AP series power supply (AP60-50) (Wayne Kerr, Farnell, HiTek APS)

[Vannystick]

So for a few months now I've been battling with the children to go to bed, no wait, battling with a Wayne Kerr branded AP6050 power supply. Its pretty common in the UK automotive industry, every company I've worked at has several, many of them failed. I've been given this one because, its failed. Over the years this PSU has been rebranded Wayne Kerr, then Farnell, HiTek and not APS (Advance Product Services), the latter of which has quoted over £500 for a pre survey quote.

I only want to get it running to power up random automotive components as I try to make them work in totally in appropriate circumstances, Webasto water heater to keep the paddling pool warm for example.

It shows UNREG as soon as the output is enabled, and no 'power' comes out of it. Sense wiring has been replaced, no obvious internal damage (water ingress, dirt etc) except C47 was damaged while investigating. All ICs are socketed, and have been pulled, visually inspected and pushed back in. There is no obvious damage to the board. This seems to be the most common problem with these supplies, and the issue tends to occur with age rather than specific usage type.

Hopefully the service manual is attached. I've been chasing around expecting to find a single at fault component. All of the ICs related to the UNREG detection seem to be working okay, supply and ground ok, relevant outputs acceptable for a good condition power supply. The absolute current limit (P2) has no effect and seems to have resistance in the range expected (a dead wiper would have been nice). I did find a few RIFA caps, they aren't swollen and cranked, no escaped smoke smell, I'm inclined to leave them alone for now.

Unless someone has seen this issue before and can direct me to the right point, I guess I'm asking for some general advice on where I can continue to stick my meter/'scope (Farnell DTV 12-14) or if I need to consider something more extreme like a trip through the dish washer (the main board, not me) or a re-flow session.

Any help, pointers, or even humours comments gratefully appreciated.

[Andy Watson]

I have its little baby brother - AP7030. This is a very similar device to the AP6050, I know because I used the 6050 manual to repair it. I can't remeber what it displayed when I first acquired it, but it was definitely a sick puppy. Both sense lines were broken. I had to bleep-out the signal from the terminal posts forward to locate the breaks - both breaks wre hidden by the shroud of the terminal block. I guess that somebody had tried the draw a load from the sense terminals rather than the main output terminals and blown the tracks on the PCB. Whatever caused it aslo took out a futher three chips attached to the sense circuitry. There were also a couple of logic chips that seemed to have failed for no apparent reason.

First identify were the high-voltage mains power is - and stay well clear unless you really need to go there!
You didn't mention an GPIB/IEEE interface but if you have this option fitted you can temporarily remove it to make debugging easier. Remove the interface board and re-plug the display connector into the obvious place on the main control board.

I would start by bleeping-out the sense lines. Check the internal power supplies. Then start at the UNREG signal - trace through the logic and find out what is causing it - don't assume that the logic is working. When testing the logic signals be aware that it operates at two different voltage levels with respect to different references.

I wouldn't recommend the scatter-gun approach of randomly replacing chips - there are too many - unless you're feeoing lucky! Be patient, the fault(s) can be hunted-down. If you must "randomly replace", go for the op-amps associated with the sense circuit first. Then try the cmos switchs (4066 ?) - these seem to be weak points.

[Vannystick]

Thanks for your thoughts Andy. Written down, your approach makes far more sense than mine! I'd been trying to work back from the 'UNREG' LED illumination, rather than considering the route cause of why it was illuminated. Then the supply arrived, the wiring on the back was not the greatest, so your fault could well be the same as mine, some investigating to be done!

I've grabbed some pictures, because life is more fun in colour. You can see on the board a suspicious track on the sense connectors, I wonder if it is that simple?

I guess I had assumed (hoped?) it would be an IC issue, but reading the descriptions for UNREG again and again, there aren't many ICs involved. I'm going to restart with your suggestions and see where I get to. It may take some time to progress, but I'll be back!

[coromonadalix]

Have you checked this explanation from the service manual

6.3.4 Low A.C. Mains Voltage Protection
If the a.c. mains input falls below approximately 70% of normal, the power supply will shut down until the normal mains supply is restored (operation at low mains could cause overheating and/or unreliable operation).
This condition is indicated by the ‘UNREG’ LED being illuminated. The green ‘enabled’ LED will remain lit, the output will be shut down and the meters will read approximately zero.


Is you main ac properly grounded an properly wired ?? up to you electrical panel   i had an 1kw amplifier who gave me some errors with an faulty earth ground

If you want to scope your instrument, be sure to have an isolation transformer,  have you tried the re-calibration procedure at section 8

Seems to be some variant of an smps phase controlled psu, seen in some sorensen  etc ... models

Is your Auxiliaries supplies good +12v +5v -5v -15volts, you have an 25khz oscillator Ic2a is the frequency okay ??

I admit  this one is a complex puppy.

[stj]

this is actually made by wayn kerr,
do you have the GPIB manual in pdf form? i'm looking for it.

if you have the GPIB pcb fitted then look for track damage from battery leakage on it - that can cause all kinds of prioblems.

[Andy Watson]

this is actually made by wayn kerr,
do you have the GPIB manual in pdf form? i'm looking for it.

So am I

I have spent some considerable time surfing for information on the GPIB interface. Wayne Kerr responded with the basic instruction manual for my AP7030, but this did not include any details of the GPIB interface.

What information do you require? I have reverse-engineered the GPIB interface to the extent of obtaining a schematic and a working set of commands. I extracted the commands by looking for ASCII strings in a dump of the EPROM code. I have used these commands to configure the power-supply and read-back data. There is also a string that returns the version/type of my power supply, and along side this string are other, unreported strings for the AP series - so I assume that this firmware is common to the whole AP range.

[stj]

i need as much as possible, i have a battery damaged board i'm fixing and then i want to make a remote "control-box" with rotary encoders & stuff. 

[Andy Watson]

Ok, I'll dig-out my notes ... but it may be tomorrow before I reply.

The GPIB board essentially "sits" on the data cable from the front panel and provides 3 channels of DAC controlled voltages to define the output voltage, current and over-voltage levels. The actual switching between GPIB and manual control is performed by the logic on the main control board.

[Andy Watson]

Here are the notes and schematics that I made about the APxxxx GPIB interface board. The firmware is based on a Z80 processor that is running FORTH.
The EPROM also contains the following unreported strings:
AP20-80
AP70-30
AP400-5
AP30-80
AP60-50
AP100-30
AP60-150
AP100-90
AP30-250

[stj]

thanks,
what's your firmware version?
i have eprom v1.5
(aphibus2 ap60-50 v1.5 3-7-92)

[Andy Watson]

The device identifies as "WAYNE KERR,AP70 -30,0,1.6"
The EPROM identifies as  "09-04-96,V1.6,4"
The label on the eprom says "AP7030 V1.6"

[stj]

interesting, did you dump it?

[Andy Watson]

interesting, did you dump it?

Yes. See attached. As far as I can tell there is no method of coding the configuration in the hardware which probably means that the AP7030 identy is baked-in to EPROM code. The battery backed RAM apparently only stores the voltage and current settings - there's no calibration or identity settings stored.

[stj]

i wonder if it's only the text string that differs or if the voltage & current limits are hard coded?

[Andy Watson]

These power supplies cover different ranges of voltage and current so there must be some re-mapping of the programmed settings (which is enetered directly in volts and amps) to the full, 12-bit scale of the DACs. I.e. there's a bit more to changing the identity, but probably not a lot.

If you know Forth you could probably unpick it easily enough. Disassembly of the first two-dozen bytes of code will land you in the threading code that is the heart of the Forth interpreter. It is also quite easy to find the start of the linked-list of Forth words and re-constitute it. Unfortunately my knowledge of Forth stops there - I don't have the detailed understanding to take it further.

[stj]

well that's more about FORTH than i know!
i jumped straight from BASIC to ASM in the early days, stuff like FORTH and COBOL seemed like a waste of money.

[Vannystick]

Thanks for the additional thoughts coromonadalix, the house was rewired a few years ago, its a little industrial but meets modern regs.

I measured out the "TX401 Auxiliary Supplies" voltages,

Marked Voltage    --->   Actual Voltage
20v    --->   18.6v
25v    --->   23.4v
10v    --->   11.16v
12v    --->   11.95v

Then while looking through the manual, with the AP switched on (output not enabled), one of the RIFAs let go in there usually exciting way! So one step forward, many steps backwards.

I'd best go put it in the shed while the stink leaks out.

[Vannystick]

Wow, can't believe I haven't updated this in over a year. Sorry.

Well, after the all the RIFAs where replaced on the high voltage side, it seems to be behaving its self. I've not really pushed it since it was rebuilt, but low 10 to 20A loads have been fine.

So, £25 later and i'm a happy chappy!

[Andy Watson]

Splendid! I do love a happy ending

[madao]

I owned a  AP60-50 with Farnell-label since few years.

It was dead, short on mains line.
Dead rectifier is why, fuse blowns.

I have replaced  rectifier and large   capacitor with  PFC module (2,8kW, 340V  DC, autoselect.) . It make it  more worth and it is better for mains.

My Farnellchen (a  german words , it means "little farnell") has  GPIB cards, i can dump  EPROM. Wait on their commando.

regards
matt

[Robert763]

Old thread, but may as well keep the repairs on these in one place.
I dug my AP50-60 (Farnell branded) out of the back of the garage where it has sat since we moved here  This was prompted by work wanting to borrow it. So plugged it in and switched on and it lit up with output LED on, UNREG LED on and voltage about 30 V and current about 25A 

Had to do a road trip that night so no work on it but thought about the problem.
Interesting that the meters were at half scale and would adjust above 1/2. This immediatly made me think of negative supply failure on the analog control. Circuit shows a -15V generated by a 7815 (yes the positive version, not a typo). Got the covers off  and GPIB board removed  for access and there is a bank of TO220 devices. Quick voltage check, +12V on C30,  0V on C31 (the two yellow axial caps in the photo). Power off and a resistance check shows 0 Ohms across C31. Whip the board out, Lift one end of C31 and it's a dead short. 100uF 40V aluminium electrolytic. The manual, which is branded Wayne Kerr, shows 100uF 16V. I put a 100uF 25V in. Putting a 16V electrolytic across a 15V supply is asking for trouble. I also replaced C30, the only other cap of that type, which is across the +12V regulator output.
Plugged it in and it works