Frank wrote:

No, that is not a promising waveform for driving any sort of
edge-triggered counting circuitry.  There is way too much ringing.  Note
that there is only about a 1.25v margin from the bottom of the ringing
after the rising edge and the top of the ringing after the falling edge.
Any edge-triggered circuitry would have to reliably trigger within
this rather narrow window under all conditions, automatically.  Not
impossible, but very far from optimum.

No idea why the T2-Mini recommends against using a termination that
matches the coax.  The TAPR blurb on the T2-Mini says, "The T2-Mini has
a single low-impedance output that delivers greater than 3.5 volts into
a 50 ohm load."  That is what you need to do -- use a 50 ohm termination
at the load end of the coax between the T2-Mini and the TICC to tame the
ringing.

Once you hook the T2-Mini up to the TICC, you can do that easily by
installing the jumper at the TICC input (JP3), to put 51 ohm resistor
R17 across the input connector.

Doing as you are now -- feeding the T2-Mini output through coax to the
input of your scope -- you will either need to switch the scope input
impedance from 1 Mohm to 50 ohms (if the scope has that feature --
sadly, many digiscopes do not), or use a 50 ohm feedthrough terminator
on the scope input.

Best regards,

Charles

On 5/31/20 8:14 PM, Charles Steinmetz wrote:

Thanks for the comments. Attached is a screen shot with the
Thunderbolt's 10 MHz output again going through the T2-Mini to the
scope, but with a 50-ohm feed-through terminator placed on the scope's
input.

Frank

Hi

Scope basics:

It is not unusual for a scope to have an input channel with a capacitance
of 10 to 20 pf. That does not sound like much but ….

A 10 MHz square wave has significant energy at 30, 50, 70, 90 MHz (and on up).
Unless a coax is terminated at all these frequencies they will be reflected and you
will get a ring.

So 20 pf at 10 MHz is 795j ohms. At 90 MHz 88j ohms. A feed through termination
probably does a reasonable job at 10 MHz …. it’s pretty awful  by the time you get
to 90 MHz. It is somewhere in the “ummm …. errr …” range at 30, 50, and 70 MHz.

At 10 MHz, 50 + 795j ohms will give you an 18 db return loss. ( the reflected signal
will be 18 db below the incoming signal). A voltage at 1/10 the input is 20 db down.
Not all of the reflected signal will be “ring”, it has to bounce off the source first. How
much bounces depends a lot on how good the source match is.

Bottom line - for a good 50 ohm termination, you need to do things another way.
One of the reasons there are compensated scope probes is to get around this
issue. There are a variety of them out there with a range of capabilities. Yes, the
fancy fet probes cost more than something simple …….

A 10 db pad ahead of the feed through is another way to deal with this. It should at
lest get you into the 20 db return loss range.

Even scopes that have 50 ohm inputs may or may not have good 50 ohm inputs.
Is that clearly spelled out in the specs? Often it isn’t …. sorry about that.

Bob

On 6/1/20 9:11 AM, Bob kb8tq wrote:

Thanks for the comments. Very helpful as background beyond this immediate project.

I have a nice JFW variable attenuator, and inserted it between the TADD-2 Mini and the scope to capture the Thunderbolt's 10 MHz signal converted to square wave with 0, 3, 6, 10 and 20 dB attenuation. Screen shots are attached (very small, but there are five of them).

Currently the only probes I have is the stock pair that came with the Siglent digital scope (an SDS1202?X-e). I also recently bought a vintage Tektronix T922 analog scope, but it didn't include probes. Thus I might be in the market for a good set of probes. If there are suggestions for any that aren't more expensive than, say, a T922 or a used Thunderbolt, I'd be very interested.

Thanks again,

Frank

Hi

There’s a lot more to setting this stuff up past the basics. Not all
attenuators have good return loss …. (and on and on). This quickly
gets pretty far of topic for Time Nuts. Simply keep in mind that the
measurement gear and measurement technique very much get
into the results of almost anything you look at ….

Bob

Frank...

What is interesting is that all of your traces appear to be identical except for the vertical scale factor, which suggests that the ringing you are seeing, which is much improved from your earlier plots, is still a little bit of a problem. I am going to guess that , as Bob pointed out, is probably due to a reflection problem in the input side of the attenuator. Could be the attenuator input impedance is not a good match to the source, or could be a cable impedance issue.

I hope that suggests some things you could try experimentally and educationally to sort it out.

Tom Holmes, N8ZM

-----Original Message-----
From: time-nuts time-nuts-bounces@lists.febo.com On Behalf Of Frank O'Donnell
Sent: Tuesday, June 02, 2020 1:19 PM
To: time-nuts@lists.febo.com
Subject: Re: [time-nuts] TAPR TICC and TADD-2 Minis

On 6/1/20 9:11 AM, Bob kb8tq wrote:

Thanks for the comments. Very helpful as background beyond this immediate project.

I have a nice JFW variable attenuator, and inserted it between the TADD-2 Mini and the scope to capture the Thunderbolt's 10 MHz signal converted to square wave with 0, 3, 6, 10 and 20 dB attenuation. Screen shots are attached (very small, but there are five of them).

Currently the only probes I have is the stock pair that came with the Siglent digital scope (an SDS1202?X-e). I also recently bought a vintage Tektronix T922 analog scope, but it didn't include probes. Thus I might be in the market for a good set of probes. If there are suggestions for any that aren't more expensive than, say, a T922 or a used Thunderbolt, I'd be very interested.

Thanks again,

Frank

On 6/2/20 2:06 PM, Tom Holmes wrote:

One trick is to move the attenuator to the other end of the cable (i.e.
at the output of the device) and then run the coax to your 50 ohm scope
input, and see what it looks like.