Serial Controlled VFD Samsung HCS-12SS59T - am I missing something?

[Ysjoelfir]

Hey everyone,
I am banging my head over that thing... I have been looking through my junkpile and found two of those rather nice (considering that it can be controlled by only four pins (+ supply)) VFD with integrated controller. Now, I looked at the datasheet, which can be found here, I am a bit confused.
First, whoever made the datasheet seems to have forgotten the nomenclature used in the pin description, as VDD and VEE are in the pinout diagram all of the sudden called VDD1 and VDISP. 

Also, if you look at the suggested wiring scheme on page 3, there are a Zener diode and R3. No values! why? Am I missing something obvious?
I mean, it is obvious that those are used to bias the heater of the VFD to a known level in regards to the grids, but which bias does that thing need? I can't find anything in the datasheet...

Also, when I connect it the way it is shown (without the biasing or with F- connected to GND doesn't matter), the VEE/VDISP section seems to behave like a short circuit, which results in R4 letting the magic smoke out. As there is no wattage given for R4 (and no usual current consumption of VEE) I can't figure out if it is just damaged or if I did anything wrong, or maybe if I just need a bigger R4...

What the hell am I missing here?
Thanks in advice.

[rstofer]

Doesn't the zener diode just set up EK - 2.0V?  Diagram on page 5

[Ysjoelfir]

That looks definitely possible. So I will go with a ~2V zener. But that doesn't explain why my resistor is burning up. Hrmpf.

[rahalph]

Hello Kai,

Have you been able to make that thing work? I haven't!

I'm going with 2V for the zener diode, too (respectively I've used 3x 1N4148 instead), so that the filament voltage has an offset of ~2V.

I've used a 5W resistor for the 470R, because it drops more than 2W. I don't even understand why they recommend 34V. With 18V and above all segments light up.

And this is exactly the problem. All segments light, the oscillator swings, but whatever I send to that dumb thing, nothing else happens.

I've tried SPI and also manual pin control. I've tried any bit patterns possible as a command (even invalid combinations). I'm sure that I got the reset and initialization sequence, the signal logic and level and timing correctly, but whatever I do: nothing!

Normally I would say that the not-connected pins (the cathodes) should have a signal from the controller but they are always low. I fiddled with the clock generator (different capacitors, even a frequency oscillator) without any success.

I would be gratefull for any advice that keeps me away from 

Best regards
Ralph

[Ysjoelfir]

Hi Ralph,
sadly I ended up at the exact same point as you are right now. With a 2.4V Diode (didn't have 2V) it lights up all segments but doesn't respond to ANY signal I send to that damn thing. I also tried many things, the same as you did, but with no success. I decided to put it down for the moment and revisit it later, had to get done with the project I intended this display to be used on, so I used a normal LCD instead for the moment. But I don't want to give up right now on that VFD... It sounds just to good to be true, a VFD with serial interface...

[cBaer]

Hi there

I have tried to get the HCS-12SS59T running today but ran into the _exact_ problems you guys have.

So i was questioning the Datasheet provided by pollin.
I'm a german and ordered there a few times, its even where i got the two VFD.
And that particular shop is not known for superior quality products.

So realizing I'm wasting my time I went hammer-time

The chip in this VFD is marked OKISEMI'08 ML9289-
(yes theres nothing behind the minus...)

A quick search got me this.
http://www.lapis-semi.com/en/data/datasheet-file_db/display/FEDL9289-01.pdf
Didnt read the datasheet yet, nor investigated whats going on.

Hope this is any help for you 

will return to this project some other day.

[cBaer]

Update:

THE DATASHEET PROVIDED BY POLLIN IS INCORRECT!

The polarity of V_EE is reversed! See the datasheet I linked before. With clipping-diodes this explains the high R4 load.

What I changed to get it working:

- connect V_DD to your Arduino, (this must be powered up before V_DISP)
- Separate +5V -> filament -> 2x 1N4007 -> GND
- +33V -> 470Ohm -> V_DISP

I'm not sure if the filament setup is correct, but it works.
Also sequencing is important. Not sure if the 'malfunctioning' mentioned in the ML9289 datasheet is permanent.

[Ysjoelfir]

Nice observation cBaer! I also know of the "quality" (or better, lack of) of some of pollins products, so I doubted some of the information, too. I was suspicious about the VEE but didn't want to investigate it just at the time being busy with other stuff. Thanks to your investigation and someone writing me in private I got it to work, which is quite pleasing. The code it needs is weird, at least in my oppinion, but it is possible to drive that thing. I will keep working on it since I have some spare time right now.

[qrti]

Hi,

I also fiddle about the Samsung Display and can't bring it to life.

5 V first and then 36 V are applied, the oscillator works, with a separate voltage on filament pins I can light all display segments. But whatever sequence I send, nothing happens. I attached a simple Arduino sketch, I'm working with.

Ysjoelfir you said

The code it needs is weird, at least in my opinion, but it is possible to drive that thing.

I would be glad if you could give me a hint.

[ale500]

2, 3, 4 people from Germany and an unknown display. I started reading this thread and after the third post, I thought this smells like Pollin... I read further and someone mentions Pollin. Why it doesn't surprise me... wait while I have a look at the stack of cheap but unusable displays I got in the last years... where did I get them... I think it was...yes... Pollin !.  They sound like good deals, they are a waste of time and money . I got mostly 320x240 and 240x64 panels, several kinds of them.

I just wanted to complain about their display "Sortiment".

[qrti]

did it ... at last ...  ...

https://github.com/qrti/VFD-HCS-12SS59T

[macboy]

Nice work.
I love VFDs. Would anyone in Germany be willing to reship a few of these to Canada?

[qrti]

Pollin overseas department?

https://www.ebay.com/itm/132336065799

[Rolo]

did it ... at last ...  ...

https://github.com/qrti/VFD-HCS-12SS59T

Nice. I see you are feeding the filament with DC, isn't that a no no in VFD land? Something with uneven lit up display. I'm also working on a VFD display and feed the filament with a 555 100khz osc and modified coil.

[qrti]

according to

https://www.noritake-elec.com/technology/general-technical-information/vfd-operation

AC and DC filament supply are both possible, with smoother brightness advantages for AC supply. I tried AC with NE555 and coil and stuff for this display and there was exactly no visual difference to DC supply. Even more, surprisingly enough and against all theory, both modes show slightly darker digits on left _and_ right ends. So I surmise that the integrated ML9289 Lapis chip compensates outer digit brightness by duty timing. Are there any other disadvantages for DC? For example relating to lifetime?

[Ysjoelfir]

hey everyone, nice work! Seems like we are finaly getting somewhere! Sorry for not answering earlier qrti, I didnt watch into here until now. But it looks like you got the problem solved, nice!
Regarding the heater voltage: it works worh DC without any problem, no visible difference in brightnes. You can even derive the heater voltage from VCC, works fine.

Nice work.
I love VFDs. Would anyone in Germany be willing to reship a few of these to Canada?

If you are still interested write me a message, I may be able to do that with my small business, just give me a quote how much you would like to have.

[texaspyro]

Are there any other disadvantages for DC? For example relating to lifetime?

Yes, Google "filament notching"

[qrti]

added coil-less AC filament supply to repository

https://github.com/qrti/VFD-HCS-12SS59T

no visible difference to DC supply in brightness distribution but hopefully avoiding "filament notching"

[Rolo]

Nice, I'm going to try your coil-less AC filament supply. Do you know the current draw of your filament?
I have a supply with a coil but I have to add windings to a stock coil manually, your circuit has standard parts.
Thanks for sharing.

[qrti]

The typical current for this filament is 125 mA at about 3.6 V RMS. With the given AC supply it draws about 100 mA at 5 V Vcc, but the display looks sufficiently bright.

[Rolo]

The typical current for this filament is 125 mA at about 3.6 V RMS. With the given AC supply it draws about 100 mA at 5 V Vcc, but the display looks sufficiently bright.

Thanks, just build a prototype on breadboard and it works on my display as well, have to tweak it a bit to get the right filament current. When I put 10 ohms in the Vcc filament it's perfect. I get a frequency of about 170KHz, your schematic says 17Khz, that's a typo I guess.

[qrti]

I get a frequency of about 170KHz ...

I checked all values, even changed the 74HCT132, but my oscilloscope shows as intended 17 KHz timing.

Replacing the HCT by a HC type provided a nearly 1:1 duty cycle. Following Noritake guide, a 2:1 duty is preferable, so the HCT type is first choice.

[Rolo]

I checked all values, even changed the 74HCT132, but my oscilloscope shows as intended 17 KHz timing

User error here, my scope's auto freq. measurement shows 170KHz, but when setting the cursors manually on the square wave it's say's 15.6KHz.
So it's seeing harmonics. I switched on the band filter on the channel and now the auto measurement also gives 15.6KHz. 

[Rolo]

I want to control the filament power from the display controller and came up with this, works on my breadboard. Resistor R5 can be used to adjust the filament power to a specific VFD display. It can be bridged for full power.