EEVblog

[@gordonlawrence3537]

Way back when I started everything was still through hole. The early surface mount kit was damned expensive because not only were the components more expensive (especially passives) but more critically so was solder paste. So a box of 2000 metal film 1% through hole was about £2GBP in bulk. A reel of 1206 5% was more like £12 so about £0.012 per resistor. Then the solder paste was about £0.008 for a 1206 too. If I remember correctly a carbon printed resistor was more like about £0.01 so around 50% cheaper. Prices for surface mount didn't really start dropping till about 1990. . The solder paste dropped in price radically around then too as did wastage. In the early 80's the wastage for solder paste was near 50% by 1990 it was less than 10% on a good line. I have known companies change pick and place machine just to reduce paste wastage in ultra-high volumes.

[@olivialambert4124]

Thanks for the feedback. I'm glad I found your comment.

[@LittleRainGames]

12 euro for a reel? That seems cheap.

[@snowstar3560]

@@LittleRainGames Did you missed the part where he mentioned through holes are the norm? So somewhere in the 80s, yeah 12euro is pretty expensive back then.

[@BenjaminGatti]

reverse the leads and remeasure. if the value doesn't change, then it's probably not influenced by active components in circuit.

[@LittleRainGames]

Good to know. Thanks

[@hectorbacchus]

I worked in bare board pcb manufacturing for years and the carbon printing was and still is an option. Hell, we are even embedding resistors into the pcb during the inner layer lamination which is a nightmare to do.

[@strangersound]

+hectorbacchus Sub-surface mount? ;)

[@hectorbacchus]

+strangersound they are embedded carbon printed resistors. The problem is getting consistent values.

[@LittleRainGames]

Why is it so expensive to add carbon stuff to boards? Maybe its only expensive for low volume/prototypes though..

[@mikeselectricstuff]

There are very few chip parts, so maybe instead of a conventional 2-stage chipshooter then fine-pitch place machine process they did it all on the fine pitch machine, which could be slow for chips. Don't know if they had fast shooters back then, but my guess would be placement time.

[@ssj3gohan456]

Not just back in the day, but now even there is such a thing as screen printing carbon epoxy. That can essentially be done for roughly the same cost as soldermasks or silk screen text - it's really not expensive at all. The reason carbon printed resistors are usually associated with high cost is the fact they're often used in hybrid substrates (ceramic and aluminum with film copper). But they don't need to be expensive.

[@goyabee3200]

My guess is that the resistors were printed long before the rest of the components were placed, maybe as an option during the pcb printing/manufacturing process.

[@JoeDesbonnet]

I worked on a project about 20 years ago which involved using a laser to trim these printed resistors. Even though there may be a high variance from the printing process, when trimmed I guess you can bring that down to 1% or less. Can't remember the details unfortunately.

[@davak72]

Interestingly enough, I saw these printed resistors in a 10USD piano keyboard toy just last week. On the pcb with the contacts on it, there were no through hole or surface mount components whatsoever! Just printed resistors (for contact multiplexing), and a ribbon to the main board.

[@VndNvwYvvSvv]

Yeah, going SMD-free is another story. That eliminates a very expensive pick-and-place machine entirely.

[@fuzzybobbles]

I'd say it was down to space. The printed resistors are all tightly packed together and looking at the size of the other passive components, I'd say it was a space issue and they had problems with the solder sometimes accidentally joining two pads together, so they went for printed resistors to elimiate this.

[@compwiz00]

The one on the end which you cut had more carbon than the ones to the left. I imagine that is why it read lower than the others.

[@arious-uk]

I really like these type of videos because they educate me about something I knew absolutely nothing about and secondly it starts a good discussions about the video contents. It would have been nice to include how these are actually printed onto the pcb. I assume it's done using similar technology to inkjet printers? Keep up the good work Dave and looking forward to another power supply video!

[@WhitentonMike]

I saw a video a while back where someone had made a DIY CNC and they were using it to route the copper off the board to make the PCB.

[@sssur32]

Sometimes screened carbon resistors were used to allow laser trimming of the resistance. The final resistance was automatically tweaked to the correct value once the circuit board was fully operational. As long as one or more of these laser trimmed resistors were required, then all of the non critical (ie, pullups,...) resistors were screened on at the same time, to reduce cost. Not sure if this is the case with your PCBA though. Many precision hybrid modules used this method.

[@wadehicks9270]

Had a guy bring a VCR in for service in the early 90's with those. however I don't recall the brand.

[@KrisBendix]

For developing something it could work well. When finally get it right, you can measure it and then replace. Only problem is that we can't get that carbon resistor on in the first place. :D

[@Morkvonork]

Because they are often in series or parrallel with other stuff which you would measure with it too.

[@emactan]

Got lost for a while and wondered why Dave was giving a depth of field primer :)

[@MrDuncl]

I saw this done in a car radio back in the early 1980s before anyone had ever heard of surface mount. Saves space and money but the resistors are about 20% tolerance.

[@maxsnts]

Resisters in Parallel decrease overall value, so because most components have some resistance, and the circuit around the resister will cause the reading to be always lower that the resister value itself. Most of the time the only advantage of measuring a resistor in circuit is to see if its open. If the measured resistance is way bigger than the resister you can be confident that its a bad resistor.

[@codebeat4192]

I think they used them because of the flat cable to keep a flat surface.

[@TheProCactus]

I have a jar ive been ashing in for 3 years, Ive mostly kept it as a slurry by adding small amounts of water. I wonder if it can be mixed with epoxy or silicone and put in to a squeeze tube. it could be handy for digital stuff.

[@WhitentonMike]

You'll be measuring other components in the circuit as well. Example: 2 resistors in parallel.

[@thewii552]

6:16 not only the semi-wide aperture but also the fact that you're doing macro AND the fact that your opteka lens isn't the highest quality probably has to do with the really shallow depth of field. I've noticed the sharpness goes way down around the edges of macro stuff, and it gets distorted as well...that might have a part...But you usually don't get that shallow depth of field shooting at 2.8 unless you're on a very long zoom or macro.

[@EEVblog]

What about component cost in 1986? It would not have been the same trivial amount as it is now, I'm sure.

[@apexmike849]

SMD components were more expensive back then - it was new tech.

[@dtiydr]

8:50 They are but pretty much only for laser trimming in equipment to get extrem precision or such.

[@AntiProtonBoy]

If you are willing to ignore the IC markings, the board looks surprisingly modern for '80s tech.

[@emmettturner9452]

I can think of a couple reasons for doing this though neither apply to this board. I first saw it when I took apart my Nintendo Entertainment System NES-004 game controller around 1986 which is when almost everything was still thru-hole so the printed resistors were literally the only SMDs. The button contacts were also carbon printed so I wrongly assumed the pull-ups were vestigial turbo button contacts. Years later I realized they were resistors and decades later I realized why: They wanted a single-sided PCB that was completely flat on the trace side for the carbon button contacts. The rest was just a 4021 shift register (DIP16) and two resistors (axial leads). The only components that would fit on the flat bottom were the printed pull-ups, and since they were already printing carbon film for the button contacts they likely saved a bit versus populating more thru-holes. Since then I’ve seen it used in another controller with a Mylar flex PCB that couldn’t have any rigid components, so there’s reason number 2. I’ve also seen it used in later designs to jump traces under a chip without a thru-hole or via (single sided).

[@Satchmoeddie1]

I have a slowly growing business of replacing SMD resistors on a certain car's ECM sub assembly board's resistors. They heat up enough to soften the solder. Maybe that is the issue. I do not want to go into a lot of detail of which car and which part. After the earthquake in Japan spares were not coming in for weeks, so I started putting 1/2 watt 100 ohm and 33 ohm resistors on these boards. That is my guess. Many were in parallel yet they still got very hot and fell off.

[@WhitentonMike]

If the board is under any thermal or mechanical stress, you could start a hairline crack. I'd try either polishing it or just replace it with a actual resistor.

[@Travellerwiz]

Is it possible that there was a sag in supply during the weeks these things were being manufactured? Maybe resistors became scarce and they used printed ones. Just a thought..

[@EEVblog]

They are now. But 26 years ago? Placement time had a large factor I'm sure.

[@QlueDuPlessis]

Those carbon printing was probably silk-screened. It would have been quick and cheap for non-critical resistors.

[@koocoo]

i wonder how long a carbon printed lasts vs a block

[@listerdave1240]

I'm guessing that when the boards were manufactured they may have been panelled together with the keypad which would have carbon patches for the keys, essentially making the printed resistors cost nothing extra. Since the resistors are essentially part of the board manufacturing process and not the assembly stage they would have been laid down before the boards were routed to be separated and long before the boards got to the pick and place machine.

[@muppetpaster]

there are no carbon patches for the keys.

[@scompo]

They could have also pick and placed the components and route out the boards tabs later maybe

[@matthewpeterson5159]

here's my question: what's with the gold circles in a bunch of the tracks? (oh they're test points probably?) (also holy cow those vias are huge :P)

[@harviecz]

This is actually quite cool. Wonder why PCB factories no longer offer this, should be very easy to do... Especialy for small batches and prototypes, where i don't do pick and place, this might save some time needed to solder small resistors.

[@skson5838]

Of course its the shape of the folded cable that is the reason. The surface mounted parts would take up more space.

[@AngDavies]

The depth of field issues could probably be solved with some kind of tilt shift setup, probably expensive though

[@RetroGamerVX]

Nice to know that even the best of us learn something every day Dave :o) When you say 'Panalisation breakout marks', what do you mean by that please? You've definatly got that thumb up with this one :o)

[@prathamkalgutkar7538]

When you Panelize multiple similar Boards into single one, the Board shape is routed into the board and small Tabs are left to keep it Connected, layer these tabs can be Cut using pliers to Seperate them leaving a Indentation or mark on edge of the board

[@hardwareful]

considering shooting things at an angle: check out tilt lenses, they are lots of fun :-)

[@WhitentonMike]

There needs to be an open source pick n place machine like the Maker Bot. I so want to build one.

[@UpcycleElectronics]

Is the carbon printed resistor the same as carbon button contacts? If panelized, boards with resistors+buttons would have made cents :-) Dave, I'm here looking for guidance on carbon button footprint creation. I have no plans to make. I'm just reverse engineering a design and in need of 'proper footprint design' guidance. I haven't come across anyone talking about having contacts made with this stuff while searching YT. I'm messing with a Nintendo Game & Watch (little STM32H7 based holiday trinket). A bit of effort went into button layout and design as contact detection is critical in the application. So what are the pros and cons of resistive contacts on pcb and silicone buttons versus gold/nickel/(soft/hard plate) and carbon/silicone, versus domes and/or devices? I'm interested in the differences in speed/noise/accuracy/mechanical (advantages/disadvantages). I know about the references to tactile dome buttons here, but I don't think the EEVBlog library covers this one specifically yet, at least not from the pcb design side of the subject. When is this even a practical thing to do? Do Proto Board Houses do this on the cheap? Probably pointless to comment like this in the sea of comments, but thanks for reading if you made it here. -Jake

[@ArduinoTronic]

The problem is not the place bit but the pick - ie component handling.

[@KarmaElectronics.]

i have a panasonic radio from the 80's with the same resisters.

[@monchiabbad]

You could have measured the resistors with the probe midway the resistor path.

[@Jawst]

I would have expected them to use a screen printing type application for carbon printed resistors.... Perhaps oven bake to set

[@goamarty]

Heat can not be the issue. Solder melts at 183°C or more (if it is öead free). You must not have temperatures like this on the PCB for prlonged time.

[@BarsMonster]

Probably these days many chips have programmable pull-up/down's, so there is less need for non-precise cheap resistors...

[@mudaserawan1457]

I can use the connector-less ribbon cable as shown in the carbon pcb on mybdesign. What is this type of ribbon called? How do you select this? Can i get this in my desired length? Do i have to strip of ribbon cable ends manually for each board? I need strong and flexible connections between my two PCBs without connectors and need minimmum verticle clearance.

[@Fiercesoulking]

I think we will see a similar technology again at least when they 3d print the whole boards as one piece.

[@WhitentonMike]

Thanks, I will check them out. The previous ones I've seen were more early prototypes and not all functions worked yet.

[@Desmaad]

Why can't you test resistors in-circuit, anyway?

[@WhitentonMike]

We know. Makes it more personal and interesting. How many electronics channels do hunter spider videos? lol I thought it was cool!

[@spacepirateivynova]

should be shooting that boards straight on rather than at an angle to emilinate the whole problem of grain vs. blur alltogether.

[@mikeselectricstuff]

resistors are nearly free but placement time isn't

[@TheEgoblitz]

Same here, I was looking at designing a single board version out of T-Slot extrusions, belt drive like the new Maker Bot, fast servos and solder paste shooter built in. Only thing stopping me is the software, I'm still learning the Arduino stuff.

[@EEVblog]

Interesting! photo?

[@TECH_GEEK10]

I'll look for a camera, but I only have a cheap 240p webcam I think

[@steamcastle]

Hey Dave have you ever worked with thick film ?

[@voltare2amstereo]

back track to the 80's or so, surface mount may have been more pricy

[@mikelemon5109]

what is that component at the left j155 at the beginning of the video?

[@hermanngoring397]

Derpy Herpy Flux Capacitor

[@mikelemon5109]

and if I blew one up can it be replaced with a normal electrolytic capacitor?

[@barbadolid5170]

+Hermann Göring Hahaha. Der Reichsmarschall ist immer richtig

[@hermanngoring397]

Derpy Herpy Yes :)

[@theSuitMusic]

...or a diy photolitography machine. Imagine being able to just print out IC's.

[@asawula]

What would your budget for one be?

[@javedkhan0258]

You are the best! Well explaned

[@michaeleric01]

video kept flipping between two pictures

[@Satchmoeddie1]

What were they thinking when the car company decided to save 13 cents and make a board that routinely has the power resistors heat up and fall off when the solder melts. It is a $40,000 car! These were essentially pull down resistors too. I do not think the carbon will melt off the board like a soldered SMD device. I can ask some of the Motorola guys or a another EEE guy who has decided to become a dentist. When I ask why he became a dentist, he said, "Let's see'em send that job over seas".

[@JohnDoe-qx3zs]

Well for the past 5 years I have seen 'em do exactly that for the higher end dental work. Patients travel tourist class to somewhere with cheaper dentist wages for expensive non-emergency procedures such as getting permanent dentures, gold teath, filling replacements etc.

[@smorrisby]

This was not unusual even in late 70s consumer audio electronics. I first saw it in early 1980 in a Sanyo ghetto blaster.

[@smorrisby]

en.wikipedia.org/wiki/Boombox

[@DJBREIT]

***** WOW you just made me feel so old.

[@InfernosReaper]

+smorrisby Oh, I was expecting that smg in Payday 2.

[@strangersound]

+imthekingofrs2 A way of life. ;)

[@KrisBendix]

An interesting way of adjusting resistor... scratch a bit. :D

[@magnuswootton6181]

if the integrated circuit is just resistors, the whole thing can be carbon and nothing but it!

[@WhitentonMike]

I thought he meant PCBs. Printed ICs? Not a chance DIY.

[@ricerob]

it's digital logic so what if the value is off by 30% 40% even 50% every thing is on or off as long as you get with in 1/2 volt + - of 50% supply voltage the logic will work plus every in put is buffered by the logic's operation as to why use printed restores well you don't have to stock them you don't have to place them

[@muppetpaster]

you still have to stock and apply carbon goo

[@leonerduk]

Right, but you only have one kind of carbon goo, and that can still make a variety of different values due to the trace geometry when you print it.

[@muppetpaster]

@@leonerduk Well the point was no stock and no placement, The way it works was not mentioned and already known.

[@TECH_GEEK10]

I have a 2001 HP printer pcb with ONLY carbon resistors!!!

[@bloubear2557]

I hated these. They're amazing and all. But imagine them peeling or rubbing off and having to replace them

[@WhitentonMike]

Very interesting.

[@hellnawnaw]

That's really damn cool

[@EEVblog]

I get easily distracted! :->

[@Floydarn]

Smart

[@thewii552]

haha, the chip making machine again. I hope that happens. Dave be quiet, I'm on Chris's side.

[@Adithyagumparthi]

The side effect is know as bokeh

[@DiegoColl44]

*****