For relatively slow (Hz to 10s of kHz) square waves of known amplitude, you can use some paralleled beefy CMOS buffers powered by a precision supply feeding a relatively low impedance multi-tap voltage divider. Output levels can be adjusted at DC with a DMM, then the buffers can be driven from a convenient function generator (or crystal controlled synthesizer if you want accruate timing). This is limited to 5v max or so with most fast CMOS, but 4000B series can be supplied with 10v or more if desired. (The 4000B series is not as beefy, though.)
toggle quoted messageShow quoted text
On Mon, Dec 7, 2020 at 06:53 PM, Tom Lee wrote:
You're not wrong about the need for lower frequency square waves if you
want to perform the full set of cal procedures. You'd have to supplement
the tinySA with a pulse generator, as you've noted. Luckily, the
performance requirements are pretty lax -- the frequency doesn't have to
be all that precise, and it's easy to produce good, clean waveforms over
the relevant frequency range. Homebrewing it would be entirely feasible,
and I wouldn't be surprised if there's a $10 pulse gen kit on fleabay.
I don't remember the frequencies needed for a 475, but ISTR numbers
between 1kHz and 50kHz. But when I homebrewed mine back in the day, I
used some TTL dividers driving a PNP (I used 2N3906s, I'm sure, since I
had a mountain of them) emitter-coupled output stage to feed a 50-ohm
termination. That stage cleans up the less-than-pristine TTL output to
provide a well-behaved square wave.
I guess you could even square up the 100kHz output of the tinySA to
drive something similar, and you'd be good to go.
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford, CA 94305-4070
On 12/7/2020 18:29, Raymond Domp Frank wrote:
On Tue, Dec 8, 2020 at 02:53 AM, Tom Lee wrote:for which Tek 465/475 adjustments exactly, except for horizontal timing as you
from 100kHz to 360MHz (beyond that you get square waves)Nice! Just an SMA-to-BNC and 50 Ohm feedthrough needed. I'd have to think
describe and observing BW, if the TinySA produces something enough like a sine
wave, which I guess it does between 100 kHz and >300 MHz. For adjusting the
high-impedance CH1 and CH2 step attenuators, a frequency much lower than the
TinySA's minimum (@100 kHz) is necessary. For edge adjustment, using the
square wave could possibly have been nice but 360 MHz (the minimum square wave
frequency) won't make it through to the vertical amp in any useful way.
Am I wrong and/or missing something?