XR2206 Signal Generator

[Vitesse]

Cheers for that audioguru, turns out somehow I'd managed to reverse the PNP's collector and emitter and I removed the 22pf capacitor and not overheating now however got some issues with the output being very noisy which I do think might have a lot to do with it being poorly shielded/grounded and might clear up once on a proper PCB and in a decent metal enclosure. Strangely it only seems to be affecting sine/sawtooth output and not the square wave.

I've attached some shots taken of my scope screen as I couldnt find an easy way to describe.

What do you guys think? or any suggestions on noise suppression?

 

[Sangoma]

Cheers for that audioguru, turns out somehow I'd managed to reverse the PNP's collector and emitter and I removed the 22pf capacitor and not overheating now however got some issues with the output being very noisy which I do think might have a lot to do with it being poorly shielded/grounded and might clear up once on a proper PCB and in a decent metal enclosure. Strangely it only seems to be affecting sine/sawtooth output and not the square wave.

I've attached some shots taken of my scope screen as I couldnt find an easy way to describe.

What do you guys think? or any suggestions on noise suppression?

There is noise or jitter on the square wave as well!!!

May possibly be to do with the scope/probes

 

[Vitesse]

I've checked the scope and probe using the calibration square wave output and they seem fine

 

[audioguru]

If you built the circuit on a solderless breadboard then each row of contacts and each jumper wire is an antenna that picks up mains hum.
The output cable should be shielded (coaxial or a shielded audio cable).

 

[Vitesse]

Cheers audioguru I was hoping that was the likely cause as I have already done a PCB layout and was just waiting for the nod from people on here before I etch a board for it. When in an enclosure I will use a shielded cable as you recommended between the board and the BNC connector.

Thanks

 

[JimB]

Looking at the scope traces, I would say that it is FMing. ie something is modulating the frequency.

Look at the triangle wave.
The traces all start at one point where the scope triggers and are spread out at the right hand side of the screen.

If you have a control to adjust the frequency, is it on long leads?
If so, try making them short, a couple of inches maximum.

JimB

 

[unclejed613]

i just found this on EXAR's website... they discontinued the XR2206 in 2011
https://www.electro-tech-online.com/custompdfs/2013/03/pdn_11-0112-01-xr2206d-1033.pdf

 

[alhoop]

I would like to know how close you can get to 1 MHz before the square wave output begins to look like a sine wave.
Al

 

[Vitesse]

I'll let you know if i ever get that far mate

I think I have found the possible source of the problem, there seems to be a lot of ripple on the +/- 12V rails, now I'm not 100% sure if this ripple is actually due to my power supply arangements or is being induced by the XR2206 circuit. I have attached a rough schematic of my PSU arrangement for it and was wondering if anyone would give it a once-over.

I'm wondering if higher values for C3 + 4 would help, additionally I think I really should have included 2 smaller capacitors just for suppression. D1-4 are 1N4002, regulators are 7812 and 7912.

 

[JimB]

My suggestions would be:

To increase C1 and C2 to at least 1000uF, 2200uF may be better, but 100uF is a bit marginal.
also
To put 0.1uF caps right at the pins of the regulators, otherwise thay can oscillate at high freqiencies and cause all sorts of problems.

JimB

 

[Vitesse]

I'll try it tomorrow morning, cheers for the advice

 

[alhoop]

Are your inputs to the regulators at least 15vdc?
Al

 

[Vitesse]

The transformer secondary is 12v however it measures around 16v after the rectifier

 

[Vitesse]

Hi Jim, increasing C1+2 and adding 0.1uF (ceramic) close to the regulators helped considerably, also added a 10uF at the far extremities of the supply rails which has further helped. I'm hoping that some of the problems might be due to the PCB laying loose on my workbench as is the transformer with everything earthed via a bolt through a mounting lug and hopefully things might improve when its in a metal box with decent grounding and shielding. On the PCB design I did try and keep the power rails and signal output well separated with a ground-plane surrounding the trace carrying the output.

 

[JimB]

Please define "helped considerably".
Does this mean that "all the problems have gone away", or, "90% of the problems have gone away" ?

Sometimes a breadbord set-up on the bench can have problems with stray signals compared with something sitting above an earthed metal sheet.

As implied by alhoop, your supplies to the regulators can be a bit marginal.
12 volts from the transformer, when rectified and filtered will give about 16 volts.
You need a minimum of 12 + 2.5 = 14.5v at the input of the regulator IC, leaving 1.5v headroom for ripple and low voltage mains.

It is good to know that you have a good improvement.

JimB

 

[dougy83]

The 12VAC is quite marginal for proper regulation. After the rectifier you'll get 100Hz ripple; increasing the bulk caps reduces this, as you have found. Make sure that the input voltage doesn't drop below the minimum specified for the 7812 (using an oscilloscope, not a voltmeter).. from memory it's around 15V (for 3V regulation drop). If this is the correct figure, you'd only have ~1V headroom. You could increase this by using low-dropout regulators

 

[unclejed613]

also, if your circuit board is "bolted to the transformer case" as you say, there could be magnetic pickup of the circuit from the transformer windings. if you look at large power amplifiers, you will see that the core of the transformer is side-on to the amplifier, which reduces stray magnetic coupling between the transformer windings and the circuits. the area marked in red in the drawing shows areas to avoid having the circuit board. think of this as a pair of cones. there is also a pair of cones to avoid on the top and bottom of the core since the top and bottom are the poles of the core.

 

[Vitesse]

OK, bit more checking done, the input voltage getting to the regulators is at +/- 20V mesured on DMM, verified by scope with very little ripple.

The waveforms seem somewhat more defined (attached pics) than before, there is now practically no noticable ripple on the +/- supply rails. I feel there may be some problem to do with frequency stability now rather than interference as if R1 is at its lowest resistance then you get the trace pictured for sine wave, however as soon as you increase R1 value it diminishes to an erratic sawtooth, R6 had to be set rather high to actually get a sine wave which I would have thought should have been somewhere close to its midpoint.

 

[Vitesse]

I've finally folded a little on this now and managed to source a Max038 very cheaply, before I commit to spending a small fortune I thought I'd give a circuit with that on the breadboard a try and see if I can get better results.

 

[audioguru]

An obsolete MAX038 is a high frequency IC that will probably mess-up if used on a solderless breadboard.
Use your breadboard for DC circuits and make a pcb for high frequency circuits.