• Please review our updated Terms and Rules here

IBM 5153 (CGA monitor) line voltage range

Andrew, as you would also expect the only difference at 230VAC would be the additional heat produced by the snubber circuit, the power resistors near the main switching transistor. Once you complete that testing, and furthermore decide to use it, my suggestion would be to reinstall the PS without the aluminium cover. I tried this and it did not affect the CRT image yet allows a lot better air flow. Still, the 5153 main enclosure is poorly vented itself at the underside.

You can just imagine that the PS designer at Tatung, assigned to the IBM contract, tested their masterpiece on the workbench, open to the air, and proudly dubbed it "job-done." Nek-min, handed over to the packaging engineer, it's in a metal box with almost no air flow. All these PS PCBs get burnt around the driver area, the 230VAC version twice as badly as the 120VAC. Considering it was produced near the same time as IBM's 5150 PS, it's not nearly as sophisticated.
Thanks Paul, although mine does have more holes in the metal case that the standard one and there are some component changes for the 230v psu which appears to keep the heat down, but there is lots of room to put a small fan to blow air through the psu if needs be so will keep a eye on it for now, also the board inside is very good with little burn at all on it, but will keep a eye on it for now

So power supply back in the monitor, time to power it up but still with my adjustable psu, and I get this...


frame collapse! I guess when the counterfeit transistor went in the psu it spiked the 115v line as well as this WAS ok before that point!

in the words of farher ted....

SO more work to do!!!
I suspected Q302 and Q303 to be faulty and sure enough they both had a base emitter short! so out they came along with Q301 just to be sure! to my surprise they all tested ok!!! so looking again D302 and D303 were short! D301 was also tested just in case but was fine! again the standard 1N4148 diode is fine in place of the 1S2076 fitted originally so that's good as I have them! Powred the monitor back up and this tine got ....


signal in and


so looking good! now my signal generator is 50hz and the monitor just won't seem to lock to that! so am wondering if the monitor is 60Hz only or if I have a sync separator locking fault, if it is 60hz only that's fine as I bought this to go on to my fluke 9100 troubleshooter so hopefully it will work fine on that!
Bit more messing with it tonight, it should lock to 50hz or 60hz positive frame sync on the D type input connector pin 9, and my pattern generator is generating 50hz! traced the frame sync pulse through to pin 7 on the HA11235 chip and it is all fine! so it is looking like the HA11235 chip has suffered as well! might have been like this when I first got it working 2 years ago I cannot now remember, or it might have bit the dust in the flashover! either way looks like a new chip is needed! Now need to source one of those!
The pictures seemed correct in the previous post. Im wondering why should it support 50Hz? I was under the perhaps naive impression that CGA is 60Hz. Does it not work on the CGA card?
Last edited:
picture just rolls and wont lock, applying sync and there is no attempt for it to lock even though there is sync at the chip, no I don't have a pc cga card, this is on a CGA pattern generator which I can change the syncs to be positive or negative 1v or TTL compatible plus mixed syncs too, and R G B to be 1v or TTL compatible so it covers all types of monitors of this vintage, this was bought as a monitor which will work on my fluke 9100a troubleshooter for the likes of testing arcade boards. Most monitors usually will lock to 50hz or 60hz signals with just a tweek on the vertical hold pot to stop it rolling


I used to repair crt monitors for a living back in the 1990's

pattern generator

Last edited:
Ok, so how this IBM monitor came about!

I have one of these little beauties as I mess with old arcade boards and these are great for doing tests from the CPU point of view!


Now the programmers station where you can program your own tests into it was supplied with a monitor / graphic card and keyboard, but most sold never had that option fitted, so are now are in very short supply to even get the base unit! these were made early 1980's

so luckily the graphic card schematic was in the service manual! and this is the only picture of the graphic card I could find online


So after trying a fruitless search for some years I decided to see if I could remake the design, over about 2 years I finally had the pcb design finished, The original used surface mount parts but DIP parts were used in my design as strangely 1) i found it near impossible to route it into a 4 layer board, and 2) the main 40 pin crt controller chips were now only available in DIP package these days.

so I had the board made


and fitted all the chips!


and it now works great on a CGA philips CM8866


but I use this monitor for other things on the workbench mainly for seeing pictures from arcade boards under repair, so thought it would be nice to find another CGA monitor for the fluke display, so managed to pick up the IBM 5153 quite cheap as it did not power up! as I have said previously in this thread that is when I unfortunately fitted a fake counterfeit transistor in the psu which blew up more than was wrong originally!!!

so searching for the Tatung PMW-811 lead me to this thread as was the ONLY reference to this psu board on the whole of the internet!! so that's why I posted on this thread!

will be nice to finally get this up and running! HA11235 chip is now ordered!!
OK so the HA11235 chip arrived and fitted, I still cannot get it to lock on a 50hz signal, so looking at the vertical hold circuit against circuits of old tv's of the same era and the datasheet of the chip I came up with a plan!!!

R307 is currently a 6.2K resistor


so lowered this to a 4.7K, AHHA! vertical hold now locks perfectly to a 50HZ signal and the range of the vertical hold is good!!

so then changed it to a 5.6k , it still locks fine but the vertical control is very near fully anticlockwise, not a problem but if tolerances drift in the monitor when warm it might loose lock and as the control cannot be turned any further it might not be able to be locked.

this does show that the description of the monitor accepting 50hz is a lie! I think I will put the 4.7K back in it as the vertical control should then still have enough range on it to be able to lock on both, will test that when I can get a 60hz signal into it!

this mod should be useful to anyone in the UK wanting to use the monitor on other things at the standard uk 50hz field frequency!!

IMG_1403.jpg IMG_1404.jpg

now locks great! and back on its original HA11235 chip too!
Managed to find a 5.1k resistor so changed R307 for that, pleased with the result, locks great at 50hz and still a bit of turn left on the vert hold pot so seems a good mod to me! Now I can use my pattern generator on it and it locks!
Managed to find a 5.1k resistor so changed R307 for that, pleased with the result, locks great at 50hz and still a bit of turn left on the vert hold pot so seems a good mod to me! Now I can use my pattern generator on it and it locks!

You could add a resistor in series with the pot so you don't have to crank it as much.
You could add a resistor in series with the pot so you don't have to crank it as much.

no that would not have worked as the voltage on pin 8 of the HA11235 needed to be slightly lower when the pot was at minimum resistance, so the only way to do this was to lower the value of R307 to achieve this, this will not affect the locking at 60hz but at lease the monitor will now work on both vertical frequency's