I mostly buy from RS. They can be too expensive for a lot of stuff but with components it usually works out pretty good.
eg I looked up that C104. I can get 100 of them for 50c which includes next day delivery to my door by security courier. That's pretty cool.
Yet, it will cost around $8.00 for the min of 100 1n9148's when I only need a few. Jaycar doesn't have smd but their 1n4148's are 60c each, so what do ya do ?
Naturally you would buy other stuff from RS at the same time to make it more worth their while but just saying, that's the way they work it. I'm sure they'd have special contracts with the couriers to even things out.
So I think I might just buy up and replace all of the caps in that part of the circuit. If it's still no good then that'll rule them out.
Oh and about those glasses, I have the cheap ones too. 3 of them in fact. 1.5, 2 and 3. I use 1.5 for driving, 2's for getting around the house or shops and the 3's for close up, and you can stack them. 1.5+2 seems like 3.5. .. 3+1.5 seems like 4.5 and so on. When I go to town I always take the 2's with me so I can double up to read a label or something. I'll put all 3 pairs on when doing close smd parts and I've gotta hold my breath for the solder fumes lol
Last edited by loopyloo; 06-08-17 at 12:51 PM.
It doesn't hurt to have a bit of a reserve stock of smd components because more and more repairs are needing them these days.
Jaycar still have packs of 10 smd 1N4148 diodes on the books for $6.95, got some a little while back (catalog # ZR1103)
Element14 offer free shipping if your order totals over $45 (used to be $36). They have warehouses all over the planet so large orders tend to arrive in dribs & drabs unless it's local stock which is typically next day delivery via courier.
loopyloo (08-08-17)
Bought some caps locally at Jaycar but after replacing 2 of them I made a soldering iron tip with 2 prongs and I found it pretty easy to remove, test and replace them. Well not too easy to replace but got away with it. Holding the little bugger while settling it into wet solder was a bit tricky but getting better at it.
I've now checked all of those caps and they're all good.
I was using my little hand held oscilloscope to check stuff but it seems to have gone off the planet. Looks like another fix it job.
What's next on the check list ?
I use some really fine point tweezers to hold smd's while tacking one end to the pad. Then it can't slip while doing the other end, lastly go back and give the first end a touch-up to be sure.
Do all the diodes check out ok IE D101, D102, D103, D105, D106, D121 (all 1N4148s)?
If the RFmin voltage is still pulled low there's a fault in there yet to be found but where.
This thing is fighting hard to stay out of action
EDIT: C114 (18nF) could have copped a good blast of EHV and it's rated at 800V - any chance of checking it with a 500V megger (out of circuit)?
I don't see how it could influence RFmin with the chips's Vcc coming from a separate source but everything is suspect at this stage.
Last edited by Skepticist; 09-08-17 at 11:08 PM.
loopyloo (10-08-17)
Yeah I used my fine point surgical tweezers to hold the smd parts.
All of those diodes test fine in circuit. I guess I can pull them out of circuit to check. Those Jaycar ones are no good by the way, They're 4mm long
A megger is one tool I don't own. I suspected C114 a while back so I removed it and it comes up spot on at 18nF.
I could just replace it to check although the closest I've got is this one but it should do. It appears to be rated 1250v.
Terrible photo , I know
Last edited by loopyloo; 10-08-17 at 12:46 AM.
Definitely try with that 22nF cap in place - shouldn't make any drastic difference as it's only +20% compared to the original.
The original could be ok at low voltage of a cap tester but breaking down with 340V applied - charges via Q101 then discharges via Q102 so the symmetry will be majorly messed up by a damaged cap.
The worry at this point is that you could get to the point where everything has been checked OK or replaced if slightly off spec and the fault is still there.
How is that Vcc voltage now? needs to be well within specs to rule out a supply problem.
loopyloo (10-08-17)
Yep, I'll shove it in-in the morning, too cold to go out resetting the breaker nowOriginally Posted by Skepticist
ha ha
Check this out, reading a 9v battery.
Last edited by loopyloo; 10-08-17 at 01:04 AM.
Looks like an interesting battery
or the ADC in the CRO has lost the plot
I get days like that - everything I touch develops a fault
Last edited by Skepticist; 10-08-17 at 01:18 AM.
loopyloo (10-08-17)
Lol...lost the plot alright, intermittent too....and does some other weird things as wellLooks like an interesting battery
or the ADC in the CRO has lost the plot
I get days like that - everything I touch develops a fault
I wrote to Banggood hoping they'll send parts to fix it or a new one. If not I'll buy a better hand held one day (when I can afford it)
Vcc is at 9.6v. The same voltage is at C509 so no losses between those points.
Changed C114 but no luck.
Last edited by loopyloo; 10-08-17 at 11:12 AM.
That Vcc is still too low, needs to rise above 10.7V to get the chip to start up. The chip's max demand on Vcc is quoted at 25mA (typically much less than that in operation) so something is overloading the supply.
How about that RFmin voltage - still suppressed? There's no definite info on whether the 2V reference is reliant on the chip's Vcc threshold but it's possible.
Need to get the gloves off and go heavy handed, maybe sectionalising the circuit by cutting tracks to work out what's causing the overload to pull that voltage down.
I'd be using an external variable 12V supply to reduce the risks of working with a live circuit and the first thing to disconnect from Vcc is the PLC circuit as there were odd results there before
loopyloo (10-08-17)
Yes RFmin is still suppressed
PLC ?
Think you mean PFC. I can pop F101 out and cut the track just before R117, Short pins 3&4 of IC102, connect 12v to R117 then start measuring.
Even more interesting is that D510 cathode has 16v on it. Maybe I should first lift R502 and see if IC501's circuit is the thief.
How's that ?
Last edited by loopyloo; 10-08-17 at 12:33 PM.
Yes PFC (slip of the digit)
See if Vcc rises high enough without the PFC connected first
For applying external 12V supply also need to bias LINE at 1.24-6V which could be done with a temporary 33k resistor from Vcc to LINE (should set ~4V). Don't forget to remove it before replacing the fuse :O
with no HV should be able to safely poke around with instruments on both sides of the IC to get measurements
It's hard to put all this in text but I'd leave the mains completely disconnected while testing with external supply (not just via F101) so there's nothing on the board that can bite you or your tools.
EDIT: IF there's 16V on D510's cathode then it would appear the major voltage drop is across Q501 C-E (5V or more) so it's not fully switching on but it IS switching from earlier probing. Check that R508 (1k) hasn't gone high and get a voltage reading on the base of Q501 with opto IC503 pins 3 & 4 shorted.
Last edited by Skepticist; 10-08-17 at 01:13 PM.
loopyloo (10-08-17)
Checked Q501 with shorted IC501 : C 16v, B 15v, E 9.6v .... R508 is goodYes PFC (slip of the digit)
See if Vcc rises high enough without the PFC connected first
For applying external 12V supply also need to bias LINE at 1.24-6V which could be done with a temporary 33k resistor from Vcc to LINE (should set ~4V). Don't forget to remove it before replacing the fuse :O
with no HV should be able to safely poke around with instruments on both sides of the IC to get measurements
It's hard to put all this in text but I'd leave the mains completely disconnected while testing with external supply (not just via F101) so there's nothing on the board that can bite you or your tools.
EDIT: IF there's 16V on D510's cathode then it would appear the major voltage drop is across Q501 C-E (5V or more) so it's not fully switching on but it IS switching from earlier probing. Check that R508 (1k) hasn't gone high and get a voltage reading on the base of Q501 with opto IC503 pins 3 & 4 shorted.
So I lifted R621, Vcc still 9.6v so IC601 circuit isn't the culprit.
Then I lifted R117, Vcc became 15v so we now know the fault is definitely in IC101's circuit somewhere. Damn !
Last edited by loopyloo; 10-08-17 at 04:18 PM.
At least some parts of the circuit are proven not guilty which helps, a little anyway
The schematic indicates C108 (100nF 50V) between Vcc and GND and also a ZD112 which is marked N.C. (not connected?) - it is left out I take it?
Check every mm of the Vcc PCB track between R117 and IC101 for problems or other connected components (in case the schematic is not 100% accurate)
And try supplying Vcc via R117 with the Vcc pin lifted just to determine if the high load is the IC or external to it.
Damn is the word - this thing is not giving up easily
Last edited by Skepticist; 10-08-17 at 04:52 PM.
loopyloo (10-08-17)
Yeah ZD1122 doesn't exist, and the track from R117 and pin 12 is only about 10mm long, it goes under the chip and the readings are the same both ends.
I wonder if I might find the fault by lifting pins one at a time, starting at pin 1 and checking that Vcc each time ? with normal power.
I'm now wondering if I've already done this in reference to pins 4 and 5. I know I checked across those resistors but can't remember if I checked Vcc at the time. I'm confused. I blame it on the mongrel bug I've had for about 3 weeks.
Blimey, I think I spend more time editing my posts than writing them in the first place. What's the count now ? arr 10,600 views....going well. Where's the moderator ? ... I want an Austech t shirt, small size please.
Last edited by loopyloo; 10-08-17 at 07:20 PM.
C108 is the only component other than the IC on short piece of track so try with C108 disconnected just to be thorough. (it's only a bypass for noise in the chip)
Lifting pin 12 would prove whether the fault's internal or external, after that try just about everything. All results so far point to internal and it's really beyond weird now IMHO
Removing C108 didn't help. Hey, have I missed a Vcc somewhere ? like at the MICOM or something ?
Thought you might like to see what we're dealing with.
Need to make a ring light for this camera I think, seen any good circuits ?
Last edited by loopyloo; 10-08-17 at 06:06 PM.
That pic is very clear but starting to look like a crime scene now
(Something evil therein lurks)
The MICOM_VDD is actually the 3.5V standby supply but enabled by the secondary 12V supply which is named DRV_VDD
The Vcc produced by IC501 is only used for chips on the primary side of the power supply board (IC501, IC101, IC601)
Ah well that's ok then IC501 is not to blame.That pic is very clear but starting to look like a crime scene now
(Something evil therein lurks)
The MICOM_VDD is actually the 3.5V standby supply but enabled by the secondary 12V supply which is named DRV_VDD
The Vcc produced by IC501 is only used for chips on the primary side of the power supply board (IC501, IC101, IC601)
I removed that last cap (C108) for testing then moved the board into the office to get that photo. Upon returning the board to the shed I remembered that cap but now it's gone
I left it sitting on the board. There's a trail of about 4 metres of concrete and 8 metres of carpet to search now......crikey !
Guess I'll just put one of the ones I bought in it
Posting Permissions
Reply With Quote
Bookmarks