|
Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members. |
|
Thread Tools |
25th Oct 2007, 10:56 pm | #1 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
10 MHz sinewave osc. - your design suggestions
I have a need for a 10MHz sinewave oscillator. It needs to produce about 350mV rms into 75 ohms, very low distortion and very frequency stable. DC supply can be anything between +5v and +15v.
Armed with this as a starting point, I acquired a 10MHz "osc.-on-a-chip" device (Epson) that produces a TTL output. (I was surprised how cheap it was!) My first attempts were to route the chip's O/P through an L/C cct. tuned to 10MHz and then to an emitter-follower. Unfortunately, this failed, since the input Z of the EF seems to excessively loads the tuned-cct. A junction FET (higher input Z) didn't help, either. Oh - and I should add - just to avoid any confusion - that this must be a 'solid-state' affair; 'fraid valves are definitely out of it for this job! Any comments on a design approach would be welcome. Thank you. Al / Skywave. Last edited by Skywave; 25th Oct 2007 at 11:02 pm. Reason: Add comment on valves. |
26th Oct 2007, 8:14 am | #2 |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: 10 MHz sinewave osc. - your design suggestions
An emitter follower loaded with 75R will have quite a low input impedance. I would suggest a low pass filter (simple 5 pole butterworth is probably fine in this context) rather than tuned LC. The EF will then probably be OK. Put a 100R resistor in series with the base as a stopper.
Any video opamp will deliver 10MHz into 75R. Old fashioned parts like EL2020 or modern ones like AD8051 are all OK. Typically in video applications the output has a series 75R resistor so that the opamp has to produce twice the voltage requied by the load. Whether you are using EF or opamp beware of capacitative loads which can cause oscillation. A series termination on the output is a complete cure. There shoul dbe loads of application notes on the Analog Devices and Maxim websites. |
26th Oct 2007, 4:26 pm | #3 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: 10 MHz sinewave osc. - your design suggestions
Thanks Jeff - I hadn't thought of this in the context of 'video'; I was thinking exclusively in 'RF-mode'
On reflection, I can see that the LPF approach makes more sense - thanks for the 'steer'. I've had to shelve this for a bit - a few other issues have arrived today which are more pressing. I'll come back here with a report at a later date. Al / Skywave |
27th Oct 2007, 5:49 pm | #4 | |
Dekatron
Join Date: Apr 2005
Location: Bewdley, Worcestershire, UK.
Posts: 4,736
|
Re: 10 MHz sinewave osc. - your design suggestions
Quote:
Would a 10MHz crystal oscillator be stable enough for your purposes? There should be several suitable circuits in Solid State Design. Phil
__________________
Phil Optimist [n]: One who is not in possession of the full facts |
|
27th Oct 2007, 9:05 pm | #5 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: 10 MHz sinewave osc. - your design suggestions
Hi Phil -
Presumably, you are referring to "Solid State Design For The Radio Amateur" by ARRL. If so, yes - tried that source. Page 19 shows a bi-polar and a FET design: both of which gave a considerably distorted waveform. A bit of background - I need to be able to lock a frequency synthesizer from an external source (which must be 10.000 MHz) and experiments to date indicate that the external freq. source must be extremely accurate, pure & stable if lock is to be achieved. A level of about +3 dBm is ultimately required (this bit should be relatively easy ) Aside. 'Solid-state' crystal oscillators - as a 'building block' - have always given me a hard time! Usually, no O/P at all is the result Free-running oscillators, mixers, amplifiers, etc. - no trouble (almost) at all; crystal oscs.: ) That's why I went for the 'osc. on a chip' approach - and at the price - very competitive. However - thanks for the 'steer'. Al / Skywave. |
27th Oct 2007, 9:35 pm | #6 | |
Octode
Join Date: Oct 2003
Location: Newbury, Berkshire, UK.
Posts: 1,770
|
Re: 10 MHz sinewave osc. - your design suggestions
Quote:
I ended up using a small Epson SM device - cost about £2.00 from Farnell ISTR
__________________
Chris |
|
28th Oct 2007, 9:05 pm | #7 | |
Dekatron
Join Date: Apr 2005
Location: Bewdley, Worcestershire, UK.
Posts: 4,736
|
Re: 10 MHz sinewave osc. - your design suggestions
Quote:
Good luck, and no doubt you'll share the ultimate solution with us all. Regards, Phil
__________________
Phil Optimist [n]: One who is not in possession of the full facts |
|
29th Oct 2007, 12:31 am | #8 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: 10 MHz sinewave osc. - your design suggestions
|
29th Oct 2007, 8:15 pm | #9 |
Octode
Join Date: Sep 2004
Location: St. Albans, Hertfordshire, UK.
Posts: 1,477
|
Re: 10 MHz sinewave osc. - your design suggestions
Could you expand on why it needs to be a very pure 10MHz? The frequency synthesizers I am acquainted with are generally constructed with digital dividers and phase detectors and they work best with a square wave. I ask because I have a number of 10MHz xtal reference oscillators looking for a good use but they produce a TTL output.
__________________
Regards, Richard, BVWS member |
30th Oct 2007, 12:18 am | #10 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: 10 MHz sinewave osc. - your design suggestions
OK.
I used the term "freq. synthesizer" rather loosely, since the application isn't that relevant; I was simply trying to give a broad, general picture of the intended use for the sake of completeness. But since you ask, this is a task arising from my employment. There are a number of pcb assemblies that I frequently encounter that incorporate a discrete, custom PLL design, tuning 88 - 108 MHz as the 'tuning heart' of a low-power FM exciter. This cct. has the facility to lock to an external source rather than its own embedded TCXO. To test the functionality of this external port, I usually use an HP 8640B - but there are occasions when this SG is tied up doing something else. The design of the PLL cct. requires the external signal type and spec. as I have stated: the design of the PLL cct. etc. is not under my control; evaluating its functionality is, however. Extensive testing with the 8640B - and using other signal sources - has determined the nature of the I/P signal required; i.e. the accuracy, stability, distortion & level that is required to achieve lock. Al / Skywave. |
30th Oct 2007, 8:23 pm | #11 |
Octode
Join Date: Sep 2004
Location: St. Albans, Hertfordshire, UK.
Posts: 1,477
|
Re: 10 MHz sinewave osc. - your design suggestions
Hmmm, you haven't got much room for manouvre there. Trying to emulate an 8640B is not easy! I agree with all the eminent posters before me that you have to start with a crystal oscillator simply to get the low noise and stability you need. Follow that with an op-amp to get down to 75R and then a LPF to get rid of all the harmonics.
An alternative to the op-amp - assuming the oscillator has a TTL or CMOS output, you could feed it into 3 or 4 TTL buffers (e.g. 74LS06) in parallel which would then drive 75R (honest- I've seen it done)
__________________
Regards, Richard, BVWS member |
9th Nov 2007, 12:29 am | #12 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: 10 MHz sinewave osc. - your design suggestions
Task now successfully completed.
I used the CMOS / TTL 10 MHz 'osc.-on-a-chip' - as referred to earlier - followed it with a 3-section Butterworth LPF and a simple attenuator to get the level down to +3dBm / 75-ohms. No additional amplification was necessary, since the O/P from the osc. is 5v. peak, producing ample O/P from the LPF. Hooked it up to the Item Under Test - bingo! It locks the PLL. Now where did I put that second-hand die-cast box to house this unit ? Al / Skywave. |