Scope Calibrator

[Wendymott]

Hi peeps. You know me... never quiet.... If I am... Im up to something. In this case..I decided,, that what was needed was a Cheap n dirty ACCURATE Scope calibrator. In association withe Terry VK5TM. I devised a reasonably accurate time pulse generator and DC Voltage source to check most scopes that reside in the Forum members places of reverence. "Workshop to you n me".
The time pulse generation was fairly easy with a P16F628A PIC.
It was the next bit that was NOT easy, well for me anyway. The original idea was to select the pulse width outputs using 7 push buttons, and Leds to show which was active. With Terry's assistance, although , he being umpteen time zones away, it was a bit difficult.
To put another obstacle in the way, my Pic programmer Pickit 3 was giving odd results. It took 3 weeks of "hair pulling" to realise the Pickit3 needs to be directly connected to the PC Board rather than a front panel connection.
Once that was sorted, I was still having trouble persuading the button code to work... thus I opted for a KISS approach.
To select each time pulse I used a 7 way rotary switch, not nice, but it works.
Everyone likes leds, so I added leds to show which position was active..
The Pic clock is 20 Mhz, that = 250 n/s per operation.
I will post the schematic on here, and if anyone wants the code, please PM Me.
Now I dont ever claim to be a code writer, as Terry will attest, and I am sure there are many out there that can do better, but it Works.
The DC reference is an AD584, which gives various voltages, but I chose the basic 10V +/- nothing...... and a 1V derived from a potential divider, as again I say KISS.
I will still continue with the 7 button idea, but for now, you have something to look at.
The waveforms displayed are a little "top" distorted, thats the "cheapo" scope probes, the important thing is the timing edges Each pulse is a value of 1 and the space = 9, thus it covers from 1 u/s to 1 sec.
I am sure some may say.. well its not a Bradley 192, but who wants one of those sitting in the cupboard, rarely used... I dont, I got rid of mine ages ago.
If I have missed anything, please feel free to comment.

[Syrinx1]

Hi, what sort of rise times are you getting? Thanks

[Wendymott]

Syrinx. Not looked. More interested in the start and stop points. Its a "quick" check box, not a scientific piece.

[Terry_VK5TM]

Originally Posted by Wendymott

The Pic clock is 20 Mhz, that = 250 n/s per operation.

200 nS Wendy

[Wendymott]

Hi Terry.... I read that its 4 clock cycles....but I will bow to your comment

[Terry_VK5TM]

Yep - 20MHz/4 = 5MHz -> 1/5,000,000 = 0.0000002S (200nS).

[Electronpusher0]

Fantastic Wendy!, I must make one.
Do you plan a pcb? I can help lay one out on KiCad for you if you want.

Peter

[Wendymott]

Hi Terry. Yes you are correct.... I must have had an "off" day.... more common these days ... but hey... thats what the forum is for...
Replied to Peter via pvt message

[Cruisin Marine]

From one extreme to another, here is the one I knocked up one lunchtime at work back in the 80's and still use it now, no expense spared as you can see lol

[Wendymott]

Like busses. Cruisin. Shake the tree and see what falls out

[Wendymott]

Hi Peeps...... Some slight modifications to the design to improve the rise / fall times at 1 u/S pulse. The rise time is now 40n/s and fall time is 50 n/s using a X1 probe. ( Much improved).
The emitter resistor of Q1 is reduced to 270R/2. I used 0805 smd's thus the power dissipation is 0.1W ish.. Q1 is changed to a 2N4401 and the base resistor is now 220R.
One other thing to note.. is that the clock crystal will affect the short time pulses, as in 1 u/S and 10u/S....
It is not noticeable above that. The real way around this is to provide an external clock..... but this project is intended as "Quick n dirty" . I can supply the code in HEX or ASM if requested by PM.

[Syrinx1]

Hi Wendy, thanks for sharing your work on this.

[commie1]

Hi Wendy,

Thanks for sharing this great work of what you have done here, I appreciate the amount of work involved to bring this project to our attention. May you continue forever

Cheers
Commie

[Wendymott]

NOT wishing to bore you good folk, but the LAST iteration is here... I was concerned that when I swapped the clock crystal, the timings were a wee bit out. As this is supposed to be a "calibration" source, I decided that it should be more stable.
Fortunately I had in stock a 20 Mhz XCO, so I have rejigged the pcb to have either the XCO or a crystal. See SCH.
The code only needs a slight change in the config line to say Internal or external clock source.
The nice thing is.. to those who are familiar with PIC Code ASM.. that tiny or even larger changes can be made to any of the timing's. 20 Mhz was chosen as the clock frequency as it gives the smallest change per clock cycle..however other crystals can be used, with a greater inaccuracy, by altering the values in the code. My final timings were tweeked using my Thandar Timer counter.

[mole42uk]

Hi Wendy, of course I have an idea…..I have a 10MHz GPSDO which locks my MI2955 to a known reference. I wonder about making an input on your calibrator to lock the internal clock to an external reference? Maybe not quite so “quick ‘n’ dirty”!

[Wendymott]

Hi Richard....getting away from the KISS theorem...but as the design is out there..see what you can do.... Im finished... moving onto other ideas..

[Syrinx1]

Quite a few PLL projects around using microcontrollers, used oxco's from decommissioned Chinese equipment available cheaply, and GPS receivers.

[Wendymott]

Hi Peeps...... I "Lied"....... well not so much a Lie as a rethink. Our friend and forum member Terry VK5TM, was assisting me with new software that would make the Calibrator a Single push button to select the pulses, rather than my simple 7 position switch select.
It was a pity for him to put in that work, not to use it, thus I re designed the pcb and added a couple of options as well.
The new PCB allows the use of either a 20 Mhz crystal with trimming cap, or a XCO. One line of code selects which version is used.
The PIC is a 16F628A and the code is available in HEX or ASM format, for those wishing to programme their own device.
As Terry or I are not interested in making this device commercially the code is free to use.
I can also supply a Gerber file for the pcb, but cannot guarantee the hole sizes for production by a third party.
Anyone interested please PM me.