advertisement


Quality passive preamp?

Spent an evening last night putting in some work on the simulator towards making my own 256 position passive using relays.... The devil's in the detail! TTL IC's are my weapon of choice for the digital circuitry. I've basically got it going in sim but things like making it impossible for it to reduce volume beyond zero... which means it goes to full volume.. are still WIP! You know how when going through Freeview channel list you get to "1" which is BBC1 but if you click once again it goes to channel 799 or some such? Yeah it works the same way and needs a digital equivalent of a hard end-stop. As it's obviously 8 bit binary I'm thinking of putting a LED volume display on it... in binary! So there will be 8 LED's on the front which you can watch count in binary as the vol goes up and down:D

There's a few potential methods of doing the actual resistor network. It's linear at the moment but I may go for the more usual log type. Then there's a choice to be made of constant input impedance or constant output impedance ladder network. Making it switchable to active is another attractive idea I will likely do.

It will probably have power on/off, vol up, vol down, mute and passive/active buttons above the row of LED's which will move "Kit in Night Rider" styley across the front panel as vol is adjusted.

It will default to zero vol on power on. Making it remember what vol it was last set to would be stupidly difficult with the technology I'm using here.
 
Jez that is a Micky Mouse approach from yesteryear. Also I'd like to see the spec that you are trying to reach for example does that '256 position' relate to a single channel or the total for both/stereo channels. In either case its far too many positions. A standard audio log pot goes from 0 to -40dB and then runs off to a short to give the HiFi nut zero volume. So taking 0.5dB steps you'll only need 80 positions per channel.

You know how I would do this and its the sort of project a 15/16 yo could or maybe has done? I'd use an Arduino controller with the necessary off shelf modules. Its coded in the 'C++' language that compiles down to machine code that is small and fast. The key thing is that the tools and code examples are free and the Arduino hardware/software is open source so the Chinese have produced their own versions (legit) of the controllers (there are several) that are a fraction of the price of the Italian kit. You ought to at least read up on whats out there and for peanuts.

Now I suspect that you'd not thought about programming this sort of kit but its been made unbelievably simple with all the hard work already written in the libraries. You'll probably find code available that would do the functions that you require and just need to bolt them together with a little massaging.

As for price? I'm using the Mega 2560 board @ around £12 delivered but the baby micro controllers come in at around £4 or less. This is for the Chinese stuff. The pucka Italian job will cost around 3 times or so more.

Cheers,

DV
 
Jez that is a Micky Mouse approach from yesteryear. Also I'd like to see the spec that you are trying to reach for example does that '256 position' relate to a single channel or the total for both/stereo channels. In either case its far too many positions. A standard audio log pot goes from 0 to -40dB and then runs off to a short to give the HiFi nut zero volume. So taking 0.5dB steps you'll only need 80 positions per channel.

You know how I would do this and its the sort of project a 15/16 yo could or maybe has done? I'd use an Arduino controller with the necessary off shelf modules. Its coded in the 'C++' language that compiles down to machine code that is small and fast. The key thing is that the tools and code examples are free and the Arduino hardware/software is open source so the Chinese have produced their own versions (legit) of the controllers (there are several) that are a fraction of the price of the Italian kit. You ought to at least read up on whats out there and for peanuts.

Now I suspect that you'd not thought about programming this sort of kit but its been made unbelievably simple with all the hard work already written in the libraries. You'll probably find code available that would do the functions that you require and just need to bolt them together with a little massaging.

As for price? I'm using the Mega 2560 board @ around £12 delivered but the baby micro controllers come in at around £4 or less. This is for the Chinese stuff. The pucka Italian job will cost around 3 times or so more.

Cheers,

DV

Totally disagree! Just because TTL was the way of doing this task back in 1975 doesn't mean it no longer works you know! It does the job very well and as I don't do software is my preferred method, especially as I'm trying to do this at zero cost and I have the TTL IC's to hand. It was good enough to put man on the moon and is well capable of controlling a few relays! To the user there is no difference... you press the "vol up" button and vol goes up, you press "mute" and it mutes... whether it is being done my a processor of by hard wired logic doesn't really matter.
It could all be done with Decatron's and relay logic for real old school!

As far as having 256 steps well why not when it only need's 8 relays and a few IC's to do it? Better than having too few steps and pretty irrelevant as you can clock it up and down at your chosen speed. Just 'cos it has 256 steps doesn't mean you need press "vol up" for 2 minutes to get to half volume! A standard rotary pot has theoretically infinite resolution.. infinite "steps" if you like, and no one has an issue with that! Hey I could add another 4 relays and a couple more IC's and have 4096 steps! And all without having to learn some sort of machine code I've never seen before over a few weeks of uphill learning curve, and then write the code, de-bug it etc.

Now the early digital tuners from Scott and Toshiba were quite something... mainly TTL, on loads of plug in cards, very early CMOS memory IC's that took maybe 4-6 of 16 pin packaged memory IC's to store the station presets etc, phase locked loop all in TTL and discrete, huge red LED displays for frequency... Lovely stuff! The Toshiba even had touch sensitive controls where you touched the glass panel in the appropriate places to control some functions! And this in about 1974-5 ish IIRC.
 
You have missed the whole point.

It took me 20s to find an amateur who built a 4 channel passive pre-amp with remote and rotary control for 60 USD. After all its only a stepped potentiometer no? No magic. There are likely to be many more better versions on the web (I can't be bothered to do this for you) but it shows what can be done quickly and at low cost by Jo(e) Public. No engineering skills needed.

https://www.instructables.com/id/Arduino-Passive-Preamp-With-Remote-Attenuator-and-/

"Skills required for this project:
  • basic electronics
  • soldering
  • arduino programming"
Why waste your time on old technology? Come into the 21st Century and use your skills a lot more productively.

Here is another Arduino based design http://www.dimdim.gr/diyaudio/la-skala-attenuator/

.........and here are ready built modules but with an on-board micro controller http://www.tentlabs.com/Components/page31/assets/Volume_Control_Application_Note_V0_AN07_19_6_10.pdf

BTW I dug out my Arduino after the pfm discussion on measuring audio Watts (What a joke that was!). I am putting together a Watt meter that uses the Hall effect to measure current (insertion loss 1.2* 10^-3 Ohm) and Voltage from amp to speakers and then calculates the actual Watts and outputs as a bar graph. It'll cost peanuts once I have the time to complete it.

Cheers,

DV
 
Last edited:
What’s wrong with getting up off your fat arses and turning a volume knob?! I have never owned an amp with a remote volume in my life!

PS Still liking the idea of a binary display, obviously!
 
As posted earlier, I'm very happy with my TKD potentiometer in a box, but I'd prefer balanced attenuation since my Mdac and amps are balanced. The balanced attenuators I've come across, DACT, Goldpoint, Khozmo etc. are quite expensive for me so I've ordered Shure inline attenuators which in effect are 3 position switched attenuators with 15, 20, and 25db attenuation, and once I've found which of the three levels get me a bit higher than comfortable listening level with the mdac set at around 0db attenuation, I'll use the mdac's digital attenuation for fine volume control. I think Benchmark use this approach for their DAC1-3 with their equivalent of the Shures built in. Anyway, the Shures are not exactly made for this purpose so I hope I don't blow anything up.
.
 
You must have missed the no programming
FWIW I'm actually trying to help Jez. Did you not see my comment "its been made unbelievably simple"? I don't know whether you have actually looked at Arduino IDE (free) to see what I mean.

Here are the basics

void setup() {
// put your setup code here, to run once:
}

void loop() {
// put your main code here, to run repeatedly:
}

...and this uses one pin on the Arduino as a voltmeter and outputs to a serial device connected to the USB port. If your computer is connected then it appears in an Arduino IDE terminal emulator Window or you could connect a TFT display with serial port. If you buy a starter kit £20 -50 you'll get everything you need with code to drive all the different modules. I have a box load of this stuff and for peanuts.

// the setup routine runs once when you press reset:
void setup() {
// initialize serial communication at 9600 bits per second:
Serial.begin(9600);
}

// the loop routine runs over and over again forever:
void loop() {
// read the input on analog pin 0:
int sensorValue = analogRead(A0);
// Convert the analog reading (which goes from 0 - 1023) to a voltage (0 - 5V):
float voltage = sensorValue * (5.0 / 1023.0);
// print out the value you read:
Serial.println(voltage);
}

A Voltmeter with just one line to set-up and 3 lines of simple code. Just in case I've high-lighted these in blue as theose starting with // are comments.

There is code like this for almost any input/output device that you can buy. So you could get by without ever having to read a manual! Its amazing the amount of effort thats gone into this project so that people can learn about computers and micro controllers at pocket money prices and without any tears.

You can see how easy this is for youngsters at school to use and build things like robots...............

Cheers,

DV
 
FWIW I'm actually trying to help Jez. Did you not see my comment "its been made unbelievably simple"? I don't know whether you have actually looked at Arduino IDE (free) to see what I mean.

Here are the basics

void setup() {
// put your setup code here, to run once:
}

void loop() {
// put your main code here, to run repeatedly:
}

...and this uses one pin on the Arduino as a voltmeter and outputs to a serial device connected to the USB port. If your computer is connected then it appears in an Arduino IDE terminal emulator Window or you could connect a TFT display with serial port. If you buy a starter kit £20 -50 you'll get everything you need with code to drive all the different modules. I have a box load of this stuff and for peanuts.

// the setup routine runs once when you press reset:
void setup() {
// initialize serial communication at 9600 bits per second:
Serial.begin(9600);
}

// the loop routine runs over and over again forever:
void loop() {
// read the input on analog pin 0:
int sensorValue = analogRead(A0);
// Convert the analog reading (which goes from 0 - 1023) to a voltage (0 - 5V):
float voltage = sensorValue * (5.0 / 1023.0);
// print out the value you read:
Serial.println(voltage);
}

A Voltmeter with just one line to set-up and 3 lines of simple code. Just in case I've high-lighted these in blue as theose starting with // are comments.

There is code like this for almost any input/output device that you can buy. So you could get by without ever having to read a manual! Its amazing the amount of effort thats gone into this project so that people can learn about computers and micro controllers at pocket money prices and without any tears.

You can see how easy this is for youngsters at school to use and build things like robots...............

Cheers,

DV


You have missed the entire point!! I do not do programming of any sort, any way, under any circumstances. As far as I'm concerned Arduino, Raspberry Pi, PIC's and microcontrollers don't exist and were never invented!! They are not a part of my "armoury" or "tool box" and probably never will be. I have zero interest in them.

Hence when requiring such functionality I use hard wired logic, usually TTL and sometimes CMOS... even discrete transistor and diode logic sometimes (all are still in every day use and still available in SMD form from any supplier. They are not in any way obsolete). I couldn't give a monkeys if the same job can be done by something a tenth of the size and cost using "modern technology".... Hence the A4 size board using loads of TTL logic that's in "The Alchemist" pre amp to decode remote control and handle all the logic for the touch sensitive controls... which are themselves done in CMOS.

This is not something for sale... it's a unit for my own use, intended to be made entirely with whatever is to hand... even if there are much simpler ways of doing it if I spent some money on it. I'd rather spend 50 hours on it than £50 on it!

Hey I could maybe buy a ready built relay based stepped attenuator with Arduino or PIC control electronics, made in China and for £20 from ebay but that is not the point! I like 60's, 70's and 80's technology and I'm a self confessed Luddite towards all the tech that has made mobile phones and lap tops etc available. I can also use parts purloined from old gear such as mil spec reed relays where each one would probably be more than the entire bought in Chifi board to buy!

The tech I use has to be controllable and manipulable by me. That's probably my main paradigm in everything I do in electronics, and without the resources and back up of a huge multinational company. There are all sorts of amazing things that can be done with a BGA IC the size of your thumbnail and with literally hundreds of connections to it but unless you are manufacturing in huge quantities in a big factory with millions of £ worth of kit then the 8 layer PTH PCB, the pick and place robot with 0.001mm precision, the infra red soldering oven and pre heating oven etc etc are not doable... then there's the time and effort in developing the software etc! Even if they were, the time, effort and money required can only really be amortised if the product is to be mass produced! ie it's the cheapest and fastest way to produce a million of something but a non starter if you only want to make one or ten!

Hence to me even a bought in Arduino board is incompatible with my own design paradigms in electronics as although I could apply it I couldn't design it or build it or repair it....(I guess if it was a matter of life and death I could...) it's cheap as chips so if it fails you bin it and use another one....but VERY difficult to work on... and VERY expensive to manufacture in terms of the plant and machinery etc needed.

As time goes on equipment tends to get smaller, lighter in weight and much cheaper in real terms, but it doesn't necessarily do the job any better than the huge, heavy kit from the past and in fact in some ways is often inferior.
The big plus point of the older stuff (apart from it just being fab!) is that if it goes wrong or out of calibration then it can be repaired and re-calibrated, even modified to improve it or add extra features, by a one man band like myself! If you look at my avatar of a part of my bench the test gear is all from between around 1958 to 1980 ish and for all the reasons I just gave... plus I just like collecting and using it! Some items were more than a house to buy new and everything still does what it says on the tin. If it was good enough at the time for the MOD, and companies like Quad, Racal, BAE and numerous research depts in university and gov then it is good enough for me!
It won't get skipped while I'm around! Which is also good for the environment.

Compare a 10WPC class D amp module the size of 10 cigarettes and only £15 to buy to a 10WPC Leak Stereo 20 and it says all the above far more eloquently!!
 
You may as well be speaking a different language DV. It might be the simplest programme task in the world but if you have zero exposure to any programming language and not even the most basic understanding of syntax it's just random letters on a page.
 
You may as well be speaking a different language DV. It might be the simplest programme task in the world but if you have zero exposure to any programming language and not even the most basic understanding of syntax it's just random letters on a page.

Yep and if you have knowledge of another technology that does the same job equally well and have the parts to hand then why would I want to spend possibly weeks learning to programme!?
 
Ive just seen this thread and would like to comment on some statements made about autoformer/transformer volume controls.

First a correctly designed transformer can have a very wide bandwidth. The low frequency cutoff is determined by the transformers primary inductance
and if high permeability materials like Hy-mu 80 are used can result in a cut off of as low as 2hz with a 1khz typical source impedance. The high frequency cut off is determined by the interwinding capacitance and is usually around 100khz with a volume control size bobbin. I think that,s wide enough for audio.

Most signal transformers of this size will show some resonant peaking at around 100khz caused by the transformers leakage inductance interacting with its winding capacitance(s). Autoformers have lower leakage inductance by their nature and show less peaking . This peaking can show up as ringing on a square
wave test signal. As this resonance is well above the audio band it is essentially inaudible. In any case this ringing can be removed by adding a suitable r/c network (zobel) across the secondary to damp the resonance but the commercial pre amps Ive seen dont seem to need use one. The load impedance will also damp the resonance to some extent.

The prime cause of distortion in a transformer will be low frequency saturation. With proper selection of the transformers lamination material, size and stacking,
plus a suitable number of windings this can be kept very low indeed. With low hysterisis lamination materials distortions at all other frequencies are also very low. Certainly low enough not to cause any coloration.

Of course all this is easily measurable and should be specified by the pre - amps manufacturer.
 
What’s wrong with getting up off your fat arses and turning a volume knob?! I have never owned an amp with a remote volume in my life!

PS Still liking the idea of a binary display, obviously!
One reason-for me - is that I listen in a suspended-from-the-ceiling 'Hanging chair from the 60s, which is a pain to get out of/stop swinging, so, once you're in it, you're happy that a 'remote' can change tracks/volume!
 
Does it run on Linux Mint?
Either that is a weak attempt at sarcasm or you are truly ignorant of how microcomputers/controllers work.

The Arduino are micro controllers. They have all the bits necessary to be a computer but they are not as there is no operating system included. Rather like the NASCOM micro computer that I built (more than 3000 soldered joints) getting on for 40 years ago. Even mainframes were initially booted from a front panel with binary switches. You would set the switches for memory location binary 0 and set the switches for the binary OP code to be loaded and press the load button. Then change the switches for memory location binary 1 and so on. Eventually your code got loaded and you then pressed the run button and hoped.

The Arduino runs on machine code as do all computers. When I first started I had to look up the OP codes and put them directly into memory but using a keyboard and VDU rather than switches. Once a symbolic assembler arrived that made life a lot easier but you had to understand all the hardware OP codes interrupts and how the stuff was addressed configured etc. So you really had to know and understand this stuff. Later we had high-level languages so the coding was much more English-like and much easier to write and understand. However the computer needed machine code to run so your program would go through a compiler that turned your high-level code into assembly language and then you'd run the assembler to get the machine code and finally you'd invoke the linker that connected the code into the operating system. The Arduino IDE takes code written in C or rather C++ and compiles it down to machine code for you. The amazing thing though is that all the complexity is hidden from you and the real code sits in libraries. This means that 'coding' is reduced to simple macros where you only need to input some basic parameters to invoke the complex code that drives the hardware. Look again at the voltmeter code above - just 3 lines! This is why 14 year olds can easily grasp this and use these controllers to interface with all sorts of hardware that do useful and interesting things at pocket money prices. They are so lucky.

I'm about to put together an audio true Watt meter once I get the time! This will measure the actual current and voltage and display the Watts going into a loud speaker. You have to remember that speakers are reactive and not a resistive load and this means that unlike a pure resistance the current can go out of phase with the voltage. In a resistor if you increase the voltage then the current goes up by the same factor. However with a reactive load as the Voltage decreases you might see the current increasing and vice versa! It only needs a Hall effect sensor and a few resistors plus the Arduino and a bit of macro code. A school kid could do this.

Cheers,

DV
 
Ive just seen this thread and would like to comment on some statements made about autoformer/transformer volume controls.

First a correctly designed transformer can have a very wide bandwidth. The low frequency cutoff is determined by the transformers primary inductance
and if high permeability materials like Hy-mu 80 are used can result in a cut off of as low as 2hz with a 1khz typical source impedance. The high frequency cut off is determined by the interwinding capacitance and is usually around 100khz with a volume control size bobbin. I think that,s wide enough for audio.

Most signal transformers of this size will show some resonant peaking at around 100khz caused by the transformers leakage inductance interacting with its winding capacitance(s). Autoformers have lower leakage inductance by their nature and show less peaking . This peaking can show up as ringing on a square
wave test signal. As this resonance is well above the audio band it is essentially inaudible. In any case this ringing can be removed by adding a suitable r/c network (zobel) across the secondary to damp the resonance but the commercial pre amps Ive seen dont seem to need use one. The load impedance will also damp the resonance to some extent.

The prime cause of distortion in a transformer will be low frequency saturation. With proper selection of the transformers lamination material, size and stacking,
plus a suitable number of windings this can be kept very low indeed. With low hysterisis lamination materials distortions at all other frequencies are also very low. Certainly low enough not to cause any coloration.

Of course all this is easily measurable and should be specified by the pre - amps manufacturer.

For a truly state of the art transformer yes that's all true... but what's the point? It's a fad, a novelty, a new way of building a mousetrap. A normal passive has vastly lower distortion still and frequency response of DC to >1MHz, no saturation issues, no need to assure precise loading to avoid over-shoot and ringing and for most intents and purposes a low enough output impedance. If you really need lower output impedance then an active buffer will give much lower output impedance than an auto transformer solution, for a fraction of the price and with vastly better performance...
 
No it doesnt have to be a "state of the art" transformer just a properly engineered one. The engineering is not rocket science. I would contest "vastly lower distortion" but no matter. You dont need to assure precise loading as I think I pointed out. An active buffer will have distortion too. One thing I didnt mention about transformer distortion is that its naturally low in high order harmonics and IMD something that active buffers often arent. A TVC merely reflects the impedances it sees and doesnt have an impedance of its own. For example if say the source impedance is 1K and the TVC is set at say -12db, which is about 2/3 on a typical switch the reflected source impedance that the load now sees is 1000/16 or 62R. Similarly if the load impedance is say 20K the pre-amp will see
a load of 20K x 16 or 320K.

I quite understand you championing your own chosen solution but I also feel the need to correct some misunderstandings about transformer engineering so
users can make an informed choice. For those though who arent interested in all this detail just listen and make your own mind up.
 
When I had Slagle autoformers in my system, I also had occasional electrical discharge when switching inputs. Is this something transformers are more succeptible to, or was it more likely a problem unique to my system?
 
No it doesnt have to be a "state of the art" transformer just a properly engineered one. The engineering is not rocket science. I would contest "vastly lower distortion" but no matter. You dont need to assure precise loading as I think I pointed out. An active buffer will have distortion too. One thing I didnt mention about transformer distortion is that its naturally low in high order harmonics and IMD something that active buffers often arent. A TVC merely reflects the impedances it sees and doesnt have an impedance of its own. For example if say the source impedance is 1K and the TVC is set at say -12db, which is about 2/3 on a typical switch the reflected source impedance that the load now sees is 1000/16 or 62R. Similarly if the load impedance is say 20K the pre-amp will see
a load of 20K x 16 or 320K.

I quite understand you championing your own chosen solution but I also feel the need to correct some misunderstandings about transformer engineering so
users can make an informed choice. For those though who arent interested in all this detail just listen and make your own mind up.

We'll obviously never agree as you too are championing your own chosen solution... A TVC manufacturer or designer possibly?
A conventional passive is technically about the closest thing to perfect in the entire pantheon of hi fi. The only distortion is that due to the voltage coefficient of resistance of the resistors and this is so low as to be immeasurable in most cases. It goes right down to DC and as high as you like.... although I guess purely academic, a passive can be made to go up to 100's of MHz if required... as in the attenuator of an oscilloscope.

There are certainly no "misunderstanding" for you to correct about transformers! Transformers do tend to have resonant peaking which, as you admitted yourself, can need suitable loading and/or a Zobel network to deal with, hysteresis distortion, saturation problems, and diminishing abilities at LF due to non infinite primary inductance.

If wanting a constant 10 Ohms output impedance then a buffer can easily do DC - 1MHz and have <0.001% THD and all for pennies.

I see the autotransformer vol control as a solution looking for a problem... whilst I agree that a well designed and executed one can sound superb, I can't see what they bring to the table that a switched attenuator, with or without buffer, already does at least as well,and theoretically better, and for less money...

They've been around for a comparatively short time as "a thing" in hi fi and it seems to me a case of that people have tried and are familiar with both passive and conventional active pre amps and they are, commercially, an attempt to gather new customers by offering a "third method" that people will buy into simply because they've never tried it before...

I adhere to the widely held opinion that transformers in the audio path are the work of the devil, and to be avoided at all cost unless there is no other choice... the output transformer in a valve amp is a rare valid example!

In pro audio certain bits of gear are lauded and fetch high prices because they have "nice", "euphonic" colouration's that an artist or producer may want to have as a part of the artistic process and this is often attributed to the transformers in the signal path, whether a Pultec mic pre amp or compressor or certain Neve channel strips. IMHO the last thing we want is further colouration's or distortions being added on replay... no matter how "nice" they may sound to some people"! "The closest approach to the original sound" is how Quad used to put it... and that'll do for me:)
 


advertisement


Back
Top