TNP #36 - Tektronix TCP0030 120MHz, 30A AC/DC Current Probe Teardown, & Failure Analysis

[@kcmichaelm]

When I was a tech in the field with IBM, we were told never to attempt to measure the resistance of a Hall Effect sensor, as they were so delicate that just the current from the meter was potentially enough to damage them. Now I’m just always nervous about them!

[@funtechu]

Always a good day when there is another teardown from Shahriar!

[@martinhodge921]

Most channels I watch at 1.25x speed. This one I must watch at .75x to keep my head from spinning. 😊

[@djmips]

And I have to rewatch a few times.

[@RabbitHoleLabs]

To remove the potting compound, try soaking in MG CHEMICALS 8310A Conformal Coating Remover. Might take a few days. Successfully removed a potted gate-pulse transformer this way.

[@AndrewZonenberg]

The general recommendation I've heard is never to buy current probes of this architecture (whether Tek or a competing vendor) used unless there's some kind of warranty from the seller, for exactly this reason. Apparently these Hall sensors are super fragile and dropping the probe from a fairly low height can be enough to ruin them. So most of the ones on ebay etc are dead and beyond economical repair.

[@MrThoriam]

So what architecture do you recommend?

[@RyanVasquez6089]

I used one for years at my previous. Vintage and they worked great!

[@PoLoMoTo2]

We just got a TCP0030A a few weeks ago at work and it has been incredible. Being able to visualize current curves in different parts of products so easily and without disrupting the circuit has helped greatly in just the short time we've had it. Has already paid for it self honestly. Very cool to see how it works! I wonder what changes were made for the TCP0030A over the TCP0030, I noticed the plastics are different but I wonder if they also maybe addressed the issue these probes have.

[@benedienst]

The Transformer subassembly and it’s issues are actually often a real pain. Nevertheless, the TCP0030 and it’s A model are still the best current probes I know. If you’re searching for something similar performance but less proprietary in connection, search for a TCPA300 and a TCP312A. It’s 100MHz and has a Tekprobe L2 output and can be used with any scope and a 50 ohm term. Funfact: The Tek VPI, TekProbe and the High end TekConnect are actually only iterations of the same interface standard.

[@shazam6274]

After going through the angst of these probes several times, I ended up using the high current ones like the Tek 6303 100A. These are quite large, but also very rugged. Haven't had one go bad yet. FYI: after doing similar deconstruction on older version of these, sometimes the thin wire is fried open by a large current spike at turn on or off of the DUT, so don't do that!

[@CNe7532294]

That construction explains it. Good news is that now we know why we shouldn't stuff our probes with wire. Temperature probably plays into too.

[@L4b3n]

Many thx for the teardown + explanation. If I can save another one from the dumpster, I will donate it as well.

[@mikeselectricstuff]

I wonder if you could get something useable, if lower performance, by transplanting an off the shelf packaged hall sensor into the core

[@Thesignalpath]

The Hall sensor is essentially deposited on the ferrite core material. It is super thin to not influence the transformer core. I am not sure if an off-the-shelf solution would be easy to integrate into this formfactor.

[@xDevscom_EE]

Nice! Real repair challenge, on mine it was just broken plastic lever, so I got lucky. These probes have impressive capabilities but require real baby care. Now we see why!

[@ileile]

Try to decompose the resin by placing it in a jar with acetone, carefully exposing it to the vapor. Takes some time for the resin to become soft and brittle. Maybe a way to closely look at the inner pcb without damaging

[@gsuberland]

Hot distilled water works better. I just depotted a solid 5cm of epoxy around a device by boiling it - it'll go crumbly and be easy to remove with a scalpel or box cutter. The water ensures that nothing ever gets above 100C, which all the parts can tolerate almost indefinitely as long as they're not operating. A quick spray down with IPA afterwards works great as a drying agent. This approach doesn't affect outside plastic housings like acetone does, either.

[@graealex]

One note - you should de-interlace the microscope footage. When you move stuff around, combs appear on the edges, and that doesn't play nicely with video compression that assumes progressive video.

[@NiHaoMike64]

Used to use those at work, they tended to break all the time. Now they're trying out probes from several different manufacturers for replacements, all of them seem to be more reliable than the Tektronix.

[@ccoder4953]

Yeah, we use those at work too and have had several break. Funny thing is there only seems to be two actual manufacturers for these. Tektronix invented this type of probe and still makes them. The other manufacturer is, I think, Hioki (OEM is probably Japanese due to the Japanese writing on the labels, and I think I read on EEVBlog forum it was Hioki). That's who, I believe, makes them for Keysight, LeCroy, Rigol, and a few others. If you notice, the current probes from all of them look suspiciously similar, apart from whatever proprietary connector they need for their scope. Although, looks like Siglent has a new series that looks a little different, so maybe there's somebody new. Or Hioki (or whoever is the OEM) just changed their plastics a bit.

[@RandomUser2401]

unpopular opinion, but Tek stuff is overpriced. Including their scopes. Bought a 50k scope from them which came with some horrible SW bugs, including an aliasing one. SW in a worse state than in a 2k Siglent scope. So for us it's Keysight, R&S or China.

[@ccoder4953]

@@RandomUser2401 I'd actually agree. We had some Tek scopes (this was some time ago, so things may have changed) where they came with stuff like caps broken off and rattling around inside. In general, the newer LeCroy stuff has been pretty good too. We have quite a few Keysight scopes, but their sub $10k stuff is sort of lackluster these days - they really need a new low end ASIC. We've also gotten some of the Siglent scopes for lower end stuff and, for the money, it's hard to match that with anything from the big manufacturers. Pretty sure the big names are eventually going to just hand that market to Siglent, Rigol and others and just focus on higher end stuff. They might rebrand some of the Chinese scopes, like LeCroy does with Siglent, but I think that's a market they increasingly have a hard time doing their own scopes for.

[@SidneyCritic]

I was thinking cracked ferrite core. We have them on automotive timing lights to sense the HV ignition leads, and drop it once and it's toast, ie, cracks the brittle ferrite core. It's hard to tell from the vid, but it looks like a glued sandwich, ie, easy to separate with a shock load. Or maybe the potting glue shrunk with age and a too thick cable snapped it open. Maybe that plastic thing in the jaws is critical, or maybe Tek has a faulty item that needs a recall.

[@user-rs8zg8ey2b]

I have this probe and the tcpa300 + tcp305 and prefer the evternal amp version, cleaner signal on my dpo4104b scope. Thanks for sharing.

[@windward2818]

The head mating circuit PCB layout small pad traces should have teardrops with entrance into a round through-hole pad (the traces are very thin). Also, a common PCB CAD layout routing starting mode has the traces from round pads come out at straight (in-line with the board shape) then make 45 degree bends. But, really if you have two round through hole pads close together, like on the bottom large trace, you can just go from the center of one to the center of the other without any bends at all. If you add to this teardrops you will get a very high strength route. Also, the marking connector square pin 1 header footprint is a hinderance in this design and leads to strange pad trance connections and mechanical stress raisers, better to remove it and put in a round pad stack. You can use a solvent to remove the potting material. You could try IPA at an elevated temperature in an ultrasonic cleaner?

[@GregUzelac]

So interesting!!

[@talsit]

Does the epoxy get soft under heat? Could you maybe heat the probe head while it's under compression?

[@Thesignalpath]

I tried, it does not soften up at all.

[@Slesar.]

Нужен обзор по пробникам тока для проводников на печатной плате. Need an overview on current probes for conductors on a printed circuit board.

[@fluffy_tail4365]

Wow it seems crazy to me that that hand assembly is the only way to make those, I wonder what prevents from moving most of the component on a single pcb and then slotting the half transformer on top of that

[@toine512fr]

So the issue is the plastic shroud preventing force on the transformer missing then?

[@Thesignalpath]

I think the plastic shroud missing because it must have broken due to excessive force...

[@toine512fr]

@@Thesignalpath Indeed it makes sense.

[@KD0CAC]

Thanks again Do you have some suggestions on dissolving different potting ?

[@JohnDoe-es5xh]

Some ketones can dissolve pottings. It's worth to try it out at these prices to analyze their compositions.

[@KD0CAC]

@@JohnDoe-es5xh , A longer version of the question - identifying a number of the types of potting --- to match the process & solvents , or physically remove only ;( . I typically attempt to repair far cheaper devices ;)

[@jyliang1917]

Just submitted a PO for this probe...

[@Txepetxcc]

Hi ! Can you make a video about Built in test , self-calibration and so on ? How can this devices or mostly oscilloscopes do that? Thanks!

[@ebb2421]

tune in next time where Shahriar builds a new and better head...

[@WolfmanDude]

Does this unit really work with a hall effect sensor? I thought hall effect sensors only work below 1MHz. Maybe they are using the hall effect only for DC and low frequency? Does anyone know?

[@scuba31]

Yes, the Hall effect sensor is just for low frequency and the transformer is for higher frequencies.

[@jeditoto3441]

Not sure how to send you messages. Could you make a video on the DMM chip 7106 and theory of operation- designed by fluke then simplified by Intersil? Very few people actually have your pedagogical skills to explain how the equipment actually works. It would help people like a lot to repair our tools 😊

[@jimomertz]

When they say “No user serviceable parts inside” they mean it! Yikes those are expensive 😮

[@testboga5991]

you could do a micro CT before disassembly! or dental x-ray as poor mans version.

[@BaconbuttywithCheese]

Do I detect a possible diy solution on the horizon?

[@McTroyd]

Oof. $2K for a probe head. A _good_ probe head, but still... My whole car isn't worth that much these days.

[@kaiklein6329]

❤ try to solder the parts tohether with wires to see if this is the issue, next step is to solder correctly together, tek did it so you can doit too.

[@xfxox]

The pricing is really odd comparing tech level with, let's say, an iphone

[@macosm7818]

Design adequate to the price. Nice older model. The plastic has cracked from age.