Quick question, when I plug in my power supply(Dunlop 18v) the lt1054 immediately start getting hot. I looked for solder bridges and the like but found nothing. My voltages on the reg and lt1054 look good, not quite sure why this would start getting so hot. Thanks, mike
I would pull all the ICs other than the LT1054 and vreg and check the temp again. Then add them back one at a time and see if it heats up. I'd start with the cheapest chips first :)
Should the LT1054 be getting 18V?
Quote from: daleykd on March 24, 2016, 06:59:42 PM
Should the LT1054 be getting 18V?
That was my first answer, then I looked at the schem and the input is feeding a 15VDC linear regulator. 15V is the listed max operating voltage for the LT1054.
Quote from: EBRAddict on March 24, 2016, 07:09:38 PM
Quote from: daleykd on March 24, 2016, 06:59:42 PM
Should the LT1054 be getting 18V?
That was my first answer, then I looked at the schem and the input is feeding a 15VDC linear regulator. 15V is the listed max operating voltage for the LT1054.
HAHAHAHA. I say "Freekout" and thought the OP meant he was freaking out because it was hot. :)
Quote from: EBRAddict on March 24, 2016, 07:09:38 PM15V is the listed max operating voltage for the LT1054.
There's your problem.
When I get back home I'll trying pulling all the other ics and test it again. When I checked the voltages the lt1054 was seeing around 14.77 I believe so the regulator was doing its job. I checked that first to make sure the lt1054 wasn't seeing 18 volts, but maybe I'll check again just to make sure. Won't be able check till tomorrow though. Thanks for the input, mike.
Ok narrowed it down to the ad633. When I put this in the lt1054 gets hot. I assume I got a bad one from China. Can anyone tell me why this chip would make the LT1054 get hot. Thanks, Mike
Check your pin voltages on that socket, post them here, maybe someone can take a look. I'm not familiar with it. Datasheet says it's a "low cost" analog multiplier at $10 a chip :)
I'm having the same issue with the lt1054 in my rust bucket project. I took out all the chips and it was still really hot. When I measure the voltages on the l7815 cv it reads 18v on the in and it jumps around on the out pin. I tried several l7815cv (socketed) and they all do this. I've even tried several lt1054 chips and they all get hot. Any suggestions. Thanks
Quote from: John S on March 28, 2016, 03:25:58 AM
I'm having the same issue with the lt1054 in my rust bucket project. I took out all the chips and it was still really hot. When I measure the voltages on the l7815 cv it reads 18v on the in and it jumps around on the out pin. I tried several l7815cv (socketed) and they all do this. I've even tried several lt1054 chips and they all get hot. Any suggestions. Thanks
If the voltage is jumping around on your output pin, something is wrong. The regulator should be putting out a steady 15V. That's the only job it does.
I'm using a L7815CV instead of a LM7815. Would that matter? Any suggestions as to what could be wrong?
Did you check the pin out?
Sent from my iPhone using Tapatalk
Are you guys using low-ESR filtering caps on your LM7815? If so refer to the datasheet because I think they need electrolytic caps with "regular" ESR.
I seem to remember people having trouble using the L7815 in some builds in the past, but I think the threads got nuked in the forum crash. I'd stick to LM7815. I probably have an extra I can send you if it helps.
If you could send me one, that would be great. Let me know how much and I'll PayPal ASAP. Thanks
I had problems with the using the CV version, incorrect voltages result! Replaced with correct version and all was well.
Well I'm pretty sure my ad633 is bad. Before I put the chip in I get -14.36 to pin 5 and 14.84 to pin 8. As soon as I put in the ad633 I get -3.89 to pin 5 and -1.21 to pin 1. I bought mine on ebay from china. Does anyone on here have a seller they bought from that I can try to purchase from. I know that's not always a guarantee but its better than stabbing in the dark. thanks, Mike
Quote from: Droogie on March 28, 2016, 05:40:58 PM
I had problems with the using the CV version, incorrect voltages result! Replaced with correct version and all was well.
Did it damage any other components?
CV suffix is fine for this build. It's the one I used (actually the only kind I have).
Quote from: gaffled on March 28, 2016, 06:55:00 PM
Well I'm pretty sure my ad633 is bad. Before I put the chip in I get -14.36 to pin 5 and 14.84 to pin 8. As soon as I put in the ad633 I get -3.89 to pin 5 and -1.21 to pin 1. I bought mine on ebay from china. Does anyone on here have a seller they bought from that I can try to purchase from. I know that's not always a guarantee but its better than stabbing in the dark. thanks, Mike
Mouser sells them.
I tried a different L7815cv- the voltage was reading correctly at the in and out. The LT1054 was still getting hot and the voltages were way off. Should I switch out the caps that come off the voltage reg to the 1054?
The LT1054 is getting hot.
Your voltages are dropping.
THESE ARE RELATED PHENOMENA (https://en.wikipedia.org/wiki/Ohm's_law).
I don't know exactly where your problem is, but when you put the ad633 chip in, it's drawing a ton of current. Check (MEASURE) all the components that attach to it. The voltages will look normal if there's no current draw (nothing to drop the voltages) so having the right voltages with the chip out isn't useful except as an indicator that some connections are there.
At this point you can pretty much eliminate the power section, filter caps, and other op amps as the source of the problem. Since there are only a few connections for that chip, though, and I have no idea how that chip works (read: I don't feel like deciphering a datasheet right now), I can't tell you what components to check, but I would at least check for rogue continuity between some pins, that you have 0V on pins 2, 3, 4, and 6, and something close to 0V on pins 1 and 7.
I just tried an LM7815 and the LT1054 is still getting hot. I'm trying to build the rust bucket here and now I'm at my wits end. I'm using the Dunlop 18v wall wart as suggested in the build documents. Please help!! Thanks
Okay, this is how you diagnose and fix this problem (relating to the Rustbucket although the process is the same for the Freekout). It sounds labor intensive but it will actually only take about ten minutes. As Jon stated - you need to find the current sink and if there is more than one. Skip any steps you have already done.
- Remove all IC's from their sockets.
- Power the board, and verify +15v from the regulator.
- Un-power, replace the LT1054 in its socket and re-power (do the un-power and re-power for each subsequent step). Verify you have -15v from pin5 of the LT1054 and +15v from pin8. If you do not, then you need to re-flow the solder joints on all the ICs and/or replace with a different LT1054. If you do, then you know each IC is getting proper power (because any one IC that had a solder joint or PCB error on the power rails would show up at this step).
- Here's the meticulous part - you need to start populating one IC at a time, check for +/-15 volts on the proper pins and check if the LT1054 is getting hot (both are likely to happen at the same time). Do it in this order: 4558, LM741, LM1458, CD4047, CD4013, CD4066, LM311, CA3080 and finally the MN3007. If at any point in this process you are not getting the proper +/- voltage or the LT1054 is getting hot, you know you probably have an IC that needs to be replaced. You should do it for all ICs - don't stop if you find one bad one (because you need to know if there is more than one problem area). It sounds like a lot of work but trust me, I've done this sort of procedure plenty and it doesn't take long.
Report your findings for further assistance if needed :)
I have 14.7v coming from the volt reg. I was getting 14.7v from pin 8 and anywhere from -3 to -8v on pin 5 (tried 3 chips). I reflowed the solder on all ic pads. I kept getting the same voltage readings. Then I switched out the ic socket and still kept getting the same readings. Are all my chips fried?
No, I don't think your chips are fried. Are you getting the low readings on pin5 with all the other IC's installed or have you tried removing them and just taken a reading from the LT1054?
I got the low readings after removing the other ic's.
What's my next move?
I had this same problem too (I think the thread was lost in the crash) and was finally able to figure it out, and it was a bad AD633. After looking over the spec sheet there is two pin outs for this chip; a JN/AN & a JR/AR. I think the cheap ebay chips are the JR/AR version (but they are labeled as the AN version) with the power rails on the 3 & 6; the one needed for the Freekout is the JN/AN version with the power coming in on the 5 & 8 pins. I finally got one from Mouser and it it works fine. So if your going to build one just get the AD633ANZ from Mouser and save yourself the headache.
Thank you. I was going crazy probing this thing. Was waiting till I had a mouser order to get a chip from there, so I'll just put this aside until then.
I had the same problem too. i'm ordering an AD633AN from mouser now. thanks for clarifying!
I made a small daughterboard to make an an AD633AR match an AD633AN's pinout, but it didn't work. either I fried the chip when I tried it before the daughterboard or the one that I got from eBay's pins didn't even match an AR's pinout. i'l have a real AN from Mouser on Monday so hopefully all of the other chips have survived.