobuddy
> 10K across the input on the primary side of the transformer, you parallel it with the transformers input impedance (1k)
The transformer does not *have* an impedance.
It is designed to work with a -range- of impedances connected to it. Sometimes they only specify "one" impedance, but that's only the middle of a range.
An Ideal Transformer would have zero internal series or shunt impedances and would work between any impedances.
The TY-250P has 172 Ohms dead resistance in it. So trying to drive 600 Ohms is just about acceptable (for 5 bucks). 1K load is better (<10% loss). At 10K load the 172r is quite negligible.
A real transformer always goes to zero impedance at zero frequency, rising with frequency. The specified impedance is probably that which gives the expected Low Frequency response. (However the Triad data does not show much effect, which suggests it was taken with a Zero Ohm source.)
The transformer does not *have* an impedance.
It is designed to work with a -range- of impedances connected to it. Sometimes they only specify "one" impedance, but that's only the middle of a range.
An Ideal Transformer would have zero internal series or shunt impedances and would work between any impedances.
The TY-250P has 172 Ohms dead resistance in it. So trying to drive 600 Ohms is just about acceptable (for 5 bucks). 1K load is better (<10% loss). At 10K load the 172r is quite negligible.
A real transformer always goes to zero impedance at zero frequency, rising with frequency. The specified impedance is probably that which gives the expected Low Frequency response. (However the Triad data does not show much effect, which suggests it was taken with a Zero Ohm source.)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.