Confusion around probing isolated circuits. Is this a safe practice?

[shapirus]

BTW, is there a "go-to" design for a DIY HV differential probe with a reasonable bandwidth and CMRR high enough to be usable for probing generic SMPS'es?

I remember there have been quite a few threads to that effect, is there some design that would stand out in being not overly complex and good enough at the same time? Surely someone tried to build one?

I've seen quiet a few! Although not many of them are properly tested or succeeded well. Some of them don't do some things I've heard is good to do, but it's designed by people who surely know a lot more than I do.

https://resources.altium.com/p/differential-oscilloscope-probe-project
This one doesn't look... extremely good? https://www.instructables.com/DIY-1GHz-Active-Probe-for-Under-20/ it's an active fet tho not differential.
https://xellers.wordpress.com/electronics/1ghz-active-differential-probe/ this one didn't really work well up to 1ghz, but it still worked at lower freq? also these days 4 layer PCBs with specific layouts barely cost a few dollars. I can do 50x50mm for 2$ at
https://hackaday.io/project/167197-poor-mans-1ghz-differential-probe
JLCPCB!! Rogers is 96$ for a 2 layer 100x50mm tho, 48 for teflon.


I thought about building one of these!

These are (mostly) different beasts: RF active probes, not necessarily differential. They serve a different purpose, not that of safely probing between two arbitrary points in DUT.

[EEVblog]

BTW, is there a "go-to" design for a DIY HV differential probe with a reasonable bandwidth and CMRR high enough to be usable for probing generic SMPS'es?

I remember there have been quite a few threads to that effect, is there some design that would stand out in being not overly complex and good enough at the same time? Surely someone tried to build one?

https://www.eevblog.com/forum/projects/cheap-hv-differential-oscilloscope-probe/
https://www.eevblog.com/forum/projects/diy-mains-voltage-differential-probe-for-smps-design/
https://www.eevblog.com/forum/projects/hv-50mhz-differential-probe-(diy-proyect)/
https://www.eevblog.com/forum/projects/my-high-voltage-differential-probe-design/
https://www.eevblog.com/forum/beginners/low-cost-diy-differential-probe/

Yes, thanks, I can use the search function .

But you didn't 

[shapirus]

BTW, is there a "go-to" design for a DIY HV differential probe with a reasonable bandwidth and CMRR high enough to be usable for probing generic SMPS'es?

I remember there have been quite a few threads to that effect, is there some design that would stand out in being not overly complex and good enough at the same time? Surely someone tried to build one?

https://www.eevblog.com/forum/projects/cheap-hv-differential-oscilloscope-probe/
https://www.eevblog.com/forum/projects/diy-mains-voltage-differential-probe-for-smps-design/
https://www.eevblog.com/forum/projects/hv-50mhz-differential-probe-(diy-proyect)/
https://www.eevblog.com/forum/projects/my-high-voltage-differential-probe-design/
https://www.eevblog.com/forum/beginners/low-cost-diy-differential-probe/

Yes, thanks, I can use the search function .

But you didn't 

Well, search wouldn't pick me the one that's most reasonable of all, if there is one, that's why I asked if maybe someone knew one from their experience.

[bdunham7]

This makes me concerned that the currents flowing to the ground terminal of my scope, although very small could potentially damage it. Is probing isolated circuits like this a good/common practice? Is this simply nothing to worry about and normal?

This is entirely normal and one of the functions of the ground is to conduct those currents.  µA currents will not harm you or your scope.

[AmeliaBuns]

Here is what I get when I go the link you provided..  I never saw an item restocked when it is marked like this.

That's because I use Amazon.ca and you're not in Canada
PS: I can see your postal code there if you're bothered by that. can be used to track you potentially.

[AmeliaBuns]

Hmm so what I'm hearing overall is that even if you're circuit is isolated, just don't probe it as is without a differential probe? and that I should ground my circuits unless for some reason, it must not be grounded:?  I'm a bit confused. the 60-200ua of current going into the ground of my scope is apparently just fine, the tiny spark is the spooky part. (not sure how that's happening)

I drew a picture to illustrate what the main problem is:

As mentioned, all inputs of a scope are connected to each other (unless you have a scope where the channels are individually isolated).
However, this also means that if you want to measure with several channels, for example, but at different (separate) potential circuits, their galvanic isolation from each other is thereby eliminated.
The consequences can range from malfunctions to destruction of the circuits.
An isolation transformer is used to isolate the potential from the DUT to the mains.
A scope should not be connected via an isolation transformer.
Why use an isolating transformer on the DUT?
If the device is not grounded because it belongs to protection class II.
These devices usually have a 2-pole power plug that can be plugged into any outlet.
However, this also means that the line can then be reversed with the phase.
If you then connect the negative terminal of a probe to the phase, you short-circuit the phase to ground – not good...


Oh yes I understand that well. My question was more in the  specific scenario of only using one probe (leaving others disconnected to stop myself from forgetting) and not isolating the scope, but rather having iosolation on my power supply side. (meaning the output is also low voltage(<30v) and isolated).

[AmeliaBuns]

BTW, is there a "go-to" design for a DIY HV differential probe with a reasonable bandwidth and CMRR high enough to be usable for probing generic SMPS'es?

I remember there have been quite a few threads to that effect, is there some design that would stand out in being not overly complex and good enough at the same time? Surely someone tried to build one?

I've seen quiet a few! Although not many of them are properly tested or succeeded well. Some of them don't do some things I've heard is good to do, but it's designed by people who surely know a lot more than I do.

https://resources.altium.com/p/differential-oscilloscope-probe-project
This one doesn't look... extremely good? https://www.instructables.com/DIY-1GHz-Active-Probe-for-Under-20/ it's an active fet tho not differential.
https://xellers.wordpress.com/electronics/1ghz-active-differential-probe/ this one didn't really work well up to 1ghz, but it still worked at lower freq? also these days 4 layer PCBs with specific layouts barely cost a few dollars. I can do 50x50mm for 2$ at
https://hackaday.io/project/167197-poor-mans-1ghz-differential-probe
JLCPCB!! Rogers is 96$ for a 2 layer 100x50mm tho, 48 for teflon.


I thought about building one of these!

These are (mostly) different beasts: RF active probes, not necessarily differential. They serve a different purpose, not that of safely probing between two arbitrary points in DUT.

Oh yes I included those too because I found them cool I wanted to make one because I don't have a spare 5,000$!
Someone here actually made one,  but it's not opensource, they sell it for 250 euroes I think? for now i'll give up on 6mhz+ noise on  my power supply and measure ripple at 1:1 and maybe 10:1 (tbh, 5mv noise is still going to be much more than any switching power supply I could design!)

[AmeliaBuns]

This makes me concerned that the currents flowing to the ground terminal of my scope, although very small could potentially damage it. Is probing isolated circuits like this a good/common practice? Is this simply nothing to worry about and normal?

This is entirely normal and one of the functions of the ground is to conduct those currents.  µA currents will not harm you or your scope.

Awesome! what about the sparks? What can I do to eliminate them?
TVS diode was one idea I had, but i'm confused about how they even happen. I'm aware of sparks from suddenly disconnecting high current loads due to inductance, but 200uA is barely anything to generate high enough voltage to ionize the air and jump to the scope! I also know during connections sparks happen with high loads as the connector surface isn't perfect and you can make a very high resistance connection and weld the tip. It might have something to do with electromagnetic fields too. not 100% sure. (if the distance is very small, the electrons might just start jumping?)

[EEVblog]

In case the OP hasn't seen it:

[AmeliaBuns]

In case the OP hasn't seen it:

Oh I've actually seen that video a few times It was the first video I watched when I bought my oscilloscope   It's a very good and useful video.

[watchmaker]

Here is what I get when I go the link you provided..  I never saw an item restocked when it is marked like this.

That's because I use Amazon.ca and you're not in Canada
PS: I can see your postal code there if you're bothered by that. can be used to track you potentially.

What privacy?

I managed several databases in the 80s and and 90s when SS numbers were still sacrosanct.  Doge gave them all away, Flock cameras, Pegasus, Cambridge Analytical, etc. OTOH, I knew enough to keep my digital footprint pretty small.  When I do deep searches on my and my wife's name, very little comes up.   And I have been on since 1983 (DarpaNet).

Back in the 1990s, when our daughter was in middle school, I showed her the router logs.  I made it clear I would never monitor them, but I wanted her to be aware that others, particularly employers, will.

In the 1950s my mother applied about my missing father to the federal office that located missing people by SSN (via Hollerith sorts!).  The reply was yes, he is alive, but we cannot tell you anything else because such info is protected by law.  I have the letters.

This lies at the heart of why many boomers prefer direct interactions.

So it goes. (KV)

[AVGresponding]

The arc happens because as you disconnect the scope ground lead, the point it was connected to is no longer shorted to ground, and the AC current flowing through* the Y capacitor allows the potential to rise to a level that an arc occurs.



*technically the current doesn't flow through the capacitor, but that's a different question.

Wait, but unconnected, the voltage on the negative/positive terminal barely goes above 4.8Vp. That's barely enough to create an arc. Why would that happen during a disconnect? I thought it's caused by "parasitic inductance" in most devices, but 96ua is so low. Is there a way to avoid this?

But the main reasons seems to be connecting two different ground clips in places you shouldn't? or the sparks.

I usually avoid using more than one probe at a time when doing this. once another probe is connected to my scope, I make sure everything's earthed and only connected to ground.

The reason you only measure 4.8V between GND and either -ve or +ve is because the meter has a low enough impedance (probably 10MΩ or so) to pull the voltage down to that. The open circuit voltage is going to be higher, and can, depending on what other circuit components interfere, reach pretty much whatever your line voltage is (unlikely to get this high though).

The best way to avoid this, if you are only going to be using the -ve rail as your scope grounding point, is to connect the -ve and GND with a shorting link on the psu; that's what they are there for. If you need to connect the scope ground somewhere that isn't the PSU -ve, you should not short the PSU -ve and GND, and only connect and disconnect the probe ground while the DUT is switched off.

[AmeliaBuns]

Here is what I get when I go the link you provided..  I never saw an item restocked when it is marked like this.

That's because I use Amazon.ca and you're not in Canada
PS: I can see your postal code there if you're bothered by that. can be used to track you potentially.

What privacy?

I managed several databases in the 80s and and 90s when SS numbers were still sacrosanct.  Doge gave them all away, Flock cameras, Pegasus, Cambridge Analytical, etc. OTOH, I knew enough to keep my digital footprint pretty small.  When I do deep searches on my and my wife's name, very little comes up.   And I have been on since 1983 (DarpaNet).

Back in the 1990s, when our daughter was in middle school, I showed her the router logs.  I made it clear I would never monitor them, but I wanted her to be aware that others, particularly employers, will.

In the 1950s my mother applied about my missing father to the federal office that located missing people by SSN (via Hollerith sorts!).  The reply was yes, he is alive, but we cannot tell you anything else because such info is protected by law.  I have the letters.

This lies at the heart of why many boomers prefer direct interactions.

So it goes. (KV)

lol that's fair. I also feel mostly the same way, that privacy is a lost cause at this point, but I also prefer to be cautious. This stuff makes it easier for non savvy people to do stuff, but honestly? yeah.....

[AmeliaBuns]

Wait, but unconnected, the voltage on the negative/positive terminal barely goes above 4.8Vp. That's barely enough to create an arc. Why would that happen during a disconnect? I thought it's caused by "parasitic inductance" in most devices, but 96ua is so low. Is there a way to avoid this?

But the main reasons seems to be connecting two different ground clips in places you shouldn't? or the sparks.

I usually avoid using more than one probe at a time when doing this. once another probe is connected to my scope, I make sure everything's earthed and only connected to ground.

The reason you only measure 4.8V between GND and either -ve or +ve is because the meter has a low enough impedance (probably 10MΩ or so) to pull the voltage down to that. The open circuit voltage is going to be higher, and can, depending on what other circuit components interfere, reach pretty much whatever your line voltage is (unlikely to get this high though).

The best way to avoid this, if you are only going to be using the -ve rail as your scope grounding point, is to connect the -ve and GND with a shorting link on the psu; that's what they are there for. If you need to connect the scope ground somewhere that isn't the PSU -ve, you should not short the PSU -ve and GND, and only connect and disconnect the probe ground while the DUT is switched off.

Ooh that makes sense. It's kind of the observer effect but I'm still a bit shocked that it does that. I felt like you'd need a good current too for  arcs to make it jump to a few kV.
My feelings towards this is so far to just honestly give up and ground the thing, and leave differential stuff to my imagination/calculations It's a shame cause I wanted to probe shunt resistors that aren't necessarily on the low side. I did connect a 20k (or maybe even 1MOhm) to avoid the arcs before hoping it'd be so small that it'd not mess with anything, but also get rid of the arcs, but that seems a bit... not great.

Or let the arcs slowly kill my scope so I have a justifiable excuse for a Rigol mho900/SDS3000x HD  (Joking ofc)

[KungFuJosh]

Or let the arcs slowly kill my scope so I have a justifiable excuse for a Rigol mho900/SDS3000x HD  (Joking ofc)

Ewwwww, don't joke about buying a Rigol.

[AmeliaBuns]

Or let the arcs slowly kill my scope so I have a justifiable excuse for a Rigol mho900/SDS3000x HD  (Joking ofc)

Ewwwww, don't joke about buying a Rigol.

I don't know much about them, but what's wrong with them? I saw dave's review on the new mho98 and it looks cool! I'd personally benefit far more from a budget dif/active/power rail probe than a fancier scope at this point. I don't need 50 ohm termination (I can use terminators or a 50 ohm at the end of my BNC for now.
I should make that BNC to IPX probe sometime, I want it so I can stop holding my probe half-hazardly above my capacitors. the solder joints for some reason don't conduct well without a bit of poking.
I'd take a non-50 ohm bnc if I could find one (with low capacitance) as a 1:1 or 10:1 probe...

[AVGresponding]

Wait, but unconnected, the voltage on the negative/positive terminal barely goes above 4.8Vp. That's barely enough to create an arc. Why would that happen during a disconnect? I thought it's caused by "parasitic inductance" in most devices, but 96ua is so low. Is there a way to avoid this?

But the main reasons seems to be connecting two different ground clips in places you shouldn't? or the sparks.

I usually avoid using more than one probe at a time when doing this. once another probe is connected to my scope, I make sure everything's earthed and only connected to ground.

The reason you only measure 4.8V between GND and either -ve or +ve is because the meter has a low enough impedance (probably 10MΩ or so) to pull the voltage down to that. The open circuit voltage is going to be higher, and can, depending on what other circuit components interfere, reach pretty much whatever your line voltage is (unlikely to get this high though).

The best way to avoid this, if you are only going to be using the -ve rail as your scope grounding point, is to connect the -ve and GND with a shorting link on the psu; that's what they are there for. If you need to connect the scope ground somewhere that isn't the PSU -ve, you should not short the PSU -ve and GND, and only connect and disconnect the probe ground while the DUT is switched off.

Ooh that makes sense. It's kind of the observer effect but I'm still a bit shocked that it does that. I felt like you'd need a good current too for  arcs to make it jump to a few kV.
My feelings towards this is so far to just honestly give up and ground the thing, and leave differential stuff to my imagination/calculations It's a shame cause I wanted to probe shunt resistors that aren't necessarily on the low side. I did connect a 20k (or maybe even 1MOhm) to avoid the arcs before hoping it'd be so small that it'd not mess with anything, but also get rid of the arcs, but that seems a bit... not great.

Or let the arcs slowly kill my scope so I have a justifiable excuse for a Rigol mho900/SDS3000x HD  (Joking ofc)

As long as you connect and disconnect the ground lead of the probe with the PSU turned off, you won't shouldn't get an arc. If you still do, it means either the power switch is not a "hard" power but "soft" power switch, or is only single pole (switches only either L or N and not both). My guess is, it is a proper power switch and this method solves your problem, though it must be pointed out that for very precise measurements of your DUT, you probably want a better PSU, as there'll be quite a bit of noise in your measurement.

Some people here don't like Rigol much because they are not very well supported, unlike Siglent for example, who take an active approach to bug fixes etc. Personally, I'm firmly in the camp of those who like well documented older kit, that they can repair themselves, should it become faulty (though I do have a Siglent 1202 as well as my Tek TDS scopes).

[KungFuJosh]

Or let the arcs slowly kill my scope so I have a justifiable excuse for a Rigol mho900/SDS3000x HD  (Joking ofc)

Ewwwww, don't joke about buying a Rigol.

I don't know much about them, but what's wrong with them? I saw dave's review on the new mho98 and it looks cool! I'd personally benefit far more from a budget dif/active/power rail probe than a fancier scope at this point. I don't need 50 ohm termination (I can use terminators or a 50 ohm at the end of my BNC for now.
I should make that BNC to IPX probe sometime, I want it so I can stop holding my probe half-hazardly above my capacitors. the solder joints for some reason don't conduct well without a bit of poking.
I'd take a non-50 ohm bnc if I could find one (with low capacitance) as a 1:1 or 10:1 probe...

I've personally only had bad experiences with Rigol, but don't take my word for it. Look for @Martin72's reviews of different Rigol scopes. Martin doesn't have any bias against them, and he gives good, honest reviews.

If you want a great budget scope, look at the SDS800X HD series. If you want something more powerful, then go up to Siglent's 2000 or 3000 series depending on your budget.

That said, only you can decide if you actually need another scope. Having tools like a good differential probe might benefit you more. I'd recommend getting advice on different differential probes and patiently wait for a good deal to show up on eBay.

[watchmaker]

Or buy a 30VDC 3A  Instek LINEAR PSU  like this.  All my PSUs are linear. Reduces  RF noise and since I am not an industrial lab, the energy savings of a SMPS are not important.

I selected this one because the seller accepts returns and it is only $75 shipped.  Even accepts offers.

https://www.ebay.com/itm/177402804844?

Also, this was just released and I found it very instructive:

[AmeliaBuns]

Wait, but unconnected, the voltage on the negative/positive terminal barely goes above 4.8Vp. That's barely enough to create an arc. Why would that happen during a disconnect? I thought it's caused by "parasitic inductance" in most devices, but 96ua is so low. Is there a way to avoid this?

But the main reasons seems to be connecting two different ground clips in places you shouldn't? or the sparks.

I usually avoid using more than one probe at a time when doing this. once another probe is connected to my scope, I make sure everything's earthed and only connected to ground.

The reason you only measure 4.8V between GND and either -ve or +ve is because the meter has a low enough impedance (probably 10MΩ or so) to pull the voltage down to that. The open circuit voltage is going to be higher, and can, depending on what other circuit components interfere, reach pretty much whatever your line voltage is (unlikely to get this high though).

The best way to avoid this, if you are only going to be using the -ve rail as your scope grounding point, is to connect the -ve and GND with a shorting link on the psu; that's what they are there for. If you need to connect the scope ground somewhere that isn't the PSU -ve, you should not short the PSU -ve and GND, and only connect and disconnect the probe ground while the DUT is switched off.

Ooh that makes sense. It's kind of the observer effect but I'm still a bit shocked that it does that. I felt like you'd need a good current too for  arcs to make it jump to a few kV.
My feelings towards this is so far to just honestly give up and ground the thing, and leave differential stuff to my imagination/calculations It's a shame cause I wanted to probe shunt resistors that aren't necessarily on the low side. I did connect a 20k (or maybe even 1MOhm) to avoid the arcs before hoping it'd be so small that it'd not mess with anything, but also get rid of the arcs, but that seems a bit... not great.

Or let the arcs slowly kill my scope so I have a justifiable excuse for a Rigol mho900/SDS3000x HD  (Joking ofc)

As long as you connect and disconnect the ground lead of the probe with the PSU turned off, you won't shouldn't get an arc. If you still do, it means either the power switch is not a "hard" power but "soft" power switch, or is only single pole (switches only either L or N and not both). My guess is, it is a proper power switch and this method solves your problem, though it must be pointed out that for very precise measurements of your DUT, you probably want a better PSU, as there'll be quite a bit of noise in your measurement.

Some people here don't like Rigol much because they are not very well supported, unlike Siglent for example, who take an active approach to bug fixes etc. Personally, I'm firmly in the camp of those who like well documented older kit, that they can repair themselves, should it become faulty (though I do have a Siglent 1202 as well as my Tek TDS scopes).

I'm not sure if this is normal, but I did just notice that my dreaded transient problem where on the lab bench side and not my PCB. I thought I did something wrong and spent so long, but it has a horrible voltage drop/spike when my esp-32 draws current (when I press reset) or does anything fancy. It's good to design a device for the worst case scenario, but at the same time.... I think I might have to start saving for a proper Siglent lab bench, or google around and see if there's cheaper ones... I was gonna buy a ta305 next month to retire my Pinecil and it's super leaky tip (the ground clip exists, but it's cumbersome) and unergonomic design (I love my pinecil, but it's not an amazing all day solder) so I won't have much money left LOL.

I was going to grab the TC22 which is far cheaper, but the leakage etc scared me... even tho I feel like it's probably nothing and fine (And the people online who report leakage are probably not measuring right)

Siglent doesn't seem to update their models often (Their latest refresh, which was my unit took them 7 years!) So i wonder if there'll release a competitor soon....  1200CAD vs 5000CAD is a huge price difference, but that's ignoring the extras the Siglent sds3000x has. I can't afford one anyways /shrug.

[AmeliaBuns]

Or let the arcs slowly kill my scope so I have a justifiable excuse for a Rigol mho900/SDS3000x HD  (Joking ofc)

Ewwwww, don't joke about buying a Rigol.

I don't know much about them, but what's wrong with them? I saw dave's review on the new mho98 and it looks cool! I'd personally benefit far more from a budget dif/active/power rail probe than a fancier scope at this point. I don't need 50 ohm termination (I can use terminators or a 50 ohm at the end of my BNC for now.
I should make that BNC to IPX probe sometime, I want it so I can stop holding my probe half-hazardly above my capacitors. the solder joints for some reason don't conduct well without a bit of poking.
I'd take a non-50 ohm bnc if I could find one (with low capacitance) as a 1:1 or 10:1 probe...

I've personally only had bad experiences with Rigol, but don't take my word for it. Look for @Martin72's reviews of different Rigol scopes. Martin doesn't have any bias against them, and he gives good, honest reviews.

If you want a great budget scope, look at the SDS800X HD series. If you want something more powerful, then go up to Siglent's 2000 or 3000 series depending on your budget.

That said, only you can decide if you actually need another scope. Having tools like a good differential probe might benefit you more. I'd recommend getting advice on different differential probes and patiently wait for a good deal to show up on eBay.

I was mostly making a joke. I can't justify spending money on a scope.
I actually currently have an SDS824x HD which I absolutely love! Just there are very few times where I wish I had a nicer scope. Can't justify it for a hobby and very occasional "need". as you said, I also think my money would be better spent on probes (Differential, Power rail, Active fet) maybe i'll design my own one day, tho there's no way i'm beating even the DP700/MICSIG for less money without learning a TON more. Those even have MCUs in them it seems... One issue with them is that their noise is too much for them to be used as a current probe (for a wired setup)

[KungFuJosh]

I was mostly making a joke. I can't justify spending money on a scope.
I actually currently have an SDS824x HD which I absolutely love! Just there are very few times where I wish I had a nicer scope. Can't justify it for a hobby and very occasional "need". as you said, I also think my money would be better spent on probes (Differential, Power rail, Active fet) maybe i'll design my own one day, tho there's no way i'm beating even the DP700/MICSIG for less money without learning a TON more. Those even have MCUs in them it seems... One issue with them is that their noise is too much for them to be used as a current probe (for a wired setup)

That's a great scope! And you might think you were joking, but one thing this website has taught me is that TEAS is real. Good luck.

What made you wish you had a nicer scope?

There're ways to get around the noise of a diff probe, but yeah, it's a noisy probe in the lower range.

[AmeliaBuns]

I was mostly making a joke. I can't justify spending money on a scope.
I actually currently have an SDS824x HD which I absolutely love! Just there are very few times where I wish I had a nicer scope. Can't justify it for a hobby and very occasional "need". as you said, I also think my money would be better spent on probes (Differential, Power rail, Active fet) maybe i'll design my own one day, tho there's no way i'm beating even the DP700/MICSIG for less money without learning a TON more. Those even have MCUs in them it seems... One issue with them is that their noise is too much for them to be used as a current probe (for a wired setup)

That's a great scope! And you might think you were joking, but one thing this website has taught me is that TEAS is real. Good luck.

What made you wish you had a nicer scope?

There're ways to get around the noise of a diff probe, but yeah, it's a noisy probe in the lower range.

What's TES? and  and how  do you get aroudn the noise of a Diff probe?

I don't remember off the top of my head right now what/when made me feel the "need" for a better scope TBH. the 50 ohm inputs can be nice but they're a bit dangerous for someone like me. I'm going to just purchase a terminator (inline) for now, and see if I can make a series terminator instead?

Another one was CAN_FD protocols etc, but TBH i've yet to actually design anything with CAN.

[AVGresponding]

I'm not sure if this is normal, but I did just notice that my dreaded transient problem where on the lab bench side and not my PCB. I thought I did something wrong and spent so long, but it has a horrible voltage drop/spike when my esp-32 draws current (when I press reset) or does anything fancy. It's good to design a device for the worst case scenario, but at the same time.... I think I might have to start saving for a proper Siglent lab bench, or google around and see if there's cheaper ones... I was gonna buy a ta305 next month to retire my Pinecil and it's super leaky tip (the ground clip exists, but it's cumbersome) and unergonomic design (I love my pinecil, but it's not an amazing all day solder) so I won't have much money left LOL.

I was going to grab the TC22 which is far cheaper, but the leakage etc scared me... even tho I feel like it's probably nothing and fine (And the people online who report leakage are probably not measuring right)

Siglent doesn't seem to update their models often (Their latest refresh, which was my unit took them 7 years!) So i wonder if there'll release a competitor soon....  1200CAD vs 5000CAD is a huge price difference, but that's ignoring the extras the Siglent sds3000x has. I can't afford one anyways /shrug.

Sounds very much like the load regulation of your PSU isn't up to the task. If you can get by with your current soldering solution, I'd suggest prioritising buying a better PSU, possibly a used HPAK one, as they are usually a lot cheaper in North America than Europe. The Siglent 3303 is reliable (barring one well known and easily fixed flaw on earlier versions), the most annoying thing being the spacing of the binding posts.

TEAS is Test Equipment Acquisition Syndrome, my signature provides an example of this...

[AmeliaBuns]

I'm not sure if this is normal, but I did just notice that my dreaded transient problem where on the lab bench side and not my PCB. I thought I did something wrong and spent so long, but it has a horrible voltage drop/spike when my esp-32 draws current (when I press reset) or does anything fancy. It's good to design a device for the worst case scenario, but at the same time.... I think I might have to start saving for a proper Siglent lab bench, or google around and see if there's cheaper ones... I was gonna buy a ta305 next month to retire my Pinecil and it's super leaky tip (the ground clip exists, but it's cumbersome) and unergonomic design (I love my pinecil, but it's not an amazing all day solder) so I won't have much money left LOL.

I was going to grab the TC22 which is far cheaper, but the leakage etc scared me... even tho I feel like it's probably nothing and fine (And the people online who report leakage are probably not measuring right)

Siglent doesn't seem to update their models often (Their latest refresh, which was my unit took them 7 years!) So i wonder if there'll release a competitor soon....  1200CAD vs 5000CAD is a huge price difference, but that's ignoring the extras the Siglent sds3000x has. I can't afford one anyways /shrug.

Sounds very much like the load regulation of your PSU isn't up to the task. If you can get by with your current soldering solution, I'd suggest prioritising buying a better PSU, possibly a used HPAK one, as they are usually a lot cheaper in North America than Europe. The Siglent 3303 is reliable (barring one well known and easily fixed flaw on earlier versions), the most annoying thing being the spacing of the binding posts.

TEAS is Test Equipment Acquisition Syndrome, my signature provides an example of this...

Lol I relate to this. I wanted a spectronometer sooo bad for a month or two until I gave up and accepted the unkown. I just GOTTA KNOW if my lights are high quality. you can't just... not know right? This is clearly an essential life purchase, right up there with food.

I wonder if the siglents will also cause the arcing and other issues, I think AFAIK 30-300uA is considered "good" leakage? Tho test equipment is probably made to have less.