Test Equipment Anonymous (TEA) group therapy thread

[Saskia]

they cancelled the order, marked it as shipped (to where ? my old address no longer exists), I don't have a tracking number with FedEx to reach out, and they did not refund the 350€.

And their branch office cannot be reached. I am pissed.

[tggzzz]

Power Designs PD2005A on German ebay:

https://www.ebay.de/itm/255327604587?hash=item3b72b60f6b:g:IjgAAOSwup9gHCiI

Check the mains voltage and schematic.

My 2020 can be changed to 240V by resoldering the connections to the transformer primary. IIRC the 2005 can't.

[mansaxel]

Power Designs PD2005A on German ebay:

https://www.ebay.de/itm/255327604587?hash=item3b72b60f6b:g:IjgAAOSwup9gHCiI

Check the mains voltage and schematic.

My 2020 can be changed to 240V by resoldering the connections to the transformer primary. IIRC the 2005 can't.

I've asked the seller. I DO have a soft spot for the 2005A, so I'm maybe not as steadfast as I ought to be.  (I have a pile of supplies that need fixin', of course. ) Will report back.

[PA0PBZ]

It looks like it has been converted

[tggzzz]

I've asked the seller. I DO have a soft spot for the 2005A, so I'm maybe not as steadfast as I ought to be.

I feel similarly about my 2020B. Seeing the lm399 inside is pleasant, and being able to tweak the supply by 10µV (not stable!) is amusing.

[mansaxel]

It looks like it has been converted

Good spotting. Question is whether it is factory or optimist...

Edit:

The service manual does neither mention field modification, nor from-factory voltage option.

[BU508A]

It looks like it has been converted

My 2005A is running on 230VAC as well:

https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg3496502/#msg3496502

[m k]

As for the DM8577, I'm avoiding ancient Nat Semi/TI parts, the Soviet clones should be a bit newer. Strange how the Nat Semi parts on ePay-UK are marked as both DM8577N & DM74188AN if they can't be programmed the same way. 

David

I was nearly caught out by that.   The ones I have are factory marked DM74188AN   DM8577N.   I was initially using the TI Databook.  I have one around here somewhere and it was easier to find on the web.   Mostly because I was curious, I did an extensive search and found the DM8577N in a National book.   Much to my horror, they are "Pin Compatible"  but very different internally.

They are awful, they even put the 74188 first, wonder how many got wasted by folks programming them as 74188 instead of 8577 over the years. 

David

Why not 74s188(10.5V) or 82s23(17.5V)?

NS manual for DM8577 is a bit uncertain but 23V is mentioned.
The procedure is for DM8574 and DM8577 pages are stuffed in between.
(169.pdf)

BTW,
Took a moment with those range relays.
(Vcc is GND2 and GND is GND4)
(where GND1-3 are the same and GND4 is -5V)

[Vince]

....snip....

I would have made a close-up picture of the screen to prove I am not lying about frequency and signal amplitude. Sadly just before doing that, I tried fiddling with the adapter to see if I could shove it in a bit more to make sure the trace does not go unstable right as I am trying to photograph it and... as I was grabbing the adapter, I got ZAPPED !!! Yes this generator tried to kill me !!! 
So no close-up picture for you sorry... instead I immediately pulled the plug !

Checked with the DMM for resistance between the two line terminals in the power socket, and chassis/enclosure.... no short, got 5Mohms and slowly climbing.

So I am not touching this generator for now ! 

Think I will retrofit an IEC power socket hidden at the back (to not ruin the looks of the face plate), and get myself a life saving earth connection....

Will look at the schematic of the generator to see if there is maybe some failing cap that could explain the mishap...

Catching up still......

Looking at your previous posts on the Metrix 920, it seems to have two capacitors to ground (aka the case) and looking at the back of the mains inlet there are two dubious looking capacitors on the sides of the inductors, I would be amazed if these weren't wax paper deathtraps/smoke bombs.
And the DMM will be of no use here, you need a proper DC insulation tester at 250V or 500V to check them, but probably best to just change them for some class Y rated capacitors anyway.





David

Thanks for doing my work Dave ! 

Just looked C11 and C12 in the parts list... they are indeed paper caps ! They are 10nF 10%. 

Will replace them of course !

Will keep the old caps so I can test them later for fun and educational purposes...

Hell I could cobble a crude tester together easily I guess !

After all these caps are simply subject to mains AC ! So all I need to test them in a relevant manner, is to connect the to mains and measure the leakage current, eh ?

Just need to put a series resistor to limit the current to a safe value, I don't know, less than 500mA, since my Metrix meter has a 500mA current range / fuse.

Will open up the generator in the coming days, if I can sort out this problem right now, why wait... there are enough things to do and fix in this lab for me not to post pone repairs I can do quickly and easily !

I'm suprised no one has said this: YOU NEED TO CHECK THE EARTH CONNECTIONS  . Even if the capactors are leaky you would not have got a shock if the units mains earth was intact (all the way back to the supply, it could be a faulty extension lead or badly wired socket).
Even if the capacitors are OK if there is no earth the chassis will be at around 120V AC. This is because the two capacitors act as a AC voltage divider. Not much current due to the low values but enough to feel if you touch it or read on a DMM to ground.
Note that replacement capacitors should be Y type rated for mains use.

Rob I don't understand what you mean ? These old Metrix as you can see on the pictures, have a two pronged power cord.  There is no earth connection to speak of 

The " earth " looking symbol on the schematic refers to chassis ground only. The front/face plate of the instrument is used as a ground plane. So these two caps connect the two mains wire to the enclosure of the instrument...

That's why earlier I was mentioning that I might want to retro-fit a 3 pronged cable with an IEC socket hidden at the back of these units, for extra safety.

[mansaxel]

Got a reply regarding the 2005: "Die Netzteile sind für 230V geeignet." So, Real Mains operation.

[Vince]

Mine is too but there is no service manual for it anywhere  ! Contacted Metrix directly and they said "yeah we remember that one, but have no info on it as we outsourced its R&D, we didn't design it ourselves ". Bummer 

The service manual for the GX5000 is proving to be very elusive! Metrix were (perhaps still are) partnered with Sefram. Sefram appear to have sold the same unit as Model 4450.

Asking for the unit's ID, via the GPIB, returns "OR-X,GX   5000". I believe Or-X are the original manufacturers. They offer the model 550 (sometimes called FP550) which appears to be the same design.

The Metrix 555 appears to have the same functionality but with an upgraded front-panel. The user manual for the 555 details the GPIB commands - all of which seem to work with the GX5000.

Wow thanks for that  !! 

I did wonder if it would be possible to find the OEM hoping he would have a manual for this thing, but the design / looks of the instrument looked like nothing else I have ever seen, so I had no idea what manufacturer to search for ! 

.. then you came !

So made probably by OR-X, and also sold by SEFRAM.   No I don't recall SEFRAM having anything to do with Metrix (in general at least), and youtr GPIB ID ping looks quite telling doesn't it... was made this mysterious sounding " Or-X " company ! Looking at the bottom of their spec sheet that you provided, it says they are based in ISRAEL ?!     Eh ?!  Why not...


Have you tried asking Sefram or OR-X directly, for a service manual then ??

[Andy Watson]

Have you tried asking Sefram or OR-X directly, for a service manual then ??

No, I've not got around to asking either company yet.  I suspect that Sefram are just re-badgers and will give the same answer as Metrix.

[Vince]

Will try contact them both. Yes I agree SEFRAM must be rebadging too. Just like Metrix they didn't have the R&D fire power to design such a nice instrument, let's be real.

[Neomys Sapiens]

Power Designs PD2005A on German ebay:

https://www.ebay.de/itm/255327604587?hash=item3b72b60f6b:g:IjgAAOSwup9gHCiI

Nah, I don't think it's worth it 

That's the better one. He had one whose specs were worse, for that he used a Brymen DMM! 

[Vince]

However I have a new problem now... looks like my Metrix Nixie counter has a problem with its LF / 1Meg input ! Signal is only 10MHz so it can take it no problem. However it would not register / count anything at all, only zeros on the Nixies, even though I have good signal amplitude, 5Vpp unloaded, and no matter how I fiddled with the input "sensitivty" pot/control. Plus, when I connect the counter, I can see on the scope that the signal goes haywire. Big drop in amplitude + it becomes completely out of shape, looking like a wiggly triangle, no more a rectangular pulse at all !  If I use the HF / 50R input instead, all is well.
So, the LF front-end has a problem.... some more repairs in the queue !
This counter is not easy to work on I fear, very compact in there. It's not going to be fun I think...   

I would try another signal source, as maybe it can't detect the narrow pulse on the low frequency input? , I know some of the old counters I've got can be a bit fussy on the input signal.

David

OK I spent a few minutes testing the counter.

Result : the /F input CAN go up to 10MHz no problem... IF the conditions are right.

Let me explain... First I connected the pulse generator only to the scope, 10MHz again. Got a fair bit of ringing. Switched the scope to 50R input... much better, looks decent now (see pic)

Have my baseline now, so I can now connect the counter as well as see what happens.


1) if I connect the HF / 50R input of the counter, it adds a little bit of ringing but minimal, still looks decent and can count just fine. (see pic)

2) Then I switch the counter to its LF / 1Meg input.  Oh oh... it adds A LOT of ringing, so much that if the pulse is too narrow, it becomes so distorted that it hardly looks like a square anymore. That's when the counter fails to count. But if you increase the pulse width a little bit, at some point the ripple becomes low enough that the counter manages to pick up the pulses.. but it still looks far from a square (see pic).

So the question is why does the LF input distort the generator output SO much  ?! 

Of course it adds more capacitance, but one would thing it would dampen the signal not make it "ring" ?

Or... the rsie time is too high, too much high-frequency harmonics, and the capacitive load of the 1Meg input low-passes the signal so much that it kills too much of the spectrum / harmonics... and the little that remains... is not enough to make it look anywhere near a square wave.

SO.... I could make one more experiment : keep the same frequency and pulse wigth, but progressively decrese rise time and see if that make the signal look more and more like a square.

Yes that would be interesting ! 

[Neomys Sapiens]

DMMCheck with LCR option ordered. I did forego getting the ISO calibration, as I think this money is better spent later during it's lifetime. Or I'll have my company do it.

[ch_scr]

(...)
OK I spent a few minutes testing the counter.

Result : the /F input CAN go up to 10MHz no problem... IF the conditions are right.

Let me explain... First I connected the pulse generator only to the scope, 10MHz again. Got a fair bit of ringing. Switched the scope to 50R input... much better, looks decent now (see pic)

Have my baseline now, so I can now connect the counter as well as see what happens.


1) if I connect the HF / 50R input of the counter, it adds a little bit of ringing but minimal, still looks decent and can count just fine. (see pic)

2) Then I switch the counter to its LF / 1Meg input.  Oh oh... it adds A LOT of ringing, so much that if the pulse is too narrow, it becomes so distorted that it hardly looks like a square anymore. That's when the counter fails to count. But if you increase the pulse width a little bit, at some point the ripple becomes low enough that the counter manages to pick up the pulses.. but it still looks far from a square (see pic).

So the question is why does the LF input distort the generator output SO much  ?! 

Of course it adds more capacitance, but one would thing it would dampen the signal not make it "ring" ?

Or... the rsie time is too high, too much high-frequency harmonics, and the capacitive load of the 1Meg input low-passes the signal so much that it kills too much of the spectrum / harmonics... and the little that remains... is not enough to make it look anywhere near a square wave.

SO.... I could make one more experiment : keep the same frequency and pulse wigth, but progressively decrese rise time and see if that make the signal look more and more like a square.

Yes that would be interesting ! 

The LF input is not terminated into 50 , hence why the "ringing" / reflection on the cable / transmission line gets worse!
See this video for a nice (& literally old school) demonstration:

[Vince]

Yeah but the scope is 50R terminated so I didn't think that putting 1Meg in  // with the counter, would make that much of a difference.. 1Meg // with 50R is still pretty much 50R 

Played with rise times. That did it, was interesting.

Since the beginning the pusle generator was set to its fastest rise time, 5ns.

So what I did was increase the puls width until the 1Meg input would regsiter. It happened with a pulse width of 27ns.

Then I backed off a little, 26ns, so the counter would not register anymore.

Then I started increasing the rise time (and fall time equally)  to see how / if it would have an impact on the shape of the waveform, and it the counter would behave any differently.

That was success.

I increased rise time very progressively, in 0.1ns increments. At 7.2ns the counter started flickering, was picking up some pulse.. and at 7.4ns, it was able to read reliably, was stable.

The shape of the signal however is barely changed to thje naked eye.. but clearly the counter does see the difference !
It was like clock work : redude pulse width by one ns : doesn't count anymore. Increase by 1ns.. hey presto works again. Same with rise time. 7.4ns works solidly, 7.2 it's unstable, 7.0ns no counting at all.

It's fun being to control pulse width and rise times, accurately, and see how the counter reacts while looking at the waveform on the scope 

[ch_scr]

Yeah but the scope is 50R terminated so I didn't think that putting 1Meg in  // with the counter, would make that much of a difference.. 1Meg // with 50R is still pretty much 50R 

Played with rise times. That did it, was interesting.

Since the beginning the pusle generator was set to its fastest rise time, 5ns.

So what I did was increase the puls width until the 1Meg input would regsiter. It happened with a pulse width of 27ns.

Then I backed off a little, 26ns, so the counter would not register anymore.

Then I started increasing the rise time (and fall time equally)  to see how / if it would have an impact on the shape of the waveform, and it the counter would behave any differently.

That was success.

I increased rise time very progressively, in 0.1ns increments. At 7.2ns the counter started flickering, was picking up some pulse.. and at 7.4ns, it was able to read reliably, was stable.

The shape of the signal however is barely changed to thje naked eye.. but clearly the counter does see the difference !
It was like clock work : redude pulse width by one ns : doesn't count anymore. Increase by 1ns.. hey presto works again. Same with rise time. 7.4ns works solidly, 7.2 it's unstable, 7.0ns no counting at all.

It's fun being to control pulse width and rise times, accurately, and see how the counter reacts while looking at the waveform on the scope 

Yes, but every cable is a transmission line. The question is: where is the T-piece? If you have the T-piece on generator or scope, the cable between generator and scope is terminated (by scope) but the other end is "flapping in the breeze". If that is the case, put the T-Piece on the counter: Now the whole line from generator with the counter in the middle is terminated by the scope on the other end.

[mnementh]

An old school buddy asked me about getting his 80's Marantz ST450 AM/FM tuner fixed. Anyone here know about these old tuners ?

The usual issues with old VFDs, dried out caps, and buttons going flaky.

IIRC, these are analog/VCO tuning with digital display derived from the VCO. Refurb should be a pretty humdrum affair, as long as the tuner string is intact and not gone stiff/flaky.

Flip side of that is analog tuner feature-set, so not a lot of demand; I see a working one on eBay Klumphzinger for 59€ :

   
https://www.ebay-kleinanzeigen.de/s-anzeige/marantz-st-450-hifi-tuner/1993128648-172-16495


hifiengine has UM & SM, but you have to register for a free account to download:

https://www.hifiengine.com/manual_library/marantz/st450.shtml

Cheers,

mnem
 

[mnementh]

https://youtu.be/31kgOCo2WhA   Hmmm... it sure does like to sing while it works, doesn't it...? 

mnem
 

I recently updated the controller board on my 3D printer. It now uses a Creality 4.2.7 board which has TMC2225 motor drivers: https://creality3d.shop/products/creality3d-upgrade-silent-4-2-7-1-1-5-mainboard-for-ender-3-ender-3-pro-ender-5-3d-printer
The printer went from being a "singer" to being incredibly quiet.

McBryce.

Yeah, I've had this printer for 4 years; a lot has changed in 3DP since I last had hands on it. Both my newer printers have Trinamic drivers; the difference in noise and finished surface quality is pretty amazing. OTOH, as this is now a 3rd printer, and will likely be only used where I need its large build volume, not a lot of motivation to fuck with it as it is working quite nicely for a CR-10 clone of its generation. 

Most likely once I get a really rigid bench/table/crate to set it up on, I'll put a 24" x 24" flagstone under it and ladder bars for the top frame and call it good.

Maybe upgrade to a kit of these for it (tho not likely for a while at $20 per each) :



https://www.cnckitchen.com/blog/bondtech-cht-high-flow-nozzle-reviewed

Thanks for enabling my addiction, ch_scr...

mnem

[mansaxel]

Flip side of that is analog tuner feature-set, so not a lot of demand...

The Marantz FM tuners of old aren't bad. If the sound and the ability to dig signals out of noise (while dealing with deafening commercial stations a bit too close) counts for something, they clearly are worth the job.  Here, we've just revived the Model 4230 Quad-receiver my father bought back in 1976, but now in Middle Boy's room, together with the AR-6 loudspeakers that were part of the same original deal, and the Thorens TD166 Mk V I bought in the 90s.  Very Classy.

There are a few scale illumination lamps out, and the scale is some 350-500KHz off, pots are a bit scratchy, all as can be expected from a 45yo consumer device, albeit a very high quality one.

I have calculated and made a simple dipole which gives full signal level on all relevant stations, so I'm guessing the FM tuner is basically OK. (but we've got excellent RF reception here, on a hill, practically line-of-sight to the main Nacka transmitter. OTOH, I can get the regional FM stations from all around lake Mälaren with my circular dipole that's feeding the rest of the FM reception units at home)

Service manual has been downloaded, and schematic of course was supplied with the unit so I now have a few reasons to learn to use all the measurement things I have.

[Andrew_Debbie]

Why not 74s188(10.5V) or 82s23(17.5V)?

NS manual for DM8577 is a bit uncertain but 23V is mentioned.
The procedure is for DM8574 and DM8577 pages are stuffed in between.
(169.pdf)

At the time, I wasn't able to find a genuine 74188 or 74S188.     There are some on eBay now, but I'm too far down the NS DM8577 rabbit hole.



The DM8577  programming:    See 1976 National TTL databook  - Page 3-101  to 3-103.

Programming sets a bit to low.   All bits are highin an unused device.   Program one bit at a time.

- Vcc 5V -- Address byte to program.   Set enable line high,  (All outputs are high when Enable is high)   20-22V pulse on the output line to program.    1µS-10µS rise time, 20mS pulse width.   (10-50mS range, with 20mS typ)



74S188 Programming -    See TI 1977 Bipolar Microcomputer Components Data book for Design Engineers     Pages 4-1 to 4-3

All bits logic lowin unused device. Programming sets a bit high.

Vcc - 5V.  Address word to program.   Set enable high.     Bit to be programmed to 0V - 0.3V.     All other bits pulled to 5V by 3.9KΩ resistors.    Apply 10.5V 100µS program pulse to Vcc.   Pulse take chip select.    There are rise and fall times for the Vcc pulse too.


I've got 90% of a design for the TI parts.   I dropped it when I saw the NS data book.

[mnementh]

Flip side of that is analog tuner feature-set, so not a lot of demand...

The Marantz FM tuners of old aren't bad. If the sound and the ability to dig signals out of noise (while dealing with deafening commercial stations a bit too close) counts for something, they clearly are worth the job...

Well, no... my point was not that they were a bad tuner; more that the street value of a working ST450 means you can't afford to spend a lot of time trying to resurrect one that's got major problems. Better to shotgun recap a working one and cut your losses. 

mnem
 

[Specmaster]

Flip side of that is analog tuner feature-set, so not a lot of demand...

The Marantz FM tuners of old aren't bad. If the sound and the ability to dig signals out of noise (while dealing with deafening commercial stations a bit too close) counts for something, they clearly are worth the job.  Here, we've just revived the Model 4230 Quad-receiver my father bought back in 1976, but now in Middle Boy's room, together with the AR-6 loudspeakers that were part of the same original deal, and the Thorens TD166 Mk V I bought in the 90s.  Very Classy.

There are a few scale illumination lamps out, and the scale is some 350-500KHz off, pots are a bit scratchy, all as can be expected from a 45yo consumer device, albeit a very high quality one.

I have calculated and made a simple dipole which gives full signal level on all relevant stations, so I'm guessing the FM tuner is basically OK. (but we've got excellent RF reception here, on a hill, practically line-of-sight to the main Nacka transmitter. OTOH, I can get the regional FM stations from all around lake Mälaren with my circular dipole that's feeding the rest of the FM reception units at home)

Service manual has been downloaded, and schematic of course was supplied with the unit so I now have a few reasons to learn to use all the measurement things I have.

Well, no... my point was not that they were a bad tuner; more that the street value of a working ST450 means one can't afford to spend a LOT of time trying to resurrect one that's got major problems. Better to shotgun recap a working one and cut your losses. 

mnem
 

Agreed, Marantz gear was very good quality, but these days everyone only wants DAB sets, and while FM is still a perfectly good platform, people just do not want it , partly because of their sheer size, more modern units are far smaller, but the audio quality could be debatable.