Test Equipment Anonymous (TEA) group therapy thread

[worsthorse]

after another couple of hours of, well, wandering around in diagnostic circles, it is time to acknowledge this:



 

while it is clear there are multiple failures that seem to involve signal and bias levels, along with some marginal parts (a) i have no idea what's really wrong, (b) i have run out of ideas as to how to figure out what's wrong, (c) i have neither the skill or knowledge to reverse engineer the circuits in a way that might give me a clue to what is wrong.

my only consolations, and they are tiny indeed.. whoever owned this before me had already done a lot of rework on this part of the board, so the problems are probably not solely related to the 130VDC incident.  and with new rigol signal generator it was easy to generate a good copy of the marker signal and do some injection testing.

i wrote a close out post on my 184 thread. hopefully someone will learn something from me recording my failure in real time.   

[med6753]

Buy another and use as a parts mule.

[WastelandTek]

https://www.ebay.com/itm/Fluke-PM6685R-PM6685R-071-Rubidum-Frequency-Counter-with-1-3GHz-option/174156265852?hash=item288c85b97c:g:g7wAAOSwuxFY2Mj5

I think I need to go turn some tricks

[bitseeker]

Checked out my 53310A a bit today. It's a little dirty and there's some burn-in on the CRT, but still looks good. Man it's bigger than I was expecting. Much wider than my 8560E, but fortunately not as long.

Can it perform an FM deviation measurement? If it does can you show me a pic of that screen? I want a unit that will do that measurement directly.

I'd assume it could, but it's a totally new piece of gear to me. You might call it an educational acquisition. (TEA provides many educational opportunities.) There isn't a screen specifically named that way as you might see on a radio analyzer.

It measures peak-to-peak frequency deviation directly, so you can easily get FM deviation from that.

Can you post some pics (or a short description) how that would be done to see if that'd help xrunner determine if it satisfies his requirement?

Pics a little later, but there is a direct measurement (shift function) on the front panel to measure pk-pk deviation in both modulation domain and histogram modes. Can also be done with markers.

Ah, OK. The manual, which I'm slowly making my way through, is rather terse and light on practical examples. Called the "Operating Reference Manual," it's very strong on the "reference" part.

For example, for Peak-Peak function it says, "This function operates on both 'vs. Time' and histogram data, and on all measurement functions. In all cases, the result is the positive difference between the largest value and smallest value on the measurement axis (within the specified analysis range)."

It's like reading a dictionary.

[kj7e]

RE: HP 53310A FM deviation measurements.

HP 8920A signal gen set to 5.0KHz deviation, HP 53310A measuring 9.933KHz P-P. Tigger says, measuring deviation is what the 53310A does best!

[mnementh]

@worsthorse: Condolences on the patient who didn't survive; congrats on your new addition to the parts pool. 

So you're really a wimp of a dwagon then ! 
Real men would just tear the split bit off, suck the blood for a bit and carry on like nutthing ever happened. 

Those are the same men that the sheep are afraid of. 

Yeah, I used to do that... then I grew up. Really hard to "see with your fingertips" to put that bolt in a hole on the back of a engine, etc when said finger is screaming in pain. Also real hard to please a woman when you can't tell that your hands are tearing great chunks out of her hide. 



Sheeps are 'fraid of dwagons too, but furr a diff'rent reason. 

mnem
Smoke 'em if you got 'em...

[bd139]

Ugh weekend two of Windows 10 upgrade hell. First weekend failure because the damn CPU won’t support Windows 10 fixed with cheap eBay core 2 jobby and now the WiFi card has no fucking drivers for Windows 10. So off to bloody Argos to buy a TP link piece of shit. This is all because the cretin who I am related to, who is incidentally wading knee deep in cash, is too cheap to buy a fucking new PC.

Edit: oh and now he’s going on about getting on WhatsApp for family photos but doesn’t want to spend more than £50 on the problem.  Currently has an ass end Nokia from the dark ages but doesn’t want a smartphone.

I am going to invoice him for this. That will make him fuck off.

Don't forget to add a line item for "consortium consultancy fee" since you brought the issue here; this should equal ~300% of the value of the entire ticket. 

mnem
*Certifiable Committee Commtitted*

£700 invoice will be issued    I don’t expect payment just the point to be got

Well an interesting thing happened. The USB ports were so slow they were unusable. Turns out the first generation USB Microsoft natural keyboard fucks the entire USB bus to 1.0 if plugged in. It has two heads so got rammed in the PS2 port instead now all the USB2 ports work. Never seen that before!

You didn't know that? I thought it was common knowledge. The USB will always default to the slowest standard device plugged into it.

Edit....you were lucky the keyboard also had a PS/2 type connector.....as well as the computer.

It’s usually more complicated than that. The thing has four separate USB controllers and data paths in it and any downgrading should only happen “per hub”. So all four controllers were downgraded to 1.0 when a USB 1.0 keyboard was plugged in rather than just on that hub.  Smells like a BIOS or windows bug to me.

Lenovo M58e for reference.

I won in the end!

[bd139]

Ugh weekend two of Windows 10 upgrade hell. First weekend failure because the damn CPU won’t support Windows 10 fixed with cheap eBay core 2 jobby and now the WiFi card has no fucking drivers for Windows 10. So off to bloody Argos to buy a TP link piece of shit. This is all because the cretin who I am related to, who is incidentally wading knee deep in cash, is too cheap to buy a fucking new PC.

Edit: oh and now he’s going on about getting on WhatsApp for family photos but doesn’t want to spend more than £50 on the problem.  Currently has an ass end Nokia from the dark ages but doesn’t want a smartphone.

I am going to invoice him for this. That will make him fuck off.

W10 on a 2 core ? 
Who thought that might be a good idea ? 

Wouldn't touch that one with a 40' barge pole !

Runs fine. I’ve been running a business off an ass end i5 for the last 3 years 

[mnementh]

Now that you've beaten it into submission... time to put it out of its misery with a TASER across the ATX & SATA ports.  Core2 Duo is literally XP generation hardware; let the fucking thing die with dignity.

In contrast, the $8 i3 HP p7-1243 was a absolutely painless upgrade; like 10 minutes to install and then allow it to do its updates overnight. When I re-installed Win10 with a cheap SSD in it, took like 5 min then another 30 minute for updates and all because I was at the bench doing other stuff and could tell it to reboot when needed.

mnem
 

[bitseeker]

RE: HP 53310A FM deviation measurements.

HP 8920A signal gen set to 5.0KHz deviation, HP 53310A measuring 9.933KHz P-P. Tigger says, measuring deviation is what the 53310A does best!

Thanks! Using the new DG831, I generated a 1MHz carrier, 1V p-p sine with a 100Hz modulation frequency and 1kHz deviation.

The 55310A shows Pk-Pk as 2kHz.

Pic is 'vs. Time' mode as compared to the histogram in yours.

xrunner, are these examples helpful?

[bd139]

Now that you've beaten it into submission... time to put it out of its misery with a TASER across the ATX & SATA ports.  Core2 Duo is literally XP generation hardware; let the fucking thing die with dignity.

In contrast, the $8 i3 HP p7-1243 was a absolutely painless upgrade; like 10 minutes to install and then allow it to do its updates overnight. When I re-installed Win10 with a cheap SSD in it, took like 5 min then another 30 minute for updates and all because I was at the bench doing other stuff and could tell it to reboot when needed.

mnem
 

It has “designed for vista” sticker on it so close  .

Yeah I can get a better PC for £40 here but the fucking owner is a total miser. Literally complained about having to replace the damn WiFi card with a £10 TP link USB piece of shit because the old one had no windows 10 drivers.

Next time:

What do you do for a living bd139?

Ah I’m a rent boy!

Conversation and obligations all ended, favours never asked!

[Cerebus]

If it's a really long split like this one, apply a little wad of cotton from a Q-tip about the size of a match-head to the wet CA; add another drop and press the excess out with a flat of a knife. The fibers will reinforce the CA enough that the nail can grow out before the CA "blowout patch" fails.

Those without experience of this beware - the reaction between cotton and cyanoacrylate can be highly exothermic. It's one thing for it to get hot, it's another for it to get too hot when it's firmly glued to your person!

Fibreglass tissue is a better choice than cotton, if you have some handy. (I keep small quantities of fibreglass tissue, carbon fibre fabric and Kevlar felt in my 'gluing things' box precisely so that I do have such things handy. Small quantities can be picked up for a very little money from the EvilBay and the usual scumbags.)

[bitseeker]

kj7e, Oculus, etc.: On the 53310A, I don't get anything when connecting the A input, only the B input. However, it measures the B input regardless which input is selected for analysis. I'm not quite sure if that's normal. It's not the Common/Separate functionality for time interval analysis, is it? I'm just doing FM deviation measurement.

[Cerebus]

Next time:

What do you do for a living bd139?

Ah I’m a rent boy!

Conversation and obligations all ended, favours never asked!

You hope.... 

[kj7e]

kj7e, Oculus, etc.: On the 53310A, I don't get anything when connecting the A input, only the B input. However, it measures the B input regardless which input is selected for analysis. I'm not quite sure if that's normal. It's not the Common/Separate functionality for time interval analysis, is it? I'm just doing FM deviation measurement.

The input is changed under the Function settings for A, B and C inputs.  I suspect your using the B input only.  You can also configure each input independently in the Configure Input menu for AC/DC coupling, input impedance, and then input voltage threshold which is very important when making measurements.   The Auto Scale will not select the correct input, but it will find the 50% threshold for the input that's selected. 

If you see INPUT above "A B C Ext" this is just to configure the inputs, not to select which to use.  You will see FUNCTION MENU at the bottom, hit that and then you will see FUNCTION above "A B C"  this is where you can select which input to use.

[bitseeker]

Turned off and started over.

If I have the cable plugged into A, I get nothing regardless which input (A, B or C) I select from the Function/Input menu. The A jack's trigger LED doesn't light up. If I plug the cable into B, then samples only appear when A is selected. Selecting B doesn't show anything.

Going to check trigger settings. Says trigger settings not available in Fast Hist mode.

[kj7e]

Turned off and started over.

If I have the cable plugged into A, I get nothing regardless which input (A, B or C) I select from the Function/Input menu. The A jack's trigger LED doesn't light up. If I plug the cable into B, then samples only appear when A is selected. Selecting B doesn't show anything.

Going to check trigger settings. Says trigger settings not available in Fast Hist mode.

Selecting the input can be confusing at first, I just had to refresh myself how to do it haha.   See my updated post above, if that does not work, I can take a video and show you.

[worsthorse]

Buy another and use as a parts mule.

I am looking at one now because I never learn. *sigh*

I may also put my attention on the shelf full of yet-to-tested TM5xx plugins, one of which is a time mark generator. It really makes more sense for me to get that gear working, as I can fit four pieces of cal gear in the space of one box full of 184. But I don't know... the 184 is really freaking cool and I am a glutton for punishment.

BTW, it turns out I don't know the guy from whom you bought the CA but I am probably going to meet him next week, at least if I decide to get another 184.

[bitseeker]

If you see INPUT above "A B C Ext" this is just to configure the inputs, not to select which to use.  You will FUNCTION MENU at the bottom, hit that and then you will see FUNCTION above "A B C"  this is where you can select which input to use.

Yep, that's the one. It is confusing since it looks similar in the two screens. I'm on the Function screen and selecting A is the only way I get samples show up when B is connected. Selecting B or C yields no samples on the B input connector.

When A is connected, I get no input and no trigger LED regardless which input I select on the Function screen. Maybe the A connector is dead. I haven't calibrated the gain on any of the channels, so this is the first try with putting signals into the thing.

[kj7e]

If you see INPUT above "A B C Ext" this is just to configure the inputs, not to select which to use.  You will FUNCTION MENU at the bottom, hit that and then you will see FUNCTION above "A B C"  this is where you can select which input to use.

Yep, that's the one. It is confusing since it looks similar in the two screens. I'm on the Function screen and selecting A is the only way I get samples show up when B is connected. Selecting B or C yields no samples on the B input connector.

When A is connected, I get no input and no trigger LED regardless which input I select on the Function screen. Maybe the A connector is dead. I haven't calibrated the gain on any of the channels, so this is the first try with putting signals into the thing.

Do this,

Power the unit on while holding any key.  This will do a clear of the self cal mem.  Then run the self cal from the utility menu.  Just need to have a good steady accurate 5v DC source to feed to channel A and B.  Will only take 5 min start to finish.

[Specmaster]

Now that you've beaten it into submission... time to put it out of its misery with a TASER across the ATX & SATA ports.  Core2 Duo is literally XP generation hardware; let the fucking thing die with dignity.

In contrast, the $8 i3 HP p7-1243 was a absolutely painless upgrade; like 10 minutes to install and then allow it to do its updates overnight. When I re-installed Win10 with a cheap SSD in it, took like 5 min then another 30 minute for updates and all because I was at the bench doing other stuff and could tell it to reboot when needed.

mnem
 

It has “designed for vista” sticker on it so close  .

Yeah I can get a better PC for £40 here but the fucking owner is a total miser. Literally complained about having to replace the damn WiFi card with a £10 TP link USB piece of shit because the old one had no windows 10 drivers.

Next time:

What do you do for a living bd139?

Ah I’m a rent boy!

Conversation and obligations all ended, favours never asked!

How much RAM has it got? 4GB is the max on that motherboard and W10 works better with 6 to 8GB as the minimum, I'm running 32GB and sometimes it can shit slow on that depending on what else I have open.

[bitseeker]

Do this,

Power the unit on while holding any key.  This will do a clear of the self cal mem.  Then run the self cal from the utility menu.  Just need to have a good steady accurate 5v DC source to feed to channel A and B.  Will only take 5 min start to finish.

Will do. Found a BNC-to-alligator cable to hook up to the HP 6114A. That should do the trick.

[bd139]

Now that you've beaten it into submission... time to put it out of its misery with a TASER across the ATX & SATA ports.  Core2 Duo is literally XP generation hardware; let the fucking thing die with dignity.

In contrast, the $8 i3 HP p7-1243 was a absolutely painless upgrade; like 10 minutes to install and then allow it to do its updates overnight. When I re-installed Win10 with a cheap SSD in it, took like 5 min then another 30 minute for updates and all because I was at the bench doing other stuff and could tell it to reboot when needed.

mnem
 

It has “designed for vista” sticker on it so close  .

Yeah I can get a better PC for £40 here but the fucking owner is a total miser. Literally complained about having to replace the damn WiFi card with a £10 TP link USB piece of shit because the old one had no windows 10 drivers.

Next time:

What do you do for a living bd139?

Ah I’m a rent boy!

Conversation and obligations all ended, favours never asked!

How much RAM has it got? 4GB is the max on that motherboard and W10 works better with 6 to 8GB as the minimum, I'm running 32GB and sometimes it can shit slow on that depending on what else I have open.

4Gb maxed out in that. Classic BSPM (basic spec porn machine) build ... mine has 32gb. Basic spec for software dev box. My old laptop had 4Gb of Ram but that was a portable terminal for AWS and i wanted to keep the juice use low.

[bitseeker]

Do this,

Power the unit on while holding any key.  This will do a clear of the self cal mem.  Then run the self cal from the utility menu.  Just need to have a good steady accurate 5v DC source to feed to channel A and B.  Will only take 5 min start to finish.

Starting with Offset and Gain at default 32 for both A and B. DC source at 5.0009V.

Offset calibration was fine. Same DAC values as before. A=21, B=22.

Gain calibration results in "Dac=64", then "Check failed" or "Calibration failed" depending which one I run. Same for A and B inputs.

[kj7e]

PM me your number, easier to do this on the phone.