PM3055 oscilloscope PSU repair (didn't work) ... (edit: got it!!!)

[Atheus]

Hi all,

Time for a new thread on this. Here's the old one...

https://www.eevblog.com/forum/beginners/scope-probes-have-no-ground-connector!/msg1175722/#msg1175722

.. but I wouldn't bother.

<backstory>
Basically I bought a "for parts" Fluke/Phillips (hopefully mostly made by Fluke!) PM3055 with '94 date code for £50. Even though it was sold for parts the photos showed a trace on the screen... I couldn't see how it could be completely non-functional (some £50 old scopes don't even power on) with that recent photo. I did see cosmetic damage, but it was just dirt, plus the plastic over/under the auto-set button was broken and this button was unusable.

I bought it anyway. The small area of broken plastic was removed and the auto switch was fully functional once repaired. Some duck tape took care of the hole in the plastic.

Now the problem was I had no probes - no working ones anyway. The fairly clueless seller had not included the ground clips with the two probes he had (one Phillips and one Hameg x10) so I bought a couple of £10 Chinese ones just for those. Couldn't find that little clip connector separately elsewhere or I'd have made my own.

Anyway, with all that sorted, I powered on and was able to get something resembling the square wave from the calibration tab. Yesss!!! What a score for £50!
</backstory>

Now, as you might know or have read, there is a known problem with this unit... at least older models. The PSU has some poly caps which go bang apparently. Obviously I needed to remove the PSU to check if I had bad parts and replace them if needed. The PSU was extremely difficult to remove contrary to most instructions old and new. I quote the service manual exactly:

"Push both lips which secure the power supply unit sideways and gently lift this unit out of the instrument"

... sideways???

It is also noted that one does not have to remove the bottom cover to remove the PSU. This instruction may have been correct the day it was made, but now it is solidly held in by two plastic clips. Again it is specifically noted one does not have to touch these on the bottom side - of course each bit of plastic is held to the main case by two lips - and you definitely do. Before I tried this method I was applying such force to the PCB that I fear I might have ripped it had I pulled it harder.

The actual procedure (now) is to pull those two clips backwards and sideways (whatever those directions mean to you - away from the PCB, and away from vertical components. Basically the only two ways they will bend). Now find some way of securing either these or the board itself and start a wiggling/pulling process to clear it JUST enough that you can press the two bottom clips inwards (towards the unit, yes with the cover removed...) which will simply snap the clips free and the board with it. only NOW can you 'gently lift the unit out of the instrument' and it still takes force.

I write all this in detail because it has taken me 3 days (only a few hours each night but still...) to do this and I could find no decent description of this procedure online. Even the video teardowns seem to miss the bit where they actually remove the PSU, and questions about how to remove it on forums usually end in "I've given up" or "I got it!" without explanation. Hopefully this post contains the right SEO keywords to turn up the next time someone has this problem.

Pic of freed PSU:



Note the cabe tie and loop of 18 AWG wire used 'wiggle' it out until the bottom tabs are pressed as below:



See the circled thing? That's the PSU board. It prevents the little tab being pushed. The board must (on my scope anyway which as never AFAIK been previously opened) be wiggled/pulled/shoved etc (while manipulating the top tabs) until the 2-3 mm clearance is achieved which is needed to press the bottom tabs. Once the bottom tabs are pressed the holders just fall away from the case. The PSU can now be removed from the case with significant force. To illustrate, I have never broken a cable tie before, even intentionally. I broke four on this board trying to pull it out; hence the use of metal wire.

TLDR; I bought a '90s 'scope and am repairing it. I'd like others to know the procedure as there seems to be nothing helpful existing on the web.

More to come...

~Atheus

[Atheus]

Part Two: capacitors...

I am told variously that all 'RIFA' caps, and/or all 'X2' caps should be removed and replaced. Pics below. You can see I have one cap marked X2 and RIFA, and two simply marked 100pf and RIFA. I intend to replace all 3 because they all show signs of strange damage (ever seen anything like this?) and they're only a few pence.

Here are pics of the offending components from all relevant sides:





Note the weird cracks etc... I am reluctant to se this scope now until these dodgy things are replaced.

I understand the X2 cap needs to be replaced with another X2 cap - I will teach myself about the nature of X and Y caps ASAP but I want to get these ordered - but the other two? They are not X or Y? Just regular 1000pf metal poly? Or can one of you see a mark I cant's see/identify?

Cheers!

~Atheus

[Atheus]

I am about to order:

- 5x metal polypropylene .22uF X2 with 22mm pitch
- 10x metal polyester 1000pF/1nF no X or Y rating (as part of a kit) with unkown pitch (but I can bodge it)

I screwed up my previous buy. Ended up with Y2 class caps which I understand will not work... thing is... that value (1000pF/1nF) is only available in Y2 rating on my go-to site as a one-off. No other metal poly. Plenty of ceramics etc at that value but they won't work... right? What about polyester rather than polypropylene?

~Atheus

/edit: is there any chance the 'Y2' 1000pF caps would do for a replacement for the existing ones? The existing ar marked 'CX1' and my spares are marked 'Y2/X1'... clutching at straws I know...

~A

[capt bullshot]

These capacitors are for EMC "only". The scope will work without them, having some elevated level of noise emitting through the mains line ... who cares if no one is disturbed by that?
So if you replace them, any technology (Ceramic, any kind of film / paper) will do as long as you get the correct capacitance / safety rating. The safety rating (X / Y / Y2) is important here because the caps are connected between the life and neutral line (X rated ones) or between life/neutral and safety/frame ground (Y rating). Their failure mode is to go silently open, never short - which is important for your personal safety.
And yes, there always are exceptions, like these RIFA ones, that go bang instead of open (more common with X rated ones).
Looking at your photos, I'd recommend to replace the 1000pF ones by Y / Y2 rated ones (don't sue me if anything goes wrong, it's just an educated guess)

[TheUnnamedNewbie]

I have a (I assume) older scope of Phillips, altho the number is "bigger": PM3244.

Mine blew the X2 cap out in the first few hours I had it. I removed it and have been using it without it since then. EMC will be horrid, but I'm willing to live with that until I completely recap the power supply.

[Atheus]

These capacitors are for EMC "only". The scope will work without them, having some elevated level of noise emitting through the mains line ... who cares if no one is disturbed by that?
So if you replace them, any technology (Ceramic, any kind of film / paper) will do as long as you get the correct capacitance / safety rating. The safety rating (X / Y / Y2) is important here because the caps are connected between the life and neutral line (X rated ones) or between life/neutral and safety/frame ground (Y rating). Their failure mode is to go silently open, never short - which is important for your personal safety.
And yes, there always are exceptions, like these RIFA ones, that go bang instead of open (more common with X rated ones).
Looking at your photos, I'd recommend to replace the 1000pF ones by Y / Y2 rated ones (don't sue me if anything goes wrong, it's just an educated guess)

Oh really? While looking at desoldering the X2 (the big one) I noticed it was in parallel. I'm going to replace it like-for-like, with this:

https://www.bitsbox.co.uk/index.php?main_page=product_info&cPath=65_81&products_id=407

The value (0.22uF) is the same, it's an X2, and it's rated 275VAC (I think... crap blurry photo on site and does not specify in text).

I have the following as options to replace the 1000pF:

This: https://www.bitsbox.co.uk/index.php?main_page=product_info&cPath=65_81&products_id=3179 - TLDR; it's a 1nF Y2 - in the other thread someone told me definitely NOT to replace non-Y with Y however, and I don't see a 'Y' on the small caps... does anyone? Or an equivalent symbol? Basically can anyone else spec replacement parts for these for me? Y or non-Y? Not that I don't trust you @capt bullshot... the name does not inspire confidence lol... but you seem to know what you're talking about.

Or this: https://www.bitsbox.co.uk/index.php?main_page=product_info&products_id=1994 - TLDR; its a set or 'normal' (non X or Y) metal poly caps which includes the value I need. I know it's overbuy but you can never have too many caps

Second opinions anyone? Are those smaller gold caps X, Y, or neither?

~Atheus

[Atheus]

I have a (I assume) older scope of Phillips, altho the number is "bigger": PM3244.

Mine blew the X2 cap out in the first few hours I had it. I removed it and have been using it without it since then. EMC will be horrid, but I'm willing to live with that until I completely recap the power supply.

Hmmm... I think I'll just desolder that right now before it explodes! Rather not have cap smoke in this rented house.

It looks ready to as much as any component I've seen - although cracks are a strange form of damage!

~A

[capt bullshot]

capt bullshot... the name does not inspire confidence lol... but you seem to know what you're talking about.

Yes, that's something I did deliberately take into account when I chose this nick for some other forum long time ago. Now I just like it.
Dealing with EMC and electrical safety otherwise is part of my regular job. But for something more than a guess on your caps, I'd have to look at either the board or its schematics too see where exactly these are located and connected.

[Atheus]

capt bullshot... the name does not inspire confidence lol... but you seem to know what you're talking about.

Yes, that's something I did deliberately take into account when I chose this nick for some other forum long time ago. Now I just like it.
Dealing with EMC and electrical safety otherwise is part of my regular job. But for something more than a guess on your caps, I'd have to look at either the board or its schematics too see where exactly these are located and connected.

Oh good idea! Due to the wide spacing and the single-layer PCB it should be easy to snap a pic of the traces and label them for you. I also have the (large) service manual but it's been wrong once already. I'll check that for schematics first.

Thanks!

~Atheus

[capt bullshot]

Oh good idea! Due to the wide spacing and the single-layer PCB it should be easy to snap a pic of the traces and label them for you. I also have the (large) service manual but it's been wrong once already. I'll check that for schematics first.

WTF  Just re-discovered that PM3055 sitting under my desktop here (queued for testing / repair), service manual already downloaded ...
Look that excerpt from the power supply schematic:
C6001 is the 220n cap, used for EMC only (as I guessed). Replace this one with a X rated capacitor.
C6002 and C6006 are the 1000pF ones, serving as a capacitive voltage divider to provide the line trigger signal. So they aren't EMC capacitors in the first place, and line trigger won't operate if you remove them. Due to their direct connection to the mains line input, I'd strongly recommend to replace them with Y rated capacitors, to provide safety. The primary function (line trigger coupling) isn't supposed to be influenced by the actual material and other properties of these capacitors, any technology will do.

Don't know the exact differences between Y / Y2 and cousins, there are other people here that I'd ask if this was an issue related to my work, so look up datasheets if you're in doubt which one is the right one. IMO it wouldn't matter at this place, as long as there's an Y rating at all.

[SingedFingers]

Whoa you just got those RIFAs in time. Moment you see cracks, shit is about to hit the fan.

TBH if you want a quick hack, just remove C6002, 6003, 6004, 6006 entirely and leave them out of circuit. You'll lose line trigger but that's it. The X2 across the line is required if you want the scope to survive a transient.

[Atheus]

Thanks all! Parts should be arriving tomorrow so stay tuned to see how badly I screw this up

TBH if you want a quick hack, just remove C6002, 6003, 6004, 6006 entirely and leave them out of circuit. You'll lose line trigger but that's it. The X2 across the line is required if you want the scope to survive a transient.

So I could _remove_the ones listed above, but I must still replace C6001? The X2? I assume so.

I think I will replace all these parts since I have those values on hand (or on order), and I haven't calibrated the machine yet, so I don't want noise etc.

Please note that 03 and 04 are not marked RIFA, nor are gold, nor show any damage. Is there a particular reason to replace these - a better reason than this being an excuse to recap the whole thing...

... on that note should I recap the whole thing? Maybe just those 2 big blue electrolytics? They show no bulging etc but they're 20 years old.

Cheers!

~Atheus

/edit: Oh yea, by the way...

I'd strongly recommend to replace them with Y rated capacitors, to provide safety. The primary function (line trigger coupling) isn't supposed to be influenced by the actual material and other properties of these capacitors, any technology will do.

Sounds reassuring....

Looking at your photos, I'd recommend to replace the 1000pF ones by Y / Y2 rated ones (don't sue me if anything goes wrong, it's just an educated guess)

Sounds less reassuring...

Is everyone sure about this? Can I get a second opinion on whether Y caps are okay here? I already have them you see, so I'd like to not have wasted that money.

I've got some reading to do on X/Y caps apparently. I am slowly making my way through the classic "Art of Electronics" (3rd ed.) so maybe they are covered here somewhere.

~A

[capt bullshot]

TBH if you want a quick hack, just remove C6002, 6003, 6004, 6006 entirely and leave them out of circuit. You'll lose line trigger but that's it. The X2 across the line is required if you want the scope to survive a transient.

So I could _remove_the ones listed above, but I must still replace C6001? The X2? I assume so.

You must remove C6001, C6002, C6006 to keep them from going bang. The scope will be fine without them.
I don't see a reason to remove or replace C6003 and C6004. You may want to have a look at C6005 if it looks crispy like the RIFAs, otherwise leave it alone. Replacing C6001 or leaving it open has some influence on the transient survival capabilities (do not overestimate this one, the risk of getting the scope blown by a line transient is rather low), but mains line EMI radiation from the scope will increase if you leave it out.
Testing C6007 / C6008 for capacity and ESR might help in the decision to replace them or not. OTOH, there's no loss (except for cost) in replacing them.
Recapping the whole instrument "just because" - I'd never do this without a reason like "this instrument knowingly suffers from capacitor plague". There's a clear reason to replace C6001, C6002 and C6006: they are known to fail soon (because of their age, brand and look)

[Atheus]

Alright... I did It

Caps desoldered:



Scratched up the underside a bit due partly to ineptitude, and partly because the merest touch of a screwdriver will make those marks. I really hope I haven't cut or shorted and traces. I'll clean it up a bit and see if the scratches 'polish out'.



And here are the new caps! Couldn't get the 1nF in a wider pin pitch so I bodged it a bit. Should be okay. More concerned about the underside to honest.



So the big one is an X2 0.22uF 275V. The RIFA was 250V. The two others are Y2 1nF/1000pF 300VDC. The RIFAs were 125VAC... which may explain some things... if this is designed for 110V American supplies only it might well blow up! Any chance all these failures happen in 230/240V countries? Should be okay now anyway.

I see no damage to other caps so am leaving them alone. I already nearly cut a trace or something. I'll be lucky if it works...

... on that note, before I power it up, can anyone confirm for SURE I have done the right thing? The X2 is right? The two others are okay being Y2 EVEN THOUGH there is no Y2 marking on the RIFAs? I trust the EE skills of those who told me Y rating would be okay but just checking one last time... nothing is upside down or back to front or something dumb? I'll never find this scope for that price again!

Cheers,

~Atheus

/edit: Underside seems okay after a clean. I think there's a thin plastic coating I'm scratching or something - not the actual PCB traces...

... so shall I turn it on? Waiting for opinions on the repair first.

~A

[capt bullshot]

Short answer: Looks OK. No shorts or breaks visible, safe to turn on.

Long answer: The scratches on the bottom side are nothing to worry about. These old PCBs have the tendency to loose their inner via plating when de-soldering components using too much force / heat. You would have noticed their remainders on the pins of the removed capacitors. They look like small tubes left on the wire that don't melt when you heat it. Since all the traces are on the bottom side, loosing them wouldn't be a serious issue here as long as you take care to ensure the pin is connected to the trace, by visual inspection and continuity testing. On this kind of PCB, traces break most often at the connection to the pad outer ring, thats the place to inspect. The scratches aren't deep enough to damage the traces.

[Atheus]

Okay here we go... board replaced without the clips - the other fixings (two sits in the plastic, very tightly fitting the PCB. The clips are unneeded and actually cause a hazard during servicing. If I hadn't been acutely aware of the fact, and watching for it, I could have very easily ripped a component off the board while trying to remove it.

Even without them it's terrible design IMHO. You should never have to get your 'big' toolkit out for this sort of thing, instead of the electronics kit, and start using things like hammers. If the scope still works I'll be very relieved! I'd give it 50/50 chance... anyway there's only one way to find out!

[Atheus]

Nope. Nothing. F****d it. Damn.

/edit: Oh actually it could be... just a minute.

/edit 2: Yesssss!!! Working. Missed a semicolon... I mean ribbon cable

/edit 3: This seems to have actually improved the operation of the device! The trace is MUCH clearer, and I can now touch the calibration tab with a probe and see a clean square wave - before it was distorted and I couldn't clean it up. Now I just touch the tip of the probe to the tab, press auto-set, and I see a nice clean square wave!

Cosmetic repair and cleaning also complete, probes have been supplemented with cheap Chinese ones (I had no ground clips and they don't sell those separately) so I have all the accessories, and while it was open I made sure there were no other dodgy looking components. Thanks for all your help guys and girls! Especially capt bullshot - I wouldn't have had the balls to do that (if I even knew about it!) if it wasn't for you. I was going to hold you fully responsible if it didn't work... just kidding - thank you!

Time to go play with my new FULLY FUNCTIONAL oscilloscope!



Note the trace is slightly distorted in the photo, but I assure you, it's perfectly square to the human eye.

~Atheus

[capt bullshot]

Congratulations! 
Probably you've improved the operation by re-seating some of the internal connectors, some kinds of them get slack over time.

Inspired from your success, I pulled the one here from under the desktop for a quick check (without opening it, I don't fear exploding RIFAs  ),
the trace can easily changed between the two "modes" by slapping on the case. I've got some similar experience from the past with an old Philips TV set that could be repaired by whacking.
So some fun to repair left for the future.

Looks like this scope has a clear and crisp trace (if it's working as it is intended to), so repairing appears to be worth it.

[Atheus]

^ Yep! That distorted trace (top pic) is exactly like what I saw before the repair. Worse if anything. I'd go for it if I was you. Now mine's working it'll give me days of just enjoying learning how to use it, but for a proper EE like you, it's just another scope. I've seem new digital scopes in operation at my old company and they're obviously brilliant and much better. I don't know what else you have, but Dave always recommends keeping a 20MHz analog scope in his blog for certain applications, and this is 60Mhz, so it should/could be useful. Enough to analyse the I²C bus, RS232 as well I think, plus the stuff you might use a 20Mhz for.

It's even got a data to PC output, and an LCD, which is priceless. The buttons digitally change the time base etc and it all shows on the LCD. Min/max voltage, Hertz, trigger mode... it's almost like a digital scope with an analog screen.

Anyway now the distortion is completely absent - except the tiny dropoff you see in my pic (camera only) - I thought it was in too fast a time base before the repair, so I was seeing the shape of the "rise and fall" (terminology?) of the digital signal when I shouldn't be... it must take some time after all. Then it dawned on me that we are (and have been for some time) capable of creating clean digital signals in the GHz range so that was a non-starter. It should rise and fall almost instantly.

Anyway it must have been those RIFA caps making noise in the PSU - the connectors were all secure. As soon as I get a close-to-reliable ESR meter I'll test the other caps and see if it gets even better

Right now though, I've got reading to do on oscilloscopes... I might even think about building a signal generator. I need a serial port too. In fact I'll probably need all the old connectors. I've bought a cable (USB to serial) for now, but I'm thinking of something like this:

https://www.amazon.co.uk/dp/B06XCD3P9L/ref=wl_it_dp_o_pC_nS_ttl?_encoding=UTF8&colid=39AC15SYOFVA&coliid=I1G3E532K1U4XK

4 COM ports, dual LAN, fanless. Cool eh?

How do you guys get around the no-serial-port thing on newer machines? Adaptor? Keep an old PC? Buy a new one?

Thanks again!

~Atheus

[Ian.M]

Assuming the test signal is good,  (check it if you have access to a better known good scope) your probe compensation although pretty close, looks like it needs tweaking slightly to flatten out the remaining overshoot.  Set the trace to occupy 90% of the screen vertically, with top and bottom exactly on graticule lines to make any tilt or over/undershoot more obvious.  Use a low capacitance (e.g. ceramic or plastic blade) trimmer tool with the probe held between two blocks of expanded polystyrene, as although the probe is supposed to be shielded, there is an opening in the shielding for the trimmer, stray capacitance affects the result, and you need it to be compensated for when its clipped on to a pin, not for when you are probing around with your fat mitts round the probe body!
After trimming, don't swap the probe to another scope or channel without re-trimming.

[Shock]

How do you guys get around the no-serial-port thing on newer machines? Adaptor? Keep an old PC? Buy a new one?

You can go usb to serial, pci serial card or keep an old PC or laptop. The advantage with keeping some old pc or laptops around is they are dirt cheap and are less likely to have compatibility issues. If you know what you are doing you can resolve compatibility issues and visualize in older OS if you have to. There are other programs like DOSBox that can help you slow down programs. You can also use USB bootable floppies/drives or make USB bootable memory sticks to access older operating systems on modern computers or even hdd partitions if your adventurous.

[capt bullshot]

^ Yep! That distorted trace (top pic) is exactly like what I saw before the repair. Worse if anything. I'd go for it if I was you.

Yes, it's worth repairing for its sharp trace, but I don't like the user interface (these pushbuttons and LCD). I prefer having knobs that you can push and turn for all the main functions like the old Hameg or Tek analog scopes. By the way: one of these buttons was stuck on the PM3055 here, blocking the whole keyboard.

Now mine's working it'll give me days of just enjoying learning how to use it, but for a proper EE like you, it's just another scope. I've seem new digital scopes in operation at my old company and they're obviously brilliant and much better. I don't know what else you have, but Dave always recommends keeping a 20MHz analog scope in his blog for certain applications, and this is 60Mhz, so it should/could be useful. Enough to analyse the I²C bus, RS232 as well I think, plus the stuff you might use a 20Mhz for.

Wish you a lot of fun with that scope! Starting to learn with an analog scope is a good thing, you don't need a 20MHz scope if you have this one. It has a double/delayed timebase, that your typical 20MHz analog scope won't have. Try to find out what you can do with the second timebase, that's a good thing to learn. Get a microcontroller dev board and build a scope clock using the X/Y mode (way much better on an analog scope than on a digital one).

It's even got a data to PC output, and an LCD, which is priceless. The buttons digitally change the time base etc and it all shows on the LCD. Min/max voltage, Hertz, trigger mode... it's almost like a digital scope with an analog screen.

Sorry to disappoint you here: the LCD does nothing but to show the actuall settings, no values to measure.

How do you guys get around the no-serial-port thing on newer machines? Adaptor? Keep an old PC? Buy a new one?

Using USB adaptor, if it works. Sometimes it doesn't. Got an old PC from scrap recently because I needed one with ISA slot to bring my old EPROM programmer and an ISA bus GPIB interface card back to life

[george.b]

Sorry to disappoint you here: the LCD does nothing but to show the actuall settings, no values to measure.

While this is true, there is one thing the LCD can help with, which I find pretty nice to have in an analog scope: you can use the dual time base, its associated highlighted "cursor" and the LCD to measure the period of a waveform way more accurately than you'd be able to with just the screen graticule.

It's even got a data to PC output

Sorry to bring another disappointment, but that port on the back outputs no data. It's not a DSO or a combiscope, it's just analog. That port has something to do with remote control and is not RS232-compatible (http://www.qsl.net/vk5bar/AHARS-Resources/Philips%20PM3055/Philips%20PM3055%20Operating%20Manual.pdf, see page 3-13E).

[Ian.M]

Its certainly *NOT* capable of  data capture, and wont expose any capabilities that aren't already available on the LCD.   There's a bit more info and a possible source of interfaces at http://www.helmut-singer.de/stock/-1776332427.html

[Atheus]

It's even got a data to PC output, and an LCD, which is priceless. The buttons digitally change the time base etc and it all shows on the LCD. Min/max voltage, Hertz, trigger mode... it's almost like a digital scope with an analog screen.

Sorry to disappoint you here: the LCD does nothing but to show the actuall settings, no values to measure.

Eh? "does nothing but to show the actual settings" - isn't that what it's for? Rather than looking at a knob and trying to read off an exact timebase... disregarding that the analog knob might not be so accurate anymore... isn't an LCD+buttons better? It shows me "actual settings". That's great! Right? Or are you saying I might set it to 1ms TB shown on the display, but in reality it is 1.1ms, and the LCD will not reflect this? It will show what I set it to - not the REAL value? I would counter that a non-LCD scope has the same problem or worse... the knob will not reflect the EXACT value either...

Yes I have dual and delayed timebase - yay

It's even got a data to PC output

Sorry to bring another disappointment, but that port on the back outputs no data. It's not a DSO or a combiscope, it's just analog. That port has something to do with remote control and is not RS232-compatible (http://www.qsl.net/vk5bar/AHARS-Resources/Philips%20PM3055/Philips%20PM3055%20Operating%20Manual.pdf, see page 3-13E).

Oh... Crap. Well, so much for that! Thanks.

I might buy a cheap (-ish, but brand new) USB DSO just for data capture. I know they're supposed to be crap, but they're getting better I'm told.

[warning - long rant about ports on PCs ahead... should probably make a new thread]

I haven't actually got a serial port at the moment. Being a (web and mobile apps mostly) software dev, and a Mac and Linux user (for work etc) plus Windows for games, my machines are always cutting edge. I write this from a new Macbook Pro, and in the other room is a tweaked gaming machine which is so twitchy it only boots for me, not my girlfriend, which has it's advantages

I don't even think I've got serial on an old retired machine... or even on an old mobo from the bits box. I used to run some old machines for fun - and also they provided cheap processing power - often 64 bit and/or RISC architecture too which was useful and rare at the time. Partly because of the processors themselves and the SMP implementation (before the days of dual core) but also because things like SCSI and Ethernet were offloaded to dedicated hardware and freed up more of... everything really. I had two Sun SPARCs and a really nice SuperMicro dual P3-S board with loads of SCSI ports, 15K and 10K RPM drives (the noise!), ECC RAM, and of course a massive selection of I/O including serial, parallel, external SCSI, dual LAN and all sorts. Wish I had that now... although it sounds like a small aircraft.

I think this is quite niche, and the product itself I can't find many reviews for... but does anyone have any experience with this sort of thing?

https://www.amazon.co.uk/dp/B06XCD3P9L/ref=wl_it_dp_o_pC_nS_ttl?_encoding=UTF8&colid=39AC15SYOFVA&coliid=I1G3E532K1U4XK

Wouldn't this be great for hardware/firmware and 'IoT' development? Dual LAN to simulate a router, two WIFI antenna ports (so you can use that as a third network or test new antenna designs), serial/COM ports of course, five of those, and all the usual as well.

I imagine this running command line Linux, sitting on my electronics workbench (rather than my desk), with the colours set to all green on black

~Atheus