IC identification help?

[Mitchekj]

Hi guys, any help would be appreciated.

I have an IC which, for the life of me, I cannot find a datasheet (or any information) for. There are no logos or anything on it, just the number:
KS58C5018
B0517

B0517 looks like the rev and date code... but cannot find that part number (or any parts of it) anywhere.

It's a 24 pin SMT chip, looks like SOIC or QSOP.
I'm not sure if it's a counter, microcontroller, or what. The pinout doesn't seem to match any of the common ones.

It's inside of a Dunlap heating pad controller, which has failed. Troubleshooting is a little hard since I have no idea what the IC is. (Only a guess, but it apears to be driving an SCR which switches power to the pad at intervals based on the heat setting you select.) SCR has tested good.

Hopefully someone knows something about this guy! Thanks!

 

[Resqueline]

The generic number is 58C5018 and I guess it's a microcontroller of some sort but I found no datasheet for it either, only chips for sale.

 

[Mitchekj]

Thanks for the quick reply!

Same thing here, could find where to 'buy' them... but no information on them.
I decided to go ahead and capture the schematic (it's a pretty simple circuit) in hopes that I could figure out what exactly was going on.

I've attached a zip file w/ a PDF (just a little too big to directly attach.)

Sorry if the schematic's a little cluttered.
I've double checked, and it appears to be an accurate schematic.. but I may have missed some connections since I'm not wanting to put a huge effort into it just yet.

Still not sure exactly what U1 is.
The LEDs (D3-D7) indicate high to low, and off, respectively.
Switch S1 cycles down, S2 cycles up.

Not really sure why they've got the SCR on the neutral line, but OK.
If I only had a datasheet for that IC, I'd have this thing working already, lol. I may need to take it into work so I can look at things w/ a scope... maybe I'd get a better feel for it then.

Any idea what U1 could be? (Looks kinda like a counter/register or something, but not sure.)
Thanks again for the response!

 

[Resqueline]

I figure the function could be done with a microcontroller but I believe it's a dedicated chip. The function remains the central issue however, and you will need a 'scope to check it out.
Pin 1 & 3 is power input. It could be around 5V going by the rating of C2. Check for absence of ripple.
Pin 5 is phase reference input. Check for a clean, clipped voltage.
Pin 11 could be a reference voltage, or a current mirror input controlling the LED intensity.
Pin 13-18 should have a square wave output in phase with pin 5, and related to the selected level. All but one should be floating along with the voltage on pin 12.
Pin 12 should have a low-pass version of the square wave, representing a phase delay.
Pin 19-23 are the LED driver outputs of course.
Pin 24 should have a short positive pulse at the selected phase timing.

The SCR should be drawn the other way around afaik. The neutral makes a nice common "ground" for the controller and the SCR's cathode.

 

[Mitchekj]

Thanks for that! Yes, I had the SCR backwards on that schematic, shoulda known better.

You were pretty spot-on there! Took it into work today and got the thing working.
C2 had miraculously shorted itself. Don't see that too often. The alum. electrolytic guy was bent into a rather odd position though... don't know if the dielectric film had eventually given way or what... but yeah, no more +5V for Vcc. Had some spare 470uF 25V guys in the shop, so all's well there.

That led me to another thing: I had it running on my bench for a while during the day to test it, and noticed that it shuts itself down after ~1 hour. I'd like to do away with that.

Here's what I can tell from snooping around:
When you first plug it in, it defaults to "Off." Pin 13 is high (+5V) and pin 24 is low. Makes sense.

You choose your heat level, and pin 13, 14, or 15 go high depending on the level you choose. (4 being hottest, 1 being coolest, 0 is off.)
13 high if 0, 3, 4
14 high if 2
15 high if 1
Pin 12 reads a ~+5V sawtooth at about 15kHz. Frequency depends on which pin (13-15) is high. Makes sense.

This results in pin 24 generating ~5V pulses at 60Hz. Drives the SCR, and viola... you get the heating coils turning the electrons into heat. (Via the half-wave rectified output of the controller.)

After about a minute or two, pins 13 14 and 15 open (floating, can see the 15kHz sawtooth on the floating pins) and pin 16 17 or 18 goes high, depending on the heat level selected.
16 high if 0, 3, 4
17 high if 2
18 high if 1

Now pin 24 outputs the same pulses, but goes low at intervals. I'd guess this is to keep the heater coils from overheating, and to maintain temperature. Interval depends on what heat level you have selected. (Makes sense.)

After an hour, it shuts off. The "Off" LED blinks, and pin 24 goes low until you select a new heat level. Pins 16, 17, 18 do not change state, neither does pin 12.

Pin 5 reads a 5V square wave, about 50% duty cycle, at 60Hz. Always.
Pin 11 reads a 5V sawtooth, at 60Hz. Always.

I'm assuming it's using one (or both) of these pins to time things. I don't see any way to disable the "auto-off" feature without affecting the output. It'd be easy enough to keep the SCR conducting, but then I'd lose the heat level control, and probably end up overheating the coils. Fire, etc. Bah.

Pins 6, 9, and 10 read +5Vdc. They are not connected to anything. Not sure what they're for, other than more inputs to the chip, like pins 7 and 8. (Could be interesting.)

I'm wondering if one or more of the Vcc / GND pins (1 - 4) are some type of enable/reset input for the timer. No way to know without the datasheet, or trial and error.

Any ideas? I appreciate all the help!

Out of curiosity, do you work in the electronics field or is this a hobby that you've been working at for a number of years?

(Also, attached an updated schematic.)

 

[Resqueline]

Ah, so it uses counters and burst firing, not phase control. I guess I would have suspected if I had bothered to calculate the time constants of the R-C links.

Trial & error is what I'd have tried to do, using a resistor between 100 & 1000 ohms to "short" out the different pins with. If it's an input it'll do the job, & if it's an output it won't do any harm.
No telling what may hide inside that chip..

I've been working in this field for decades, but it's also been a life-long hobby. Got my first soldering iron at 10.. Why, does it show?

 

[Mitchekj]

It shows a bit. I've noticed you give sound advice in most threads here, from what I've read so far. Nice to see people sharing their expertise.

I got the timer changed. (Good old trial and error.) It was the RC network on pin 11. Aparently the chip is counting the transitions, and 60 Hz comes out to 1 hour. I'm assuming C7 was a 0.1uF, that would give the time const. for 60 Hz. It's a small 0805 smt cap, so it doesn't say.

Anyhow, I changed the RC to about 6 Hz so now I should have a good 10 hours or so. It's been running for a few hours sofar, so it looks to be a good "fix." all other functions are working as normal.

Thanks again for the help!

 

[Resqueline]

Heyhey, that's good work! So pin 11 it wasn't related to the actual mains frequency, only happened to be close..

I changed the clock frequency of a calculator that way once that I felt turned itself off too quickly..

Thanks for the appreciation btw. but you have done a good job here yourself now.!

 

[Mitchekj]

Hehe, thanks. Come to think of it.. that is an excellent idea. Same deal with my calc. I may end up picking your brain for that mod when I get around to it.

I had noticed pin 11 being slightly off of 60Hz (something more like 62Hz if I recall,) but didn't think twice about it (oscope tolerance I assumed.) Seems that it's staying on more along the lines of 8 or 9 hours, not 10. Just as well, though I would like to know why. I do know I don't want to completely remove the timer, see below.

I just have to keep in mind: mods may have unintended consequences! e.g. removing the piezo from the UPS at work which we use to isolate the scopes. That constant BUZZ BUZZ BUZZ whenever you unplug the UPS gets so irritating. Remove the piezo and all is well! Only problem is, it's so peaceful that I always forget to plug the damned thing back in to recharge when I'm done.

Cue the hour or two wait next time I need to isolate my scope.

edit: argh, I'm dim. I had my fiancée time it today. Six hours on the dot, and it turns off. I reran my math, and I have set the rc for 10Hz not 6. That's what I get for doing simple calculations in my head.

 

[Resqueline]

It was a Casio FX-180P. I was also annoyed that everytime a fellow student had borrowed it they dutifully turned off the power switch. It turned out to have a single resistor determining the clock frequency. I tested it out with a potentiometer and found the fastest it would run w/o "hanging" and the slowest acceptable for pushing the keys. Then I rewired the power switch into a turbo/snail mode switch instead. People got quite confused when turning off the power had no apparent effect.

I know about the UPS issue. I've even observed people turn off that noisy thing..
Then someone else would complain later that the server wasn't working.. Oooppss..
Try replacing the buzzer with a high-intensity LED instead, or remove the tone circuit so the piezo element just clicks instead.
Please be careful when isolating the 'scope instead of the device you're scoping on.. (It's a mortal sin you know..)

Oh yes, head math hurts, I know that too..

 

[Mitchekj]

I like the LED idea, that may be in order.

Isolating the supplies w/ an UPS tends to be tiring, as the larger (250-500W) ones like to drain the batteries post-haste when under full load. We really need to get an isolation xformer for just that kind of thing. Problem is talking the finance folks into letting us buy anything, lol. Small company and all.

 

[Resqueline]

Ask them about the cost of their life insurance for you, as they're forcing you to risk your life to do the job..
That might get their financial tuned minds into the right tracks..

 

[bluegrassbandit]

It shows a bit. I've noticed you give sound advice in most threads here, from what I've read so far. Nice to see people sharing their expertise.

I got the timer changed. (Good old trial and error.) It was the RC network on pin 11. Aparently the chip is counting the transitions, and 60 Hz comes out to 1 hour. I'm assuming C7 was a 0.1uF, that would give the time const. for 60 Hz. It's a small 0805 smt cap, so it doesn't say.

Anyhow, I changed the RC to about 6 Hz so now I should have a good 10 hours or so. It's been running for a few hours sofar, so it looks to be a good "fix." all other functions are working as normal.

Thanks again for the help!

Hello Everyone,

I have been scouring the web looking for a method to delay my timer from one hour to 10-12 hours or so.

From reading this post, I believe I have finally found that it is possible.

I am just wondering what exactly was changed on the pin 11 RC Network to switch it to 6Hz instead of 60 Hz?

Sounds like a great fix for what I need.

Thanks Again,
Al

 

[Mitchekj]

Oh wow, making me dig through my memory, are ya?

Let's see, I know I did about 10Hz, (by accident,) which gave me 6 hours of on time. So to get 10 hours, we'd need to set the RC time constant to 6Hz.

I would change:
R11 -> 360k Ohms, 1/8W
C6 -> 0.47uF, 10V (16V+ may be easier to find, which is OK to use)

That will get you right around the 10 hour mark, using common value components.

If memory serves:
For R11, I used a carbon film 1/4W axial leaded resistor (if I recall, it was an SMT resistor originally, had I been patient I would've ordered the right one, 0805 package I think.) I just folded the leads over and soldered them to the pads, then bent the body down to fit nicely into the case. (There's plenty of room in the controller case, though.)

For C6, I used a small 0805 ceramic SMT cap. You MAY be able to get by with a radial aluminum electrolytic, should be pretty small, but you'd have to bend the leads, etc.

Now, keep in mind: this is altering the design of a device which may create a fire hazard if something goes wrong. I believe I took every precaution, even ensuring that the SCR was still being triggered as it had been, thus pulsing the heating coils as intended, instead of holding them constantly on. (Which I would advise you do, as well, you never know what can happen when you tinker with something, though you would need a scope.) BUT! It was designed with the one hour timer for a reason. I personally feel that the one hour was far too conservative, but I can't take any responsibility for anything that happens as a result of screwing with their design if anyone does this! I accepted the risks when I was trying to figure this thing out, even let it run at work for a few days in a nice non-flammable area, switching it on each morning.

That said, so far it's worked great!

Edit: Oh yeah, is this for the same type of heating pad? I had it in my mind you were talking about the same one. This one was a Dunlap, not sure of the exact model number. I would NOT recommend you change those exact components if it's a different heating pad. For that, we'd have to go through the same steps of figuring out what your particular pad is doing.

 

[bluegrassbandit]

thanks for the quick reply.

i'll get to work on it as soon as i order the parts.

it is the exact same pad, and i agree, one hour is a bit too conservative. i think 10 hours should be fine, considering i never put the settings above 2 for heat.

i do understand, agree completely the risks involved with modifying something of this nature. this particular heating pad will always be used in a supervised setting, so it shouldn't be a problem.

thanks again for the speedy reply!

 

[TheAndyH]

Timer elimination

Hey all. I see the great detective work you did in finding the R/C timer in the circuit. What I would like to do is eliminate the timer for an "always ON once turned ON" state.

I need it for incubation of baby chicks (cant use a light at night) and for seed germination.

Please let me know if anyone is willing to help out... hope the thread isn't to old.

Thanks!
Andy

 

[Mitchekj]

First: I wouldn't advise doing that. I don't know the heating coils' ins and outs. They may combust after X hours of constant operation. This is NOT safe... that being said, anything I say is to be used for educational purposes only, not for practical application. And always make sure you unplug the unit before tinkering.

Since I don't know exactly what's going on in that IC, other than that it's counting the transitions on pin 11, I'd try removing C7 (carefully, make sure to save it just in case.) See what happens. If it freaks out, or shuts down, try installing a 100k-or-so ohm resistor as C7.

If that freaks it out, keep the 100k-or-so ohm resistor as C7, and remove R11 (also keep it, just in case.)

If that freaks out, well... that will require some thought. Maybe directly grounding pin 11, etc. Let me know how it works out!

 

[joetronics]

I have been looking for this for more than a year now, and you guys just helped me out with all your work.
Only problem I am having is your calculations regarding the time constants: You are saying that 160k and .1ufd = 60 hz for a 1 hour shutdown time. I have tried to make those calculations work, and just cannot - so maybe I am dense - but, something is not right somewhere..... when using the RC time constant formula for those values I come out with .016 seconds - so how do you guys get 60 minutes out of .016 seconds???

 

[Mitchekj]

Aaah,

Freq = 1 / T

1 / 0.016s = 62.5Hz
or 62.5 cycles (on-off) per second.

My guess is that the chip is counting the hi-low transistions, and when it gets to a number (225,000?) it says to turn off.

Slow the frequency down, and you increase the time.

 

[joetronics]

Oh yeah, forgot about that one....

Anyway, I have been thru all the chip sellers in China, and I cannot find any who have the datasheet for this IC. A few have been telling me that Kendin was the original manufacturer of this IC, but they were bought out about 6-7 years ago and the new company does not have the datasheet for it.
Someone has to have it, because these IC's keep showing up in heating pad controllers, and I know that engineers do not throw darts ar a wall to come up with operating specs.
And, of course trying to get a datasheet out of Companies that make this product is like pulling teeth.
Take me back to the 50's, when a schematic accompanied every electronic device sold.