EEVblog

[@EEVblog]

More testing needs to be done on this of course, temperature testing for contrast consistency across multiple devices for example.

[@martinsalko1]

EEVblog just a dumb question, but have you tried reseating the display? I can see you have the "precision pinholes" which are a pain in the ass, as they never have the right hole size ( they're either too small or too big )

[@EEVblog]

It has nothing to do with the contacts seating

[@johnkeates9434]

Does the voltage adjustment in the video use the power pin on that holtek chip or the VLCD pin? And is there something to measure on the LCD side? Or does the method of powering LCDs not result in a good measurable voltage-at-LCD?

[@EEVblog2]

The supply to which Vlcd was connected without divider, so it was at max. Vlcd cannot go above the supply, so Vdd must increase.

[@EngineeringVignettes]

How do the other 4 look? Is the quality consistent? Cheers, - Eddy

[@JerryWalker001]

I have had that problem a few times and it was almost always board layout where some tracks had higher capacitance than others and it caused some segments to 'lag'. I also had one where there was some crosstalk between a couple of tracks.we generally avoided ground planes under the tracks controlling the segments.We always now specify a lower voltage display and feed it through a resistor and this allows us to select the right values. We also find variations between different vendors for the 'same' design.

[@vant4888]

HT1622 is outdated crap, STM8L could handle those LCDs much better. The problem with segments is that the IC supplies constant current to LCD waveforms but different LCDs have different segment capacitance (and probably line resistance). And it gets worse with higher line multiplexing. To solve the problem you can 1. Increase voltage 2. Increase current (STM8L can do this but low power consumption goes through the window) 3. Lower LCD update frequency 4. Add dead time after segment update to allow the segment to charge properly (STM8L can do this too).

[@amd64online]

Line resistance is definitely what I would be inspecting.

[@ZomB1986]

Just like v ant said, I think the common is snaking for too far on the glass, making the capacity and resistance too high and the signal too weak.

[@kruppin]

Have you tested with the other four lcds and it's the same, or with another batch lcd driver ic? Just to rule out other causes than bad manufacturing of the lcds.

[@EEVblog]

Yes, they are nearly identical.

[@johnfrancisdoe1563]

EEVblog But did you test the relevant LCD pins with scopes and meters to determine if it's the test board underdriving the pin or (as another commenter suggested) that part of the LCD having a different equivalent RC circuit than the others? Because such details may allow design adjustments to eliminate the issue over the final power rail tolerance.

[@themightyjosh]

6:39 Dave can't help himself drooling all over it.

[@ikbendusan]

hahaha i saw it and had to see if someone mentioned it in the comments

[@6Diego1Diego9]

it's not drool....

[@monad_tcp]

is so beautiful he's crying

[@6Diego1Diego9]

yea.. something like that

[@EngineeringVignettes]

No, he is just _spitting mad_ at the bad contrast at the nominal voltage. - Eddy

[@craign8ca]

I used to troubleshoot and repair electronic gear with LCD's and Zebra contacts. They were very difficult to get just right. Glad you decided to get them with pins instead.

[@PlasmaHH]

I see this phenomenon in a lot of LCDs, even in professional gear, it seems that nobody really cares about it.

[@TheHuesSciTech]

If the manufacturer can design LCDs to work at 3.3V or 5V supplies, and this one was specified at 3.3V but works better at 3.7V, could you get them to adjust {whatever it is they adjust to move between 3.3V and 5V} downwards a bit so that the production LCDs would work nicely at 3.3V and save you the extra regulator?

[@DemetryRomanowski]

It would be interesting to see another set of prototype LCDs spec'd for 3v operating voltage and see if it still needs that 300mV up to make it look good.

[@EEVblog2]

Yeah, that would be interesting.

[@fouzaialaa7962]

maybe because that particular pin is driving 5 segments !!! and not 4 like the others

[@PokeNationOnline]

Dave should check this!

[@johnfrancisdoe1563]

fouzai alaa Wait isn't this a 256 segment display with each pin driving either 8 or 32 segments?

[@ChipGuy]

Last time I got custom LCDs they gave me 15 samples containing 3 different versions. So I could pick the exact type for my controller (MSP430F449) since I did not have any chance to increase the voltage to 3.6 volts. However, tooling and handling was $1000 in my case. That's 10 times more. I am sure they could fix that for you using a different liquid crystal type, for another 5 samples and $100.

[@hitonokonoko4717]

It depends type of your LCD Display or LCD display Module. For example this manufacturer www.yaoyulcd.com offer 150$US for LCD glass and 150$US for Handmade Acrylic LED backlight. 5pcs of engineering samples are free and plus shipping cost. LCD cost is always same all model TN, STN, FSTN and VATN. If someone offer more than 200 USD, probably your LCD supplier are not a manufacturer. In China there is many different trader buy and sell Custom Design LCD Displays and modules. Also if your LCD display contains LED backlight, cost will be higher. Because there are many different type of LED Backlights. Injection mold LED Backlight tooling will be higher than standard Acrylic cut LED backlight. In addition, manufacturers have a minimum order quantity for mass production. Unprocessed LCD glass dimensions are approximately 360x300mm. The manufacturer must take an order that will correspond to a minimum of 600 pieces of glass(each glass 36x30cm). According to these dimensions, you can calculate your minimum order quantity.

When you have 1/8duty, does it mean that you have 8 common masses ?

[@jordanjohnson714]

Can't wait to see this project when it is finished.

[@iamdarkyoshi]

Where'd the DC bias test video go? It just dun poofed yesterday

[@EEVblog]

It was a figment of everyones imagination...

[@FennecTECH]

No it wasn’t.

[@FennecTECH]

I Bet the LCD was fine and he has to hide them to not destablize the lcd market

[@EEVblog]

Because the "bad unclear" answer is more fun watching people speculate :-P

[@VerstehenSieMathis]

Maybe we can get a short version on EEVblog2, or at least the result commented to the old video? We're curious!

[@russellhltn1396]

Two comments: What's the voltage on the drive chips themselves? (anything significant lost in the traces? I've learned that lesson the hard with way TTL stuff.) Second, was it clear from your communication with them that the 3.3V was the drive level to the driver chip or the drive level to the LCD? I'd think they'd be expecting it to be the drive level to the LCD. Otherwise, they have to work out how much drop there is the the drivers you're using - which may or may not match the information they have.

[@jojoposter]

Yeah it looks like the lcd is not seeing 3 v3. But i'd think raising the vdd and fixing the mistake shouldn't pose much of a problem.

[@frollard]

At the higher voltage the refresh 'attack' and 'recede' times seem to be much much improved also. It slowly fades in at the start of the video on changes...and sharply changes at the 13:15 mark.

[@bsb0011]

Dave, for those of us following along at home, is there any chance you will be posting a complete set of submission and design files for us to use as templates in our own design?

[@EEVblog]

Only when it's done.

[@user2C47]

The μSupply should be patented.

[@EricBuschdet]

Can you disclose any details on the upcoming μSupply? Is there a forum thread?

[@EEVblog]

Nope, no forum thread.

[@noland65]

There may be more to it than just that single line. Mind the rightmost digit when it goes to a "2" exhibiting different behavior than when displaying a zero. (E.g., the top right segment is solid on 0, but faded on 2.)

[@EEVblog]

Yes, but that single line was the worst.

[@kruppin]

Yes, but when it's showing a five there instead of the zero, it looks good, which makes me believe it's a problem with the driver ic. Would be interesting to see if it behaves the same with another batch driver ic. But since a bit higher voltage solves it it really doesn't matter anyway.

[@sasan.]

Tanks Dave for such huge information you provide here

[@TheDefpom]

If you need someone to bug test and review the uSupply once it is ready, let me know 🤓

[@daab889]

since not all of the segments are faded under 3.4 volt, lowering the voltage for contrast setting seems less ideal.

[@OmarMekkawy]

The 3.6V might be got from an LDO or a tiny DC-DC converter chip with some resistors. But if you are using a lithium batteries this might be a problem because its voltage drops under load.

[@reubenbridges]

Damn, not a one hung low product. 3.6 volts is best, Xcellent video.

[@donreid358]

Surely you checked the waveforms on the LCD pins??

[@randomelectronicsanddispla1765]

Yeah! We finally get to see the finished product!!!

[@mata7648]

The main reason I like your videos is that OK is never good enough. As an analog circuit designer, I am always trying to achieve one more digit accuracy. A lot of digital circuit designers do not understand this obsession.

[@Kalvinjj]

Well can't blame them since 0.4 is close enough to 0 and 0.6 is close enough to 1

[@userPrehistoricman]

pi = 3

[@josech5743]

In The End it doesn't even matter

[@PyroRob69]

If it's worth doing, it's worth overdoing.

[@NivagSwerdna]

@12:50 I like that too. You're being too tight at 3V3. Drive it with 5V and add a VR to VLCD. Job done. Keep the pins they are nice.

[@EEVblog]

Yep, easy solution.

[@zusurs]

Is there any reason to use these segment displays, apart from low current draw applications? I mean - these days you can get full color 320x240px or even higher resolution raster LCDs with built in drivers for just a few dollars, and you can display literally anything on those. I mean - cost- and functionality-wise there is no reason making your own segment LCD these days. Am I missing something?

[@JKPhotoNZ]

I'm guessing that the main reason (other than power cons) is the effective resolution. To match that res may require a lot more pixels than 320 wide. That said, I do like displays that simply use an I2C or SPI interface!

[@tomvleeuwen]

In a previous video, you made a setup where you DC-biased an LCD. Is that LCD still working? Any updates?

[@darikdatta]

He posted the conclusion video to that experiment yesterday then took it down before I got to watch it.

[@YTANDY100]

@random did it fail completely or could it be used again with ac bias ? ,did dave try it ? , i would like to know this as i saw the notification in the bell list but the video is gone :-(

[@C0wb0yh3nk]

See EEVBLOG2 the video is called "shit happens"

[@gotj]

Hey, Dave, get in touch with the guys of ADVChina, perhaps they can arrange a visit+video of that LCD factory. They live in Shenzen. Or with Charbax, but I'm not sure where he's located.

[@infinitepower6780]

Or Strange Parts

[@SianaGearz]

Ooooh Strange Parts is so sweet

[@electricengineerwannabe5499]

XD Cmilk

[@RobertSzasz]

I assume the process they use pushes just about all the per part costs in the design/Cad and final processing to attach the pins. (A few cents per piece for the materials and processing if physical masks aren't needed)

[@worroSfOretsevraH]

For me it's a bit confusing the exact same font (and small V/A letter at the end) of the two display lines. Either you have to be really familiar with the display, or always check the small letter at the end to see weather you are looking at the voltage or current display. This is why LED displays use different colour blue/red for the two values.

[@irnavas]

Hi Dave, It could be that you are operating the LCD in a brighter environment than it’s nominal. Did you also specify the ambient brightness? Anyhow, the final product may benefit from adjustable voltage.

[@EEVblog]

No spec on ambient brightness. Reflective LCD's don't have a maximum brightness.

[@irnavas]

Since the dynamic range of the display is quite low, they should have some idea of the ambient light while building it for the specifications, e.g., it will become unreadable in indoors if it is designed to work under sunlight.

[@GabbeeAnimator]

The black parts looked like objects with a chemical that makes it fade away and appear but it’s actually the liquid crystal showing it

[@Skinkworks42]

Does that dev board have a Type-C connector on it?

[@willyarma_uk]

What would happen if you requested a change to 3.0V and asked for another set of prototypes? how much would that cost?

[@AMalas]

Willyarma would it cost more? I assume it's just a different material and the tooling is same?

[@karsnoordhuis4351]

You can even get custom lcd panels (with pixels) and even oled panels these days.

[@SwitchingPower]

This display is not compatible with a zebrastip because the contacts are on the bottom glass, for a zebrastrip you need the contacts on the top glass.

[@axellitoral1503]

Here we go we can make game and watch replacement!

[@wpherigo1]

Who was the manufacturer?

[@DM-fz3ly]

Hey Dave, David, If you guys are still working on your battery powered power supply then you might consider coding it such that it can keep track of how many Amp Hours it has supplied (consumed out of the supply battery) and therefore be a sort of gas guage on when the battery needs recharging. Nice display!

[@kristiandawe85]

Dave is so excited he is spitting on the lcd

[@pichacker]

Where did the bargraph go? And would the duty cycle have an effect on the contrast?

[@deerel]

How is 6 O'clock a tilted angle? By pure logic it would be to be looking straigh down on it, as in looking at the number 6 on a clock. Or what am I missing?

[@EEVblog]

www.pacificdisplay.com/lcd_view_angles.htm

[@deerel]

EEVblog Thanks! I learned something new, nice! So the "6 O'clock" thingy is not a viewing angle thing. It means that the display are made to be view from the bottom. In it self its not a definition of the angle. That makes more sense.

[@EEVblog]

Yeah, it's a bit of confusing industry specific lingo.

[@Graham_Langley]

+drlunderberg Many years back I visited a bleeding edge tech firm where they'd set a nice LCD clock into the back edge of the visitor registration desk. Problem was the display had a 12 o'clock viewing direction and was almost invisible when viewed from 6 o'clock. Pointed it out to their tech director but never went back.

[@6Diego1Diego9]

where's little Bobby Dezzla?

[@ilyakaryagin4754]

That's a nice bit of kit.

[@urugulu1656]

can this sort of issue be because of a half ass-design for the testing board? we both know that it doesnt need that much to drive segments (or in your case not drive them)...

[@EEVblog]

The "half-arse" design is using a dedicated LCD driver chip.

[@hossammoghrabi1021]

we need another video showing how to write software to drive this thing... which is more important than hardware

[@Daveyk021]

Would an off the shelf dot matrix lcd with the same dimensions have been a viable much faster option? How much more driver software development would there have been on your end? You had a complex cad development of the board layout in this case. What is the disadvantage of a dot matrix lcd? I imagine it’s not as sharp?

[@petrosmodellbau671]

Do you think a much smaller display would be even cheaper? I need a display to show the destination for my TT gauge model railroad trains - to fit into 15x5mm area and preferrably with I2C or SPI behind... Maybe LCD, LED or OLED...

[@makesomethingsmartertt474]

Hello Dave how is it going? could you tell us how to handle LCD of old calculators?

[@GeorgeTsiros]

i'd get a 120x120-ish lcd and be done with it off the shelf, future proof, works for other projects too

[@LucasGarrow]

Line the edges with small incandescent lamps for backlight 👍

[@MDealer]

If it would be an F35, then, no biggie, straight to release. But, since it's just a power supply prototype, should replace that display. Also, keep the pins. I've had so many displays lose contact with those flimsy rubbers over time due to either oxidation from high moisture or just the rubber perishing.

[@leocurious9919]

Yes, please dont go with those "zebra stripes". Touch them once and thats it, it will never work again... :D

[@roberthorwat6747]

A pleasure to watch!

[@ingmarm8858]

Is the Holtek providing the advertised drive voltages at the LCD?

[@xjet]

Or you can use a pixel-addressable dot-matrix LCD and control it with a microcontroller so that you're not limited to fixed segments. That seems to be the way many manufacturers go, especially for small runs where the software development is a "one time" cost and it means they can use off-the-shelf LCD units. It also allows for things like line-graphs, histograms, etc and other things which are impractical with a fixed-segment display.

[@ionstorm66]

No where near as readable as a fixed setup display. Also these will be cheaper if he orders a few 1000 at once.

[@EEVblog2]

Sure, if that's important to you that's the way to go.

[@Graham_Langley]

Looks to me like a it could be loading issue on that segment drive line - does it have a bigger area to drive than the others? I see the datasheet gives a 1:5 difference in drive sink currents at 3V and 5V which would seem to explain the problem going away with a slight increase in drive voltage. Edit: I see *v ant* beat me to it 3 hours ago.

[@tubelife70]

Curious that the data sheet specifies 64Hz. When did D/C power supplies begin producing sinusoidal output signatures or have I missed something?

[@EEVblog]

LCD's are driven by an AC signal.

[@DEADB33F]

Out of interest, what are the bulk prices like now you've paid for tooling?

[@mikelemon5109]

Do you sell those displays by any chance?

[@jamesmdeluca]

I presume the microSupply uses a rechargeable Li-Ion battery. The second VA display could be for a second output or a memory reading!!?

[@hrnekbezucha]

I was really shocked last time when you said you'll get the chips from AliExpress/baba. Did you do it? And where is the line of what you would consider getting from Shenzhen and what you wouldn't?

[@stevensexton5801]

@EEVblog, Were you going to build a product with the display?

[@martinda7446]

That just looks fantastic.

[@guguslangit2230]

hi... its posible to read data what com in LCD...?

[@HammadKhanYT]

Can somebody guide me how to drive custom LCD (generate waveform), Do we have to generate biase voltage in multiple common LCD and do we have to generate these waveforms ourselves? Is there link of a book or an online tutorial? Thanks

[@laconeccionvideo]

Very nice!!!, how much will cost order for an small production?

[@BOMBOVA]

FAB Interesting, much appreciated, still using them myself

[@krawutzimon]

Design approved by David Ledger :)

[@prashantandsapna]

i loved this 3 part series. big thumbs up. (i know i am 2 years late ) . was the final product ready and is there a video on it?

[@CliveChamberlain946]

Dave, will the uSupply project offer the 12-o'clock position? or will you take a vote on adding-in some kind contrast adjustment (as offered on a lot of electronic devices?

[@EEVblog]

No, we'll stick with 6 o'clock. Maybe add contrast, don't know until we do more testing. The chipset doesn't support contrast internally.

[@johnfrancisdoe1563]

EEVblog But imagine this in the mounting position of other bench supplies. 12 o'clock is the common viewing angle, not 6 o'clock like a handheld meter.

[@XanCraft21]

What website did you use to make the lcd display?

[@TwoHandsInPants]

What would be the price per unit when you order 5k?

[@jw200]

Minimum order is 5 pcs?

[@jamisusijarvi646]

Why you dont tell manufacturer of these?

[@HamzaKhan-es2fi]

Can anyone tell me what software would be best for desinging the UI for the segment LCD like that?

[@tookitogo]

Seeing the process is fascinating. But I agree that a backlight option would be good (not every usage scenario includes good lighting.) And taste aside, this should have had a UX designer's touch: the layout is super cramped, which harms usability.

[@johnfrancisdoe1563]

Antonio Tejada Not that cramped. And wasting 50% or more of UI space on blanks like so many "designers" do is horrible for usability. As users we waste so much time scrolling, zooming and straining our eyes because someone taught designers to duplicate the visual rules for large printed static text onto tiny electronic varying text. Look at stuff commonly accepted 50 years ago for lessons in what works for the human eye instead.

[@KevinOsborne1987]

More more more! My favourite kind of content

[@asagk]

Beautiful!

[@neome5970]

I am looking for a glass display manufacturer. May I know what's the name of that LCD company? Thanks!

[@kadiraynur2636]

Hello, I need a specially designed monochrome tn screen, I want to have it produced, but I cannot have it produced because I do not know how to draw or design it. Can you help me?

[@OC35]

What happened to the CC display part?

[@EEVblog]

It wasn't switched on

[@lordcrunk4790]

awesome, could an interactive screen like this be created but one that has the glowing red numbers on a black background?

[@EEVblog]

If you want glowing red, get LED or OLED displays.

[@Kalvinjj]

Would likely only take a new back polarizer at 90° from the current and transmissive/transflective with a red backlight. Would be cool as heck, but I can imagine this one here being more visible under all circumstances.

[@RobertSzasz]

EEVblog if they could do custom OLED displays at similar costs... Drool...

[@simonstergaard]

Strange Parts should go and take a look at the factory.

[@taranpalsinghpanesar3364]

Can you help with the coding method ?

[@JeepNetwork]

Hey I’m just wondering if you will ever post on EEVDiscover again?

[@EEVblog]

That's the plan, but I'm lazy

[@electronic7979]

Excellent

[@joopterwijn]

Do all have the same problem?

[@flo47able]

Thank you for using usb c! Hope that is not just for the prototype :)

[@nrdesign1991]

Super interesting series

[@onlyrgu]

133 US$...Whaat?? Yesterday i just ordered 5 4layered PCB for the same price(7day lead time)!! LCD looks good but I think information is cluttered all over the screen..Not that great from an UI/UX perspective!! Otherwise great job!!!

[@UtkarshPatel4]

Can help us in programming

[@jlebrech]

Would make for a cool cycle computer

[@tubical71]

i just had a chicken dinner, thanx dave....now i want one of these displays...! ;)

[@searchinghead]

I think this is a usb type c connector!

[@zer0b0t]

Looks great

[@kuningkucing6250]

4:24 thanks..

[@gladson9343]

Hello Friend! My Kawasaki Z800 motorcycle display broke and I can't find it, could you give me a solution that I can do?

[@ledhunters]

I would like to have your micro power supply but I am afraid it will be more expencive than if Fluke made it, like with EEVblog multimeter. Ask Fluke if they would be willing to slap their brand on it, it would be cheaper?