Calling WRGKMC or other electronics geniuses--preamp question

[irishstu]

Hi all,

 

I'm in the middle of trying to make a pre-amp that will go from a record player to the line-in of an amp. I'm aware of the RIAA curve, etc.

I'm trying to make this as simple as possible and I'm not OVERLY fussy about getting the RIAA curve 100% accurate. Close enough will do.

 

Anyway, here's what I've been trying so far. I've got an 18v DC adapter going into a simple voltage splitter using a 741, and a couple of resistors and capacitors, like the one found here: http://www.paulinthelab.com/2012/06/op-amp-power-from-single-supply.html (and which I've attached).

 

From there, the Vcc+, virtual Gnd and Vcc- feed the simple RIAA preamp circuit found here: http://www.learningelectronics.net/circuits/riaa-phono-preamplifier.html (also attached).

 

That then goes straight into line-in on the amp.

 

My question is, do I need any other components on here to stabilise things? I've found that when I plug this in, and just one output it connected, it works pretty well. When just the other output is connected, it also works great. However, when both outputs are connected, I get a very loud sort of clicking sound which I'd say is about 6Hz, give or take (it actually sounds a bit like when your hard drive has just died). Any idea why I might be getting that sound, what I can do about it, or basically any other ideas?

 

Thanks for any and all advice you may be able to give me. I do have an electronics background, but I've forgotten a lot of stuff.

 

Stu

 

[WRGKMC]

I'm definitely no genius but I think I know what you're problem is.

 

I'm guessing you are taking a stereo output of two amplifiers and feeding them into a mono like input jack of an amplifier?

 

I believe you need a mixer circuit to combine the two stereo outputs and make them a single mono feed.

 

The clicking noise is being caused by one channel feeding back up into the other. Y jacking two signals together

does not make them feed in one direction through a mono plug. one channel can feed back up into the other leg

of the Y jack and overpower the output of that channel and even damage it.

 

The best way to get both channels to feed into a mono jack would be to mono the stereo signal from the record cartridge

going into a mono phono preamp. You shouldn't get much problem running the cartridge mono like that, but you could get some phase issues.

Otherwise you have to buffer/mix the output of the two preamps before its combined to mono.

 

A simple method of doing this would be to use a resistive bridge or cable.

 

The picture below shows two 470 ohms in series with the channels and one 200ohm resistor that takes the Y junction to ground.

This should kill most of the feedback.

 

The better way is to use either buffers or a dual input line transformer.

A good Jenson transformer would likely provide the highest fidelity.

 

http://www.jensen-transformers.com/as/as078.pdf

 

I'd have to dig around for a schematic using dual buffers to combine the two channels, but it

should be easy to find. All audio mixers safely combine channels. Finding something super

simple would be what you want.

 

 

[Mr.Grumpy]

Most bi-polar op-amp circuits don't bother with the "active" voltage splitter and simply use a resistive voltage divider and a couple of capacitors to stabilize the voltage. I guess the advantage of the active voltage splitter is that it's more "stable", I don't really see the benefit vs. the extra complexity...See what happens if you unplug the 741 (it's socketed, right?) and jumper pin 3 to pin 6.

 

I would also use coupling capacitors on the circuit inputs and outputs. Maybe a 1 µfd for the input and a 4.7 or 10 µfd on the output. I suggest using a film cap for the input and a bi-polar electrolytic on the output. Yes, having capacitors in the signal path will limit your bass response, but trust me that's a good thing with records. Turntables and the cartridge DO pickup subsonic signals and feed them to the pre-amp. These can't be heard, aren't musical, but they will eat up amp power, and make your speakers' woofers work harder. And if you want to record, these subsonic rumbles will use up some headroom too...

[irishstu]

Thanks, guys. I really appreciate all of this information, because I've never really dealt with this particular type of circuit and am encountering a variety of hurdles that I've never seen before.

 

First is the voltage splitting. I had wondered if the 741 was really necessary, so thanks for the push I needed to try the circuit without it, Mr.Grumpy.

 

WRGKMC, the output signals are going into separate channels on the amplifier, or rather they are SUPPOSED to be going into separate channels. Perhaps that's something I should check. All through the pre-amp, the two channels should only be sharing the power +/- and the GND (virtual from power supply, connected to the input/output plug shielding, but not to the chassis itself.) The transformer looks like a great solution, but I'm trying to do this with a minimum of cost, so unfortunately that option's out. I'm hoping I can find a "good enough" solution, as there appear to be many supposed solutions on the Net.

 

I'm actually looking into this circuit at the moment: http://phonoclone.com/diy-pho5.html (picture attached to this post)

 

Seems even simpler, while still having very accurate RIAA equalisation. My only question with that one is about the "split 12V supply". I'm guessing for that I'd need a 24V DC split in the middle, rather than a 12V DC, split into +/- 6V, right? Would I need to regulate the power supply, or are the op-amps fairly forgiving with this sort of thing?

 

 

Sorry guys, one more question. Some circuits recommend grounding the phono plug shielding to the chassis, where others recommend separating them. Any thoughts on that? No matter what, I'd have a grounding screw on the chassis for the ground wire from the turntable anyway.

 

 

Again, thanks for ANY thoughts you might have on this. It's very much appreciated.

[WRGKMC]

Its a matter of weather you are running a balanced signal or not.

A differential preamp can be run balanced with a separate ground.

Most other circuits use frame ground for both the signal ground and DC ground.

This acts as a faraday cage to prevent the hot signal wire from acting like an antenna and picking

up magnetic noise like AC hum and amplifying it.

 

Balanced signals are normally used for signal transmission

over longer distances because you can convert high current low voltage signals to higher voltage low current.

This reduces the resistance to the signal over longer distances like mic cables and you have less signal quality

losses from the cable resistance. This shouldn't be needed with a phono cartridge because it should be less than 20'

of less from the amplifier.

 

By the way you haven't mentioned what kind of amp you'll be feeding. That's a key element for sure.

If its something like a guitar amp or PA, all the worry about fidelity is pretty useless because that gear wont be producing anything

remotely close to high fidelity anyway. Audio is a chain and any weak link in that chain bottlenecks the sound quality.

 

If your concern is cost, I wouldn't even be building a phono preamp from scratch.

I'd just buy one for $20 http://www.ebay.com/itm/Pyle-Turntab...-/390761300738

and be done with it.

 

If I wanted some good tones I'd go with tubes. They're a bit more pricy but the benefits in dynamic response and harmonics is worth every penny.

http://www.ebay.com/itm/6J1-x2-HiFi-...-/111332468005

 

http://www.ebay.com/itm/Little-Bear-...-/291094887533

 

Of course if this is an education exercise, that's a different matter. I don't want to discourage anyone from learning how to build from scratch

because that's where you learn to scrounge and make things work with what you got as well. I've done all the scrounging and tweaking

I need to in my life because I did it for a living. If I do build something it comes down to price, and the work involved getting the parts together.

I'd rather buy a kits like these so I don't have to deal with making things work.

 

I also know when I buy parts separately, figuring in my time, shipping costs, buying a housing and all, I cant come close to some of the kits sold.

 

http://www.ebay.com/itm/RIAA-Compens...-/321180975382

http://www.ebay.com/itm/12AX7-ECC83-...-/221383887584

 

Of course building kits is just like painting by numbers. There isn't a whole lot of challenge just assembling something.

But I worked in electronics so long, especially in the parts end of it, I don't see scrounging parts together as any kind of fun.

I also don't want to be bothered making components do things they weren't designed to do. They either fail to do it or they blow

out. Both are a waste of time and money. I'd rather just use a great design to begin with and use the exact parts that work best.

 

I'd rather let the other guy test out every generic transistor out there, run it through a load line analyzer and figure out which ones

function best in the circuit. I know how to do that stuff and will do it repairing my own gear but again its too much like work.

You do that stuff because you had to on the job making generic parts work because original parts were impossible to get.

 

I still like Refurbing old vintage gear however. That is if I can find the parts I need and it isn't in such bad shape it requires a complete rebuild.

[Kazinator]

Hi irishstu,

 

I second the recommendation to use coupling capactors at the input and output of the circuit. Moreover, the one between the stages is not necessary. As it is, your ground is actually 9V above the real ground, and you are coupling this back to your source device (phono cartridge or whatever). If your circuits ground (it's true ground, what you're calling -VCC) is connected to the ground in the source device, then you're sending 9VDC back into the output of that device. Moreover, when you connect both outputs, you compound the problem.

 

The 6 Hz clicking is "motorboating". Since it depends on an interaction with the source, it is difficult to hypothesize about it. Motorboating is an oscillation which has the same causes as any other oscillation: the right combination of phase shift, feedback and gain. It tends to plague single-supply amplifier circuits which have parasitic feedback paths, due to the ground reference voltage not being a true ground, and easily influenced by things like currents in the output stage.

 

I would recommend a redesign of the first op-amp stage to an inverting configuration, like the second stage. This will have a number of benefits.

 

• One is that the phase is preserved through the circuit (probably not a big deal at all; only shows up as a problem if you send the same signal through multiple paths, some of which invert phase, and later mix it, or send to different speakers.)
  
• More importantly, inverting op-amp configurations work better in single supply designs, because the + input of the op-amp is not involved in the feedback loop. If both your op-amps are inverting, then the only thing that the reference voltage from the 741 is used for is for the + inputs of the stages.
  
• By using an inverter for the first stage, you can easily turn it into a mixer (google for "virtual earth mixer" and you will see why). With just a couple of additional passive parts, you can build a correct circuit that handles two inputs and mixes them into one.
  

 

Another thing is: use the real ground for return currents!!! Do not reference the input and output connections to the fake ground. Use the phantom ground only for generating the + input reference for inverting op-amp stages. I.e. your input and output jacks should be grounded to the real ground, and not go back to the 741 circuit. Any use of the virtual ground for dumping return currents creates feedback voltages, because it is not a true galvanic, zero ohm ground. In other words, you can't just close your eyes and pretend you have a dual voltage supply, and just cut and paste dual voltage circuit approaches: treat it as a single supply device which processes a level-shifted around a mid-voltage reference.

 

A note about that 100 pF capacitor on the input: that looks like it is intended to block some RF from entering into the circuit. It assumes that the source device provides some impedance (so that the capacitor can form a voltage divider with that impedance, thereby implementing a low-pass filter). This is not super-critical, and it may be worth playing with the value to see if it is affecting the audio. If taking it out makes the sound noticeably brighter, the value is too high.

 

The use of the 741 is not critical if you follow these advices. What the 741 does is provide a better reference which appears to have a lower impedance than just a stack of resistors and capacitors. It is better able to source and sink current. But if you only use it as a reference voltage, and not for returning currents, then these properties are not critical.

 

An op-amp can be used to make a higher quality voltage reference which improves power supply noise rejection. But that requires a more complicated circuit. The way your voltage reference uses the 741 is just as a buffer. Whatever power supply ripple there is on that buffer's input is simply copied to the output. So the only thing which is smoothing power supply ripple is the resistor-capacitor network. The op-amp is not helping to improve ripple rejection. What can be done to improve ripple rejection is to change the circuit a bit and use that 741 as an an active low-pass filter. An active filter can use relatively small capacitor values to achieve a very low cut-off frequency (compared to using the same components passively) to really kill the supply ripple. It can "look like" you're using 1000 uF or 10,000 uF caps, without actually using such values. Instead of all that extra complexity, I'd much rather just use a center-tapped transformer to give myself a simple, passive dual-voltage supply (or perhaps with a pair of voltage regulators).

 

One last note: are you sure that the record player's outputs are just completely passive connections to the phono cartridge? If they are already amplified or buffered, then you probably don't need a RIAA filter.

[agedhorse]

If using a single ended power supply, you are missing the appropriate input bias network. The original schematic assumes a bipolar power supply from the bian network shown.

 

The phono pickup is an inductive source, no input coupling caps are necessary. IF there is motorboating, the inductive source needs to be considered with any interactions.

 

In practice, a phono pickup source is a balanced source, that's why there are 5 wires coming from a turntable. They are left+, left-, right+, right- and shield. I would expect all kinds of noise challenges without using a balanced connection with a non-current carrying shield (that should extend all the way to the headshell.