开云体育

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Zentronics42@...
Sent: Monday, December 6, 2021 10:44 AM
To: [email protected]
Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running.

Specs are always as tight as possible with in reason. If I am affecting a repair I would like the unit to meet factory specs but I am picky in this area. However for this one the owner would like to hook it up to the GPSDO that is in his lab. So I will be looking at making it an external ref unit. Need to dig in to the documentation to see how this was done but that is an option on these.

For me personally I will be adding some rb to the lab at some point. But I am building a reasonably specified cal lab at home.

Zen

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Tom Lee
Sent: Monday, December 6, 2021 12:41 AM
To: [email protected]
Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running.

Hi Zen,

I'm not sure what specs you want to achieve, versus what you need. A 25Hz bang-bang amplitude out of 10MHz is 2.5ppm. Again, for calibrating a scope timebase, that is huge overkill. Even for a lot of telecom applications, 2.5ppm is just fine. If you're just chasing digits because you can, ok; you get to choose how to spend your free time. But if you think that 2.5ppm is a problematic error, I'd be curious as to why.

--Cheers
Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070


On 12/5/2021 19:51, Zentronics42@... wrote:

Thanks again Tom for the math behind the results. I was getting lost in the decimal places. It looks like on the tight end of the spectrum I can get a 20ppm 1 Mhz crystal for under $5. From a reputable source. I know on the Type 184 because I set the crystal in mine it has a +/- 25Hz swing on 10Mhz due to the heating and cooling of the oven. It is never rock stable and constantly moving. I start with a 20ppm unit and see what some long-term testing will tell me.

Zen

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Tom Lee
Sent: Saturday, December 4, 2021 11:12 PM
To: [email protected]
Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running.

If by "IC crystal" you mean "clock oscillator module", you won't have to worry about accuracy (including drift) unless you want to use the TG501 as a timing reference for tweaking, say, a frequency counter. Any commercially available quartz-crystal controlled oscillator will produce an output within 200ppm of the nominal frequency (and most will do considerably better than that). Add to that a worst-case drift of, say, tens of ppm per degree C. For adjusting scope timebases, that's plenty good enough. Here's the math: CRTs can't be expected to do significantly better than 1% linearity, but let's be generous and pretend you've got one that's 10 times better than that. That's 1000ppm. Any quartz oscillator will have much better accuracy than that, even accounting for drift from temperature and supply pushing. So, any error you see will be due to the scope, not the oscillator.

The -01 conversion is at least electrically straightforward, judging
from the schematic. The core of the generator uses a 1MHz reference.
You can use whatever you want instead, as long as you add appropriate
electronics to produce a 1MHz signal. For a 5MHz oscillator, you use
the
divide-by-5 capability of a 74LS90. There are also many excellent 10MHz oscillators out there. You may use those, again, with an 'LS90, but now configured to divide by 10.

-- Cheers
Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070


On 12/4/2021 19:45, Zentronics42@... wrote:

Well it turns out it was a cam switch issue. The unit I have been working on now has a brain. But it is still in need of a heart. I get good time marks on all ranges now. Is there any information on the -01 conversion? I have the service manual but the board lay out is very bad in that one area. Updating to a 5Mhz reference might be the way to go. Also the 1Mhz IC crystals do I need to worry about drift on those much or is this a moot point?

Thanks for the info Harvey, It looks like the only dead IC in the unit was the main REF. This one is confirmed dead as the outputs of 2 of the gates are stuck high and wont switch even with suitable inputs.

Zen

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Harvey
White
Sent: Saturday, December 4, 2021 11:04 AM
To: [email protected]
Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running.

These are straight TTL ICs. The logic 1 level is roughly 4 volts or so. All that a logic 1 needs to do is back bias an (equivalent) input diode. Look at DTL logic (old!) to find an equivalent circuit, and then start replacing the input diodes with (typically) a multiple emitter transistor for a nand gate.

The 16 ma for a chip is also not too out of line, in my experience.

What I'd be looking for would be either slow switching times at the
chip outputs or significantly lower 1 output voltages, or higher than
normal
0 output levels.

Should two TTL outputs be shorted (that are not open collector), bad output levels are to be expected even from good chips.

Note: the output high driver of a TTL chip is effectively an emitter follower with the collector and base resistor going to VCC. There's a limit to what the transistor can pull up, so the 1 output is rated at about 400 ua source, and the standard TTL output (a saturating transistor to ground) is typically 16 ma. Since the output sinking transistor is driven (more or less) by the full VCC through an effective resistor, it gets plenty of base drive.

Harvey


On 12/4/2021 9:40 AM, Zentronics42@... wrote:

Shaun,
Just tried to order the Tauntek but the boards are sold out. I was informed that there will be a delay in getting new boards so. But no worries I am on the waiting list. I will take a closer look at the drum switch. I would imagine it is a cam switch or something in the IC's that the cam is switching. I tested U330 but totally unloaded the single IC is drawing about 16 ma which seems a touch high to me. And when I wire a gate to get a logic level flip the (high) output is only 4Vdc on some of the gates and this is with the IC out of the unit on a breadboard. So it cant get to 5V totally unloaded. VCC is 5.2 to match the TG 5V rail. This is why I am thinking the IC is marginal. Though I could also be asking to much from an IC of this vintage.

Zen

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Shaun M
Sent: Saturday, December 4, 2021 9:23 AM
To: [email protected]
Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running.

Zen,

I had a similar problem (double time marks) on a TG501 unit several months ago. It turned out that one finger of the drum switch assembly was not opening when it should have due to mechanical damage. I found the problem using the switch matrix in the SM which allowed me to figure out the proper position of each switch finger. The matrix is located on the “Display and Switch Details” page of the schematic section in the SM.

On the subject of vintage IC testing: The Tauntek tester lets you see pin voltages and overall current for a given chip as well as the usual logic testing.

Shaun M