No.036 - Digimess PM6003 Power Supply Repair

[@captiveimage]

First time I've seen a vid like this, where the design of the device circuitry is not only analysed, but outright rejected and rebuilt. Awsome! I hope Digimess were watching. They should be giving you a nod for the redesign 😉

[@SpeccyMan]

Digimess seems quite an appropriate name in this case. Great to see you back Ian.

[@cfytcf]

Seriously, who would call their power supply something like that?

[@bobford100]

Hi Ian, Recently came across your channel when searching for Digimess vids, and immediately subscribed. Great channel, keep up the good work. I've had a couple of the Digimess PSU's for a year or so now. As with yourself, bought direct from Digimess via Ebay as non workers. One is a PM6003, and the other a PM10001, and I put them on the shelf waiting for motivation to investigate and repair. Both have identical internals, but with subtle component changes to suit each supply range. After seeing your repair video, I then had motivation to investigate. So, both PSU's are dead, no displays, no output. I immediately checked the TVS on both units, and the PM10001 was reading around 10 ohms, the diode in the PM6003 read fine! So it would appear that the TVS is a weak point on these items. I will now replace the TVS, and see if that fixes the PSU, or if I have the same problem with the Opto supply arrangement too. As for the unit where the TVS checked out fine, I'll leave a comment here on any extra components that have failed. I must confess I'm not 100% sure I understand the opto circuit bias problem fully, but should I have the same problem then I'll have to study the circuit some more. Cheers Bob

[@artursmihelsons415]

Coolest thing to see is that not only i, in some cases, putting paper under the board to prevent shot circuit when testing faulty device. :) Thank's for great video and st-by supply update schematic.. I think, that this same problem may be in other brands, similar bench power supplies.. And that's mean that this video may be useful even later for other's, when repairing other brand supplies.. ;)

[@Ogma3bandcamp]

Good to see you back and sharp as ever.

[@vojtechdvorak2219]

Thank you very much for the schematic and explanation.

[@IanScottJohnston]

Thanks for the nice comment.

[@alpcns]

Always a pleasure to watch! Thanks for sharing.

[@makeitwork7522]

Quality video - well explained and fixed. Thanks for sharing.

[@saarike]

Nice video with good information about those power supplies! Thank you.

[@jameshauser1507]

Very nice fix Ian.

[@stevenbliss989]

Tracking "hybrid" PSU, very simple and a bit novel ....now to watch the rest of your video :)

[@philiprogers5772]

That was a master class.

[@jeffescortlx]

I love videos like this. Keep them up!

[@GadgetUK164]

Excellent! Shocking that their design there for that "standby" circuit was so deficient. Makes you wonder if that's quite a common fault and they know of the deficiency in that rev board. Also lucky that over voltage didn't kill anything else - I imagine any digital components would take offsense to 9v for a period of time!

[@IanScottJohnston]

The ICL7107's are 6v max.......but lived!

[@tbbw]

I love videos like these... More you two! A dedicated viewer demands it! ( Or i'll haunt you while bored... ever seen the library scene in ghostbusters? imagine me doing that to your component trays )

[@MrDomyboy]

Nooooooo, not the component trays. I would cry hysterically if someone did that to me haha.

[@flatfrog69rr71]

Hi Ian You must have the most tidyest workshop in the world!!

[@IanScottJohnston]

It's only because it is a brand new build.......gimme 6 months :-)

[@SeanBZA]

The diode would probably be better replaced by the proper snubber, as in the datasheet, as this will be more reliable, but at the expense of slightly higher power use in the auxiliary power supply. Better would be to use a TL431 in the classic SMPS circuit to get the regulation correct, though you also have to remember the TOP series must run with a secondary load, they have very poor load regulation due to the feedback method they use ( hard to get the duty cycle very low if the feedback is also the run power supply) so need a defined minimum load. no load they can drive the secondary to very high levels, also a fault when the secondary side output capacitor fails, or even that small capacitor in the feedback circuit. Best to replace that with a better unit right off the bat, and add a 6V8 5W Transorb diode across the 5v rail as well.

[@FixDaily]

Case to say that you have a DIEode :) First video that i saw on your channel, sounds like a channel of my interests. Great work keep it up.

[@DanielMelendrezPhD]

Some months ago I got one of these from eBay as "for parts or not working" for £5. I fully recapped it, changed some diodes, transistors, and repaired a bunch of traces that had burnt after a massive short. Success! It worked like a charm and I was able to pull about 9A without a problem at a rock-solid stable voltage. HOWEVER, one day while trying to calibrate the max current (shorting the output terminals) the supply blew up with a massive bang in front of my eyes! Smoke, flames, and bad spirits came out from the PSU. I was so pissed that I immediately disassembled the little beast for salvaging some parts and promised myself to never ever trust a crappy chinese PSU. Sad story....

[@mgphotostudio]

Omg hahaha epic!!!!😂

[@bobford100]

Hi again. Just an update on fixing the HV10001. Replaced the TVS reads 10 ohms), replaced the TOP224 (O/C all leads) and replaced C101 (very low capacitance) fixed this variant of these PSU. Did not have to alter the opto supply arrangement, however this unit does have a 0.1uF cap from C105 -ve to ground. Very odd PSUs these Digimess. Cheers Bob

[@jetraid]

Excelent job. I notice something strange when you turn on the PSU the first time, if you look the amps indicator is going up without any charge connected, maybe that's the indication of a over voltage who cause the wrong indication until the switching psu protect itself.

[@xgentecchannel2309]

Would have been interested to see what that D104 zender looked like on the curve tracer but all in all nice repair good work.

[@TheDefpom]

Obviously this was a design flaw, good to see you figured it out and corrected their error.

[@IanScottJohnston]

Yes, looks like it.......just wasn't expecting to do this though in what should have been a straight repair job......especially on a 152 ukp psu! I'll point Digimess to this vid and see what they say.

[@ryebis]

Digimess, named perfectly.

[@FireballXL55]

I noticed the voltage output before you modified the standby PSU went above 62V so maybe an over voltage trip kicked in.

[@jp-um2fr]

Just found you - perfect. I'm on my 7th P.S. The latest has 2 x 37V DC 10A transformers, a 20A (17A ?) buck, 4 x LT1083CP in parallel regulator, and 2 x 6 TIP35C regulators. In other words a hybrid 4 channel around 5A each at 36V or 2 at 10A. 9 fans auto regulated, temp. gauges and thermal cutout on each regulator + V & A gauges. The 2 x TIP35's regs. are happy to run in parallel at 32V 20A. The front panel is a nightmare (for me), around 70 odd connections. Why build it? All part of a learning curve and good fun. Did valve amps 50 years ago (EL34's - lovely beasts). Iv'e had 410V across my nose (valve days) and lately 245V right across the chest, not good at 72 years old - OOPS!. Had a look around your channel and you now have an ardent fan.

[@IanScottJohnston]

Greetings!......I've had a few guitar amps with el34, el84 and 6L6's, but never done any repair or design on them. Ian.

[@robertcalkjr.8325]

Good job. Thanks Ian.

[@stevenbliss989]

I think there is a mistake on the schematic at T202 rectification stage???

[@chriswouse7713]

Nice vid. Got a 3006 in my repair bin that doesn’t work under load, so will be trying to find that schematic and poke around (carefully)

[@tonybell1597]

Top work Ian... thanks.....

[@svsv9]

Hi, I have same PSU but transformer version, I replaced the potentiometer, and the psu still float the regulated set voltage I need it to work on, example: it will jump from 5v to 3.4 back to 5.5 and so crazy on, sometimes it may get stable. I hear how relays switching while start get unstable. Some help, what could I check?

[@shoaibnadeem]

Hi! I have a 0-60v 5A power supply. Its output somehow got short for like 30 seconds and now it is stuck on 20-21 v and not going up or down. What to check / replace ?

[@tukler2]

Great vid! Thanks!

[@sureshlingabathina]

excellent, subscribed.

[@PapasDino]

Wonder what the zip-ties are for around the two large brown electrolytic caps? Doesn't appear they're attached to anything. Also, with all that real estate available on the bottom PCB I guess it's cheaper to manually add solder to the trace after the fact to increase its current capacity instead of simply designing it with a larger trace to begin with? Would have been interesting to be a fly on the wall during the design review for this puppy!

[@IanScottJohnston]

Dino Papas - It's one cable tie that loops around both capacitors as a way to hold them together.

[@WaltonPete]

IanScottJohnston So the solder joints can fracture in transit on both of them simultaneously, rather than just one!

[@twobob]

It amuses me how non-catastrophic a few volts sounds, in stark contrast to the usual impact. Of course it would be like running an old b+ at 200% and maybe 500V sounds more "deadly enough"; as a guy who sucked up an EHT voltage strap hit "once upon an idiotic time" I can report a high excess of electrons is as shocking as it's name suggests: "Shocking".

[@101blog]

Well another good Video. Thanks for making an SMPS repair fun and interesting ;-). Where did you get the circuit diagram from Ian ? Did they Volunteer it or did you have to ask?

[@IanScottJohnston]

Both!........I asked for it before the PSU arrived, but one was sent with it anyways.

[@charlesdorval394]

Congrats mate! :)

[@ameristanbouli5063]

Hi man My power supply current reading keep going up and down what could it be the problem ?

[@motoqueirosemmoto]

Did you find a Terminator at 15:09?!?

[@CliveChamberlain946]

Kind of think they should have put a snubber cap near that TVS you replaced..?

[@IanScottJohnston]

Yes, I saw that too.

[@TechBench]

Nice vid, as always. What does the power-on curve look like (without and with load). Any spikes or other surprises?

[@IanScottJohnston]

Hmmm, might do a follow-up on this, or at least post the info on my website. Thanks.

[@SmithKerona]

Hi Ian, great repair video! One request please, could you post the schematic in pdf format in your website? I have the same supply and would love to have the schematic. Thank you!

[@IanScottJohnston]

Done.

[@uwepolifka4583]

I think the modification gives more power to the optocoupler´s LED. When this LED needs more power than before to give feedback to TOP226 the coupler ist out of specification what means it is defektive. I would have tried a new optocoupler first.

[@IanScottJohnston]

I do have compatible opto's in stock but it seemed to check out on the diode test so left it.

[@ntag411]

Line input power should be monitored and turned off when the power is not appropriate. I've worked in power supply manufacturing and data was collected for line/load regulation, ripple, isolation and efficiency.

[@LJ45Chimera]

Great Vid, Thanks :)

[@ahmedibrahim2512]

can u send me the schematic please

[@GADELHAS82]

And what about the Farnell LT30-2 Power Supply project?

[@IanScottJohnston]

GADELHAS82 Too much revolt that I was losing a nice analog psu, plus it was taking too much time........so it's back on my work bench in its original guise! Must admit the analogue movements on it are quality items.

[@uK8cvPAq]

That's a bit shit then, you think there's something else off or could this happen to any of these units in service?

[@IanScottJohnston]

uK8cvPAq Don't know......funny thing is, if you look at the schematic later in the video I have drawn on a cap on the problem 5v circuit, they had hacked this in place so I am thinking this was their attempt to get it working.......hmmmm!

[@uK8cvPAq]

Maybe they could get capacitors a cent cheaper than resistors that day!

[@nyannyan8243]

"Digimess PM6003" - Don't mess with Digimess!

[@simontay4851]

Looks like the zener in the small 5V PSU had gone high. Not a very good design. Obvious that they designed/built it as cheap as possible and didn't follow the data sheet. Should've used a proper TL431 regulator. Its a switch mode PSU but would I right in saying that the main output is regulated linearly with that MOSFET and the opamps or do the front panel controls change the feedback resistance/PWM duty cycle of the TL494 SMPS control chip.

[@IanScottJohnston]

Simon Tay I did change out that zener during the repair but it made no difference........was one of the reasons I opted for the data sheet circuit.

[@IanScottJohnston]

I never measured any of this to see it working, but I think the TL494 ia acting as a pre-regulator. It's picking up the voltage "across" Q401 the main output FET (+ before, - after) and adjusting the duty cycle T202 in order to keep this to a minimal voltage......hence the small'ish heatsinks required considering it's a 60V/3A PSU. The voltage/current control is all done with the op-amps via the main Q401 FET........but like I said, a guess.

[@mrjohhhnnnyyy5797]

The SMPS section is just WEIRD. Halfbridge on the primary, and the rectifier from a single switch forward on the secondary, why? Full wave to the transformer, half-wave rectifier on the secondary. It's a digiMESS indeed :D

[@IanScottJohnston]

We have a saying in Scotland...."oh ma heid!"

[@rogeronslow1498]

Mr Johnnny. I thought the same when I first looked at the schematic. If you look closely at the secondary rectifier circuit you'll see two output inductors across the secondary of the hf transformer. They are wired in a classic current doubler configuration. It doesn't help matters that they placed the output diodes in such a way that it initially looks like a forward topology but is actually half bridge with a current doubler. Why they chose to do post linear regulation using a series mosfet is beyond me when they could easily have dispensed with the linear regulator. This is a classic example of a very poor Chinese design.

[@IanScottJohnston]

I thought about that too and just wondered if the same finite control would be achieved.

[@rogeronslow1498]

IanScottJohnston The only advantage of the post linear regulator is a faster transient response. But really, the switch mode circuit can respond in under 1ms easily anyway. The hf ripple will not be removed by the post regulator as some people seem to think.

[@mrjohhhnnnyyy5797]

Roger Onslow Oh, now I get it. The schematic is a bit deceiving. As for the linear regulator - it does knock out ripple, if you have one with nice PSRR. It won't save you from the common mode noise though.

[@aaa000777]

Must be short hard for Digital Mess !!! LOL

[@himanologanor4093]

while the design is very poor (no surprise here), the real problem is that there are outthere a few thousands already sold and working (as it`s mine) so yours had a problem inside and it`s STILL inside as you only dragged it down somehow to work. If it was a design flaw now it should be thousands of guys returning them with out-caps blown but they aren't.

[@IanScottJohnston]

As it was the manufacturer I bought it from I could have been sold an old/early version, or a version that hadn't been modified to work more reliably/better. Who knows!......Digimess have yet to get back to me.

[@standishgeezer]

They're still selling/trying to get rid of these (& 30v 6A versions) with the same blurb.....and probably the same design fault. Bit dishonest really as they are probably fully aware of what the issue is.

[@leppie]

Chicken dinner

[@fapcrafter7395]

digital mess

[@titodejesus6238]

Nice video but TOTALLY OUT OF MY KNOWLEDGE AND SKILLS.

[@drteeth7054]

"No transformers as such"! Buzzzt! Wrong!

[@user-rs8zg8ey2b]

Un-subbed, my comments get deleted or not shown.

[@IanScottJohnston]

Not by me. I'll have a look at the back end and see if KZbin algorithm is doing it. Advice: you can always contact me via my website any time you feel this is happening. A shame to just un-sub.

[@user-rs8zg8ey2b]

@@IanScottJohnston my reply to this just did not show up either.

[@IanScottJohnston]

@Ozzy3333333 this one did. Hmmm, i will look at backend when i get out to workshop soon.

[@IanScottJohnston]

I looked at the backend and there are no posts there by you. YT do delete them automatically after 6 months so. Not much more I can do I am afraid!

[@user-rs8zg8ey2b]

@@IanScottJohnston thanks, maybe if I put a link to my blog at rcgroups it deletes the post,,, Idk.