Test Equipment Anonymous (TEA) group therapy thread

[med6753]

Lots of interesting TE today. 

The TEA closet has been cleaned out. On the left is a now blank area which will accept a scope cart once built. The assorted boxes containing Tek circuit boards and other parts are now in the garage in snap lid totes which should keep the majority of the moisture out. All the parts scopes except one are stacked neatly on the right along with two on the upper shelf with the parts only plug-ins. I had to sacrifice one scope unprotected in the garage. It's the Type 561B I got in PA when I visited nixiefreq. That scope was already full of corrosion and mildew so it won't make much difference. And it gave up it's CRT for the functional Type 561B. But I may at least cover it plastic.

[ch_scr]

[...]

The new one uses simple divide-by-four circuits to provide in-phase and quadrature of the 10MHz input signals (now at 2.5MHz). These are fed into simple XOR gates and RC lowpass filtered. You get two triangle waveforms at 90° phase shift, these are input to an AD7793 high-resolution ADC. Do the math and get the frequency difference.
The new unit has three inputs, one reference and two frequencies that are compared to the reference. So I can monitor two devices vs. a third one at the same time (hopefully saving a few of these 24h debug sessions ).

Great results and project in general   
To me it sounds like it is related to a Dual Mixer Time Difference DMTD?
Where the two signals to be measured are mixed with an offset oscillator and the resulting waveforms LPF'd and compared by a counter?
With your contraption they are mixed each with the reference as well? but as I and Q components and the output digitized,
but you have no amplitude difference (it's all square anyway?) so only phase information is what is relevant and used?

Thanks!
I've read some papers about the DMTD (before and while I built the first one).

Anyway, the first one (100MHz PLL) isn't a DMTD, I've had some ideas on my own that I wanted to test:

One 100MHz PLL is offset by a slightly variable 125Hz signal, the other one is just 10x the input frequency. Then it's mixed down to a 125Hz beat frequency. Within the uC, there's another I/Q mixer that mixes the 125Hz beat frequency down to DC. Its output drives a control loop that keeps the phase constant (e.g. Q is alway zero) by varying the 125Hz signal frequency. As a result, the 125Hz offset signal frequency minus 125Hz equals 10 times the frequency difference of the inputs. The problem is the not-so-good 125Hz subtraction stage (by an I/Q mixer) to offset the first 100MHz PLL, it produces quite an amount of noise.
Next iteration will use DMTD principles.

The new one directly compares the input frequencies by mixing them down to DC (all digital). Each input is fed into an divide-by-four circuit with I/Q output. Two XOR gates do the mixing part (one input I / Q, other input I only), after RC filtering in a triangle wave proportional to the phase difference of the input signals. ADC and software measure du/dt of the mixed down signal and calculate frequency difference. Theoretically one needs only the I (or Q) part, but the triangle isn't very linear at the ends (near GND / VCC), so software automatically switches between them, and by having both signals one can derive the sign of the difference. Amplitude is "1" by definition.

Would this approximate a functional diagram? Please excuse the unstandardized symbolical expression.

[bd139]

Lots of interesting TE today. 

The TEA closet has been cleaned out. On the left is a now blank area which will accept a scope cart once built. The assorted boxes containing Tek circuit boards and other parts are now in the garage in snap lid totes which should keep the majority of the moisture out. All the parts scopes except one are stacked neatly on the right along with two on the upper shelf with the parts only plug-ins. I had to sacrifice one scope unprotected in the garage. It's the Type 561B I got in PA when I visited nixiefreq. That scope was already full of corrosion and mildew so it won't make much difference. And it gave up it's CRT for the functional Type 561B. But I may at least cover it plastic.

Looking good 

Mines a shit show at the moment. Unpacking everything today finally and playing Tetris.

I’d forgotten in a flurry of “I’ll worry about that later” I chucked all my bags of parts loose into a plastic box and they’re all mixed up now 

BTW there’s about 700 bags 

[mnementh]

   Lots of interesting TE today.     The TEA closet has been cleaned out...

Looks a lot like my old Closet of Doom  back in Houston... 

mnem
As Ifni is my witness, I will have closets again...!

[mnementh]

Looking good  Mines a shit show at the moment. Unpacking everything today finally and playing Tetris. I’d forgotten in a flurry of “I’ll worry about that later” I chucked all my bags of parts loose into a plastic box and they’re all mixed up now 

TETRIS...? As in TEA Glossary TETRIS, or did you finally score another 54645A...?

mnem
We will not speak of my current living room... it is Ground Zero for the moving & sorting going on ATM... 

[bd139]

Glossary Tetris. Only way to fit stuff in here. Managed to get all the consumer electronics boxes and left over PC bits in the top section which is a win.

Right coffee break over. Back to the cupboard of doom.

[Cerebus]

Strange that somone into electronic music doesn't know the difference between a voltage divider and a decade box. 

Probably a drummer at heart. 

Ancient joke:

"How do you know when the drum riser is level?"
"I don't know. How do you know when the drum riser is level?"
"The drummer's drooling evenly out of both corners of his mouth."

[med6753]

Looking good 

Mines a shit show at the moment. Unpacking everything today finally and playing Tetris.

I’d forgotten in a flurry of “I’ll worry about that later” I chucked all my bags of parts loose into a plastic box and they’re all mixed up now 

BTW there’s about 700 bags 

My parts are only sorted by component type (capacitors, resistors, etc) and no accurate inventory so every time I need something I have to physically sort through them. Maybe some day I'll take the time to do an inventory. 

[bd139]

Nah don’t do it. Not worth doing an inventory. It’ll always be wrong.

I just sort mine into boxes by large types and then value or alphabetical order in the boxes. If I haven’t got something it’s quick enough to find that out.

[mansaxel]

Looking good 

Mines a shit show at the moment. Unpacking everything today finally and playing Tetris.

I’d forgotten in a flurry of “I’ll worry about that later” I chucked all my bags of parts loose into a plastic box and they’re all mixed up now 

BTW there’s about 700 bags 

My parts are only sorted by component type (capacitors, resistors, etc) and no accurate inventory so every time I need something I have to physically sort through them. Maybe some day I'll take the time to do an inventory. 

Bit of a pity that the only existing software solution for that involves PHP, because that's not allowed on my computers.

I might get arsed to try to make something about it, but I'll probably build a CLI or curses interface for it instead.

For that 3270 ERP feeling.

In other news, I might have a lead on an actual zSeries machine, but I doubt I will be able to house it, so might need to let it slip.
That's enheavyment par excellence.

[Saskia]

Argl, a z series....

I don't need it, I don't need it, I don't need it ..
The desire is strong in this one ...

[bd139]

Taking on a zSeries is like having two additional children. Walk away 

[Saskia]

No need for me to get pregnant at my age

[Saskia]

Ok, did 10 km today and lugged 8 moving boxes full of books from 3rd floor into the car. Enuff for today

[med6753]

Taking on a zSeries is like having two additional children. Walk away 

Only two children? 

[med6753]

Ok, did 10 km today and lugged 8 moving boxes full of books from 3rd floor into the car. Enuff for today

I'm still hauling boxes/totes to the garage. I do some then rest. Don't need no repeat of the past few months.

But I still have to get the Type 561B out there.  At least it has no CRT or plug-ins but still has that heavy power transformer. 

[capt bullshot]

[...]

The new one uses simple divide-by-four circuits to provide in-phase and quadrature of the 10MHz input signals (now at 2.5MHz). These are fed into simple XOR gates and RC lowpass filtered. You get two triangle waveforms at 90° phase shift, these are input to an AD7793 high-resolution ADC. Do the math and get the frequency difference.
The new unit has three inputs, one reference and two frequencies that are compared to the reference. So I can monitor two devices vs. a third one at the same time (hopefully saving a few of these 24h debug sessions ).

Great results and project in general   
To me it sounds like it is related to a Dual Mixer Time Difference DMTD?
Where the two signals to be measured are mixed with an offset oscillator and the resulting waveforms LPF'd and compared by a counter?
With your contraption they are mixed each with the reference as well? but as I and Q components and the output digitized,
but you have no amplitude difference (it's all square anyway?) so only phase information is what is relevant and used?

Thanks!
I've read some papers about the DMTD (before and while I built the first one).

Anyway, the first one (100MHz PLL) isn't a DMTD, I've had some ideas on my own that I wanted to test:

One 100MHz PLL is offset by a slightly variable 125Hz signal, the other one is just 10x the input frequency. Then it's mixed down to a 125Hz beat frequency. Within the uC, there's another I/Q mixer that mixes the 125Hz beat frequency down to DC. Its output drives a control loop that keeps the phase constant (e.g. Q is alway zero) by varying the 125Hz signal frequency. As a result, the 125Hz offset signal frequency minus 125Hz equals 10 times the frequency difference of the inputs. The problem is the not-so-good 125Hz subtraction stage (by an I/Q mixer) to offset the first 100MHz PLL, it produces quite an amount of noise.
Next iteration will use DMTD principles.

The new one directly compares the input frequencies by mixing them down to DC (all digital). Each input is fed into an divide-by-four circuit with I/Q output. Two XOR gates do the mixing part (one input I / Q, other input I only), after RC filtering in a triangle wave proportional to the phase difference of the input signals. ADC and software measure du/dt of the mixed down signal and calculate frequency difference. Theoretically one needs only the I (or Q) part, but the triangle isn't very linear at the ends (near GND / VCC), so software automatically switches between them, and by having both signals one can derive the sign of the difference. Amplitude is "1" by definition.

Would this approximate a functional diagram? Please excuse the unstandardized symbolical expression.

Yes, this is correct.

Edit: nearly correct, the output is CAN, not USB.

[bd139]

Taking on a zSeries is like having two additional children. Walk away 

Only two children? 

Ok two teenagers?

[BU508A]

Where did you get those prototyping SMD boards? Rademacher? Roth-Elektronik?

[BU508A]

Lots of interesting TE today. 

The TEA closet has been cleaned out. On the left is a now blank area which will accept a scope cart once built. The assorted boxes containing Tek circuit boards and other parts are now in the garage in snap lid totes which should keep the majority of the moisture out. All the parts scopes except one are stacked neatly on the right along with two on the upper shelf with the parts only plug-ins. I had to sacrifice one scope unprotected in the garage. It's the Type 561B I got in PA when I visited nixiefreq. That scope was already full of corrosion and mildew so it won't make much difference. And it gave up it's CRT for the functional Type 561B. But I may at least cover it plastic.

Have you added some silica-gel packets to the boxes? I'd recommend this to get rid of the inital moisture in the boxes and absorb the moisture which will come in newly.
Those boxes are also prone to deforming under load, so they are not as sealed as you may think.

[med6753]

Lots of interesting TE today. 

The TEA closet has been cleaned out. On the left is a now blank area which will accept a scope cart once built. The assorted boxes containing Tek circuit boards and other parts are now in the garage in snap lid totes which should keep the majority of the moisture out. All the parts scopes except one are stacked neatly on the right along with two on the upper shelf with the parts only plug-ins. I had to sacrifice one scope unprotected in the garage. It's the Type 561B I got in PA when I visited nixiefreq. That scope was already full of corrosion and mildew so it won't make much difference. And it gave up it's CRT for the functional Type 561B. But I may at least cover it plastic.

Have you added some silica-gel packets to the boxes? I'd recommend this to get rid of the inital moisture in the boxes and absorb the moisture which will come in newly.
Those boxes are also prone to deforming under load, so they are not as sealed as you may think.

That's an excellent idea and it just so happens that I have a 10 pack of 10 grams each. 

Yes, they do deform under load and I'm being careful when stacking.

[capt bullshot]

Where did you get those prototyping SMD boards? Rademacher? Roth-Elektronik?

This particular one: Scrap box @ work
In general, they're called SMTpads-3U-Thin SP3T-50x50-G by BPS (BusBoard Prototype Systems, www.busboard.com). Usually I order them from mouser.

[mnementh]

mnem's 3DP Corner:

   https://a360.co/3yoCRfS   

This is probably my most time-consuming 3DP project yet; completely reworking the guts of my UltraSabers lightsaber to fix all the cut corners it came with, and upgrade to 18650 power source with BMS and charging circuit. Original Obsidian e-gyro controller only does sound and lighting effects and used 14500 cells in a battery holder. 

(SNIP)



And here's the money shot; this lightsaber has e-gyro controller with flash-on-clash, motion sound effects, fully-programmable sound library, 18650 power source with more than 2X the power density of original, and charge/BMS circuitry the original design never even thought of. As a bonus, it works perfectly with the 8.4V charger I already have for my bicycle headlight.

As usual, full details and 3DP Nerd-Stats can be seen over in the 3DP thread:

https://www.eevblog.com/forum/3d-printing/3d-printer-yet/msg3618846/#msg3618846

mnem
And my son is more excited aboot it than I am...

[mnementh]

Taking on a zSeries is like having two additional children. Walk away 

Funny; that's what my wife said aboot marrying me...  before we even had our kids.

mnem
Aye, I cannae deny it; may as well own it.

[Wolfgang]

Taking on a zSeries is like having two additional children. Walk away 

Only two children? 

Ok two teenagers?

... rarther some very care intensive senior citizens with a lot of progressive deficiencies.
 Sample: you need to talk JCL to me.