Pink noise generator circuit mod

[culomax]

Kris,

I hope I haven't offended by asking the question of simulation. I was just curious to know how you have worked it out and I have every confidence in what you have designed!

Yes, would be good to chat on Skype once I have the circuit built and ready for testing. Hopefully can make a quick video to show the results also.

Thanks again for your time.

PS: If you think I should order additional/spare components can you let me know.

 

[KrisBlueNZ]

I hope I haven't offended by asking the question of simulation.

No, not at all. It was a fair question. I am aware, myself, that people in my kind of work would normally use simulators. I never used them at my previous job, and there were times when using one would have picked up mistakes and saved time.

Recently I tried the free simulator at a site called circuitlab.com and wasn't very impressed with it, and I'm currently not in a position to buy a good quality simulator package. Maybe one day that will change.

Yes, would be good to chat on Skype once I have the circuit built and ready for testing. Hopefully can make a quick video to show the results also.

Cool. I'm glad you have a good attitude to this project. Don't bother with a video unless you want to when it's all up and running. I'm happy to help. It's not as good as working in the field again, but it's better than nothing

PS: If you think I should order additional/spare components can you let me know.

Sure.
It could be worth getting some other JFETs (they're less than a dollar each). Other types that look suitable apart from BFR30 and BFR31 and are also ex stock Digikey in unit quantity are BSR56, BSR58, BFT46, 2SK3666-3-TB-E, DSK5J01P0L, and DSK5J01Q0L. You could get one each of a few of those alternatives. But they're not easy to change round, being SMT.
You might want to get a couple of 200K trimpots (or regular potentiometers) for playing around with the delay and ramp-up times.
If you don't have a stock of assorted value resistors, you might want to get a bulk resistor assortment, especially if you're planning to do more project work.
When you order your electrolytic capacitor for CT, aim for one with 20% tolerance or better. Aluminium electrolytics aren't particularly accurate. Tantalums are more accurate but tend to go leaky so probably best not to use them!

Everything there should be ex stock in unit quantities at Digikey. Mouser probably have everything as well.

Edit: You might want to get a pushbutton to connect across CT for testing. That would be the easiest way to restart the delay and ramp. You don't need a pushbutton though, you could just solder two wires and bend them to touch each other to discharge CT.

 

[culomax]

Ok cool.

I'll add to the list. I was not aware of DigiKey before. I was going to use RS Components as they are local to me but guess it doesn't matter either way.

Last thing was I need to work out the component layout for the stripboard. I was hoping to find a simulator that can do this for you. I download LT Spice but it seems a bit crap. Guess I'll work it out as I go along.

 

[KrisBlueNZ]

Google "stripboard layout". There are several packages, including a web-based one. You could also do it on paper.

RS Components have a few N-channel JFETs too. Their best bet is the J310G which is a TO-92 part. MMBFJ309LT1G looks suitable but is SMT. Other ones I found on the RS Components site either have no data sheets, or the data sheets are too brief.

RS also have the MC34071 op-amp and they're bound to have everything else as well.

You might also want to get an 8-pin DIP IC socket for the op-amp.

Edit: I just checked and they don't have the TLE2426. You'll have to get it from Digikey or Mouser.

 

[culomax]

Ok so here is my component list. This is to make two circuits - one for each speaker.

Im not sure what the power rating should be on the resistors. Is 1/2W sufficient?

Also what rating the push button switch should be - would this be sufficient: link

Components List

Resistors - 14 Total
2 x 33R
8 x 100k
2 x 1M
2 x 4.7k
2 x 100k - Variable
4 x 200k trimpots

Capacitors - 8 Total
4 x 47uF
4 x 100uF - 25v 20% Tolerance or less

Diodes
2 x 1N914

IC's
2 x TLE2426CLP
2 x MC34071CP - 8-pin DIP IC socket
2 x BFR30/31
2 x BSR58
2 x DSK5J01P0L

2 x Power Switch
4 x Crocadile Clips
2 x Strip Boards

 

[KrisBlueNZ]

Im not sure what the power rating should be on the resistors. Is 1/2W sufficient?

Yes that's fine.
The pushbutton you linked to is fine but put a 1K resistor in series with it to limit the peak current - it's only rated for 100 mA. I would just use a couple of wires that you can touch together. That's a lot cheaper than a pushbutton!

2 x 100k - Variable
4 x 200k trimpots

The 100K pots are to set the output level. I assumed you would use trimpots for those.

2 x Power Switch

I'm not sure what these are for. I assumed you would connect the noise generator circuitry directly to the amp, so when the amp is running, the generator would be running.

Another issue. Do you intend to have a switch to disconnect the noise generator signal from the amp inputs? I think those powered speakers are too valuable to be permanently converted to noise generators with no option of using them for other purposes.

4 x Crocadile Clips

Are these to get power from the amplifiers?

You'll also need wire, plugs or more crocodile clips to get the signal into the amp inputs, some way to attach the stripboard to the noise generator board such as nuts and bolts, and solder.

 

[culomax]

Hi Kris,

I want to mount the push button switch on the Speaker cabinet for switching on/off the PNG. I have called them power switches on my list which is probably a bit confusing. Thought the led on the switch would be useful for knowing the current state of the PNG.

I will use trimpots for the Op-amp gain control. I just knew it as a variable resistor before I looked at the component suppliers website. What sort of current do you think the circuit might draw just so I can be sure I'm picking components with correct ratings?

The croc clips are just for testing purposes to supply power and take output to the amp. Thinking about it I probably need 8 in total. For testing I will just feed the circuit through the 1/4" jack input and run off an 12V battery. All going well I will hard wire the PNG signal internally to the 1/4" Jack input.

Was thinking to hot glue the circuit internally somehow or maybe screws would be better. Will figure that out when I open up the cabinets and see what space is available.

Thanks again for your input.

 

[culomax]

Just had another thought...I should keep room for a voltage regulator on the board just in case. Would it be worth just adding it to the board either way as I dont think the speaker manufacturer's are going to supply me with the schematic for the amp.

 

[KrisBlueNZ]

I want to mount the push button switch on the Speaker cabinet for switching on/off the PNG. I have called them power switches on my list which is probably a bit confusing. Thought the led on the switch would be useful for knowing the current state of the PNG.

Sure, you can switch the power to the noise generator on and off, but:
1. The pushbutton switch you linked to is a momentary pushbutton. It conducts WHILE you HOLD it in. It's not a push-on, push-off switch. So you would have to hold it in to keep the noise generator going! You need one like http://www.digikey.com/product-detail/en/GPB527B2L02BR1/CW150-ND/2349745 which I describe as "push-on, push-off". Digikey call it "On-Off", as opposed to "(On)-Off" for the momentary type, where the brackets around "On" indicate that it's momentary, i.e. you have to hold the button in.
2. Turning the noise generator off isn't the same as disconnecting its output from the input of the speaker amplifier so you can put another signal into the amplifier. If you want to be able to do that, you really need to disconnect the output of the noise generator from the amplifier input. You don't really need to power down the noise generator, as far as I can see, unless you want to be able to restart the ramp-up without turning the whole speaker off. So you could use a push-on push-off button in series with the connection from the noise generator output to the amplifier input. If you want an indication of whether the noise generator is enabled or not, you could use a switch that has two clear positions, such as a toggle, rocker or rotary switch, or use a push-on push-off button with two poles (two separate sets of contacts) and an illuminator, with the second contacts used only to switch the illuminator.
3. As I said, if you want to be able to feed other signals into the amplifier, you need to disconnect the output of the noise generator. Even if the noise generator is not powered up, the output stage of the op-amp will partially conduct in response to the signal source you're feeding into the amp, and will cause that signal to be distorted.

I will use trimpots for the Op-amp gain control. I just knew it as a variable resistor before I looked at the component suppliers website. What sort of current do you think the circuit might draw just so I can be sure I'm picking components with correct ratings?

For the trimpots? They carry almost no current. Anything you find will be suitable. Ten-turn trimpots are available for very fine control and they'll probably be more stable when subjected to vibration. They take a long time to adjust though! You might want to get a trimpot adjuster, which is like a small flat-blade screwdriver with a skirt attached to it, so it doesn't slide off the adjustment screw.

The croc clips are just for testing purposes to supply power and take output to the amp. Thinking about it I probably need 8 in total. For testing I will just feed the circuit through the 1/4" jack input and run off an 12V battery. All going well I will hard wire the PNG signal internally to the 1/4" Jack input.

OK, remember what I said about disconnecting the signal from the noise generator circuit if you want to use an external signal source.

Just had another thought...I should keep room for a voltage regulator on the board just in case. Would it be worth just adding it to the board either way as I dont think the speaker manufacturer's are going to supply me with the schematic for the amp.

Yes, good idea. You might want to get some 78L15 devices, on the assumption that you'll be able to find a supply rail in the range 18~35V somewhere in the amplifier.

Edit: Check out this pushbutton: http://www.digikey.com/product-detail/en/GPB527B2L02BR1/CW150-ND/2349745 It's way over-rated (20A current rating, probably meant for automotive use) but it has an illuminator and it has two poles (two sets of contacts).

 

[BobK]

2. Turning the noise generator off isn't the same as disconnecting its output from the input of the speaker amplifier so you can put another signal into the amplifier. If you want to be able to do that, you really need to disconnect the output of the noise generator from the amplifier input.

Or simply connect both outputs through a 1K or so resistor to add them.

Bob

 

[CocaCola]

Or simply connect both outputs through a 1K or so resistor to add them.

There are many passive or active mixer circuits to be found, but attenuation will happen with most so that has to be considered and factored since you are trying to push the limits...

 

[culomax]

1. Lol...I didnt spot the switch function. I was more concerned about the power rating.

2. Good point about disconnecting the PNG circuit completely to prevent distorting other input signals. I would still like to power down the PNG circuit without switching of the amp as it will be switched on off quite a few times during sound insulation testing. I think I would like to use a 2 pole switch with illumination. That link you posted looks good, cheers.

Just one more thing, (I should stop saying this!). I'm gonna put this out there see what you think. If I could add an RF controller to the circuit that could wireless control the circuit the would be just incredible. I've not looked up the price for these things but do you think it would be hard to incorporate. I'm guessing it just acts like a normal switch when in operation. I guess I would need to make space for the antenna somewhere on the cabinet but price will determine if this is a viable option. I will go look at cost now. Thanks again for you input.

Update: I found this Fob and Rf receiver on RS components. Would this do the trick? They seem inexpensive. I looked on DigiKey but I find that site a lot harder to navigate when you are not entirely sure what you are looking for. Anyway what you guys think?

Ok, on second thoughts I should probably get the PNG circuit up and running first. Will look into the RF control at a later stage. If somebody could provide guidance on what allowance I would need to make to have the option of adding RF control at a later stage that would be much appreciated.

 

[KrisBlueNZ]

BobK, mixing the signals together is not the answer, because when the noise generator circuit is not powered, its output will interfere with the signal coming from another source. The junctions in the op-amp's output stage will not be biased, because the op-amp has no power, but they will conduct on parts of the waveform from the other signal source, because that waveform will back-feed through the mixing circuit. Changing the ratios in the mixing circuit would help but as CocaCola says, will attenuate the signal, unless an active mixer is used (which is overkill IMO), and even with attenuation, a passive mixer won't eliminate the distortion problem, and this is a high-quality audio application.

A simple solution would be to use a relay, with its coil powered from the 15V rail that supplies the pink noise generator main circuit board, and its contacts connected between the circuit output and the amplifier input. This would also ensure that the 15V rail collapses to zero quickly when power is removed, because the relay coil draws significant current. 15V relays aren't available but you can use a 12V one with a series resistor. Here's a nice cheap 12V reed relay: http://www.digikey.com/product-detail/en/8L01-12-001/306-1023-ND/301658
The series resistor needs to drop 3V at 24 mA (the rated coil current) so from Ohm's law it needs to be 125 ohms, use 120 ohms. I've updated the schematic.

Regarding RF control, I agree that would be a great idea. The devices you've found probably aren't the best suited though. They are effectively a serial data link that operates up to 115kbps. You feed serial data in the transmitter and it comes out the receiver. For your application, I would use more of a command-based system - the transmitter would transmit a command for ON, and a command for OFF. (Or you could use a single "toggle" command.) I don't know of a suitable system offhand but there's sure to be something out there. As you say, you could add it later.

 

[CocaCola]

A simple solution would be to use a relay

That could result in some nasty *POPS*

 

[KrisBlueNZ]

Pops occur when there is a DC voltage across the contacts before they close. For example if I put the contacts between the coupling capacitor and the 100K resistor to ground. In this situation, if the contacts are open, the coupling capacitor doesn't charge up, because there's no DC path to ground. Then when the contacts close, a DC spike of half the supply voltage appears on the output as the coupling cap charges up.

The way I've shown it, the relay is between the resistor to ground and the amp input, which should also sit at 0V, so there is no DC voltage across the contacts. To be sure, you could add another resistor to ground on the other side of the relay contacts, i.e. between OUT and GND, just in case the amp doesn't have this. I'll add this to the schematic, ready for the next upload.

 

[culomax]

Guys,

The way I'm reading this is that you are having a healthy discussion and I dont need to add a relay to my circuit.

The PNG circuits have arrived and I want to order the additional components so let me know if there is anything else I'm missing.

 

[culomax]

Regarding RF control, I agree that would be a great idea. The devices you've found probably aren't the best suited though. They are effectively a serial data link that operates up to 115kbps. You feed serial data in the transmitter and it comes out the receiver. For your application, I would use more of a command-based system - the transmitter would transmit a command for ON, and a command for OFF. (Or you could use a single "toggle" command.) I don't know of a suitable system offhand but there's sure to be something out there. As you say, you could add it later.

We have another piece of kit the works on RF control with an on/off toggle. Im going to open it up today and see what module it uses.

 

[KrisBlueNZ]

I think you should get the relay that I linked to. I don't think there will be a problem with pops and I think it's a tidy way to disconnect the circuit's output from the amp input when the circuit is powered off. They're only USD 1.80 each.

Edit: And a 120 ohm resistor and a 1N914 diode for each relay.

 

[culomax]

I think you should get the relay that I linked to. I don't think there will be a problem with pops and I think it's a tidy way to disconnect the circuit's output from the amp input when the circuit is powered off. They're only USD 1.80 each.

Edit: And a 120 ohm resistor and a 1N914 diode for each relay.

Ok, so I take it I just need a SPST Switch for the circuit? Will the relay discharge the JFET capacitor quicker so if the circuit is switched on/off quickly the JFET gate will still ramp up.

Another thing is all the JFET's I can find are surface mounts apart from J310G . Should I just stick with this one?

You mentioned 'plugs' in a previous post but I'm not sure what they are?

 

[KrisBlueNZ]

Ok, so I take it I just need a SPST Switch for the circuit?

Yes. That one needs a current limiting resistor for the blue LED illuminator. Assuming you want to power the illuminator from 15V and operate it at 20 mA (fairly bright), the resistor needs to be R = V / I = (15 - 2.8) / 0.02 = 12.2 / 0.02 = 610 ohms, nearest preferred value is 620 ohms. (the 2.8 in the formula is the forward voltage of the illuminator LED.)

Will the relay discharge the JFET capacitor quicker so if the circuit is switched on/off quickly the JFET gate will still ramp up.

Yes. Also, the illuminator LED running at 20 mA will help the voltage rail to drop quickly at power-off.

Another thing is all the JFET's I can find are surface mounts apart from J310G . Should I just stick with this one?

Yes, the J310G looks fine.
BTW they are somewhat static-sensitive, so you have to handle them carefully. My suggestion is if they come with a bit of conductive foam, keep a small bit of foam connecting all three pins together close to the package, and when you've finished assembling the board, you can just rip the foam out.

You mentioned 'plugs' in a previous post but I'm not sure what they are?

Those were to plug into the amplifier input sockets, if you wanted to do it that way. But if you're going to solder your wires directly to the sockets inside the enclosure, you don't need plugs.

I've just noticed a problem. If you want to turn the noise generator on and off while the speaker amplifier is powered up, there WILL be a problem with pops. Actually, thumps would be a better word. This is because of the time needed to charge and discharge the coupling capacitors in the circuit. I had originally intended that the circuit would be powered all the time, so the start-up delay in the amplifier would hide the initial thump. But I see why you want to be able to start and stop the noise while the speaker is running. So we're going to have to do something about it.

I suggest some changes to the circuit to include a switch that controls the charge and discharge of CT directly. It would also control the relay in series with the output signal. This would become the on-off switch and it would be the one you replace with a relay when you add radio control. The whole circuit, including the noise generator board, would be powered all the time while the speaker amplifier is powered.

Is that OK with you? Let me know and I'll post an updated schematic. The changes will not be large.